You all know about solder, most of you use it all the time. But there are a lot of kinds of solder now, and a lot of soldering accessories.
Have ample supply of dry rags, and/or a holder with a damp sponge to clean dross and filth from the tip. I prefer cotton rags over the damp sponge
There are many shapes and sizes of soldering ironn tips. You may be surprised by which ones are better for what. Usually I am perfectly happy to buy cheap crap stuff from importers, but a good soldering iron is worth some extra time (to find a used weller) or money (for a quality new unit). Contemporary irons are much worse than decades old ones available second hand for a given price - in terms of tip lifetime, build quality, and reliability. I would take a weller with a Curie Point tip over a some newfangled one. (Though I prefer the Weller irons with the adjustable temperature control). Sooner or later even with the most abuse resistant irons around, you're going to need to replace the tip - but as you move beyond simple through-hole soldering, you'll want to have multiple options at hand.
It makes a huge differece to have the right tip. A good iron will have high-temp plastic on the nut to unscrew the tip holder, permitting you to change the tip without waiting for the iron to cool. Be sure to dump out the tip onto either a damp sponge of a scrap PCB or other heat resistant material. It leaves scorch marks on an unprotected desk. If using a weller iron, I recommend using real weller tips, from, for example McMaster Carr, not the cheap knockoffs.
- Conical and nearly conical (often called chisel) ones are great for through hole work. Although youd think the small ones would be excellent for SMT, soldering SMT parts with more than 2 pins one pin at a time is a losing propeosition. You need to use a larger tip and learn the technique called drag soldering. It looks like magic, but is far easier than it looks! There are many vieos on YouTube.
- Knife-shaped ones are spectacular for drag soldering of SMD parts with closely spaced pins. While it may seem counterinutitive to get a 3/8th inch wide tip when working with 0.5mm pitch pins, you don't want to be soldering the pins one at a time. Soldering one pin at a time is a losing proposition even on SOIC, much less QFP, QFN, etc. The knife point is also highly effective for rows of pin header, as a similar technique can be used.
- Cylindrical/Bevel ones with a ~ 45 degree angle at the end - are great for SMD drag soldering where more precision is needed and for rework - you want the ones with only the face tinned.
- "Gull wing" and similar ones designed for drag soldering. Like the beveled one described above one but ith a depressio nin the middle of the bevel to hold some solder. I have not beeen impressed with these - they carry more solder than I want and I end up having to use braid to remove some.
- 30 degree round - 30 degree bent tip ostensibly for precision work. These seem to get eaten away be the solder rapidly (poor job of iron plating?) do a ranther poor job of gettign the heat you need to the tip. Looked very promsing. But alas, did not measure up.
- Precision needle shaped ones - sounds and looks great, does poor job transferring as much heat and has the same problem as the ones with the 30 degree bend.
I use a Weller WSD81 soldering iron (and love it, and a contemporary iron sits next to it unused); for this specific iron I grabbed the following tips, mostly from McMaster:
- LT KNSL - the knife shaped ones. This is usally whats in my iron. Gives you a sharp point for semi-precise use or a broad flat side for drag soldering, clearing bridge melt solder then drag tip away from part), and lets you remove most SMT resistors and ceramic capacitors with ease, because you can touch both ends at once.
- LT 22CP, 33CP - tinned face only bevel. Amazing for drag soldering. 22CP is the more useful imo.
- LT 1L - long conical tip for precision work.
- LTS, LTA, LTB - your basic soldeiring iron tips for through-hole and wire-to-wire, wire to board work.
- LT4 - not from mcmaster, but sold by US sellers on market sites. good compromise of precision and heat transfer
- A second tip holder to expedidite changing tips (at least for LT series Weller. AliExpress has 'em dirt cheap and they're indistinguishable)
- Get one extra broad chisel tip, these are helpful for dumping large amounts of heat into challenging joints.
And the ones I don't like much
- LT 1S - as noted above, poor job transferring heat, amd short lifespan.
- LT 1SX - as above with 30 deree bend. These look nice and precise, but don't transfer heat well and seem to have a very short short life. I barely used mine and it was already starting to lose material....
- LT BB, CC, DD - Bevel tips... but tinned some of the way up. Retains, then transfers to the board, too much solder more often than not.
- Gullwing styles - I really wanted to like these. They're similar to the bevel tips only with an indentation to hold some solder, but the tinned-face-only bevel tips seem to work better for the same tasks. These just want to transfer too much solder.
- Do not expect the same quality when you buy cheapo tips from China - they usually don't last as long, especially the way I abuse them. As I said above, your iron is your most important tool, and the tip is the focus of it.
If you are lucky enough to have one of theose teal green weller irons (the later models with more conventional colors are disappointing - they were bought out and quality slipped; ebay is your best bet for the good ones, unless you get lucky buying one second (or more likey 3rd, 4th, or 5th) hand at a brick and mortar used tool store like I did. A good set of tips for the WSD81 that the damned fool at said store sold me for $30 wound up costing over $100. Both were worth every penny and then some), I recommend buying quality tips from McMaster Carr if you have a Weller iron. The prices are high, but your soldering iron is the most important tool you will own. Contemporary irons are either eyewateringly exopensive or of miserably poor quality, with crappy tips with short life times. Usually weller tips can be abused in ways that would make a cheap tip cry. I've done all of these things to my Weller tips without any problems....
- Leave it on overnight at least one night a week.
- Turn iron on. Use. Go on vacation for a week, come back to find iron still on. Wipe tip on tamp sponnge, add some solder to it and begin fixing the electronics I broke while on vacation.
- Use as heat-source to melt through inconvenient plastic parts (wipe often with a rag and keep re-tinning with just a touch of solder)
I do these things ALL THE TIME, leave the iron on more than 50% of the time (the auto-shutoff failed ages ago - on the WSD81, the intensely annoying auto-shutoff will usually fail if you leave it on over night, then immediately try to use it, after which you are freed from the "feature" that turns off the iron right when you've got everything lined up to use it). In spite of all this abuse, I go through at most 1 tip per year! The weller tips are built like tanks!
See below under paste for pros and cons. Only suitable for SMT parts, but a remarkably expedient way to solder them, and the easiest way to rework single sided all SMD boards. Just be careful - it's very easy to overheat the board with a hotplate and discolor it. For rework, one can also set it to a lower temperature and then use hot air or an iron to heat the part the rest of the way You can get two tiers of quality withoug busting the budget:
-
$12 or so gets you a fixed (too hot) temperature one, with thermally conductive metal legs that leave char marks on he surface of the table (I found I needed 3 layers of FR4 to keep the area under the legs from discoloring). These nonetheles permit rework that you could not dream of otherwise! You need to work quickly if you don't want to overheat the board and turn the white solder mask and silkscreen or silkscreen sickly salmon color.
-
For around $65 shipped, I have been extremely happy with a UTH-945C from aliexpress. Make sure you get the right voltage for your country.
There exist two routes towards an affordable reflow oven. They both have their problems.
You can modify a toaster oven. This is very cheap, and kits are available. They usually need a booster heater, and the whole package in in the neighborhood of $100. The results are not so great though.
These also suck, but significantly less. The T962 is the one people usually use. Be sure to read up on the steps needed in order to make it usable and be prepared to implement them. At minimum, some tape inside must be removed and replaced with kapton, and the miserable stock firmware replaced with slightly less miserable third party firmware. The firmware can be updated with a serial adapter. I found that putting small round-headed screws into the slots on the bottom of the drawer also helped, because otherwise the thermal mass of the drawer works against you. I still have trouble reflowing SAC305 though.
Most of you know this already: Aliexpress and similar marketplace sites (to varying degrees) are what you would expect to get if you had a giant flea market, filled with people from all over the world... Many of the people you will do business are honest, upstanding citizens. But because there is no barrier to entry on aliexpress, many of the sellers really don't have any background knowledge of what they sell, so even if they're not trying to be dishonest, they may tell you things that are wrong. There are also a few straight up crooks, who are extra-visible because they know how to game the system either by selling (inferior and misleadingly described) items for less, and/or through a deep understanding of SEO. This is very much a "Caveat Emptor" situation - the dishonest sellers can get away with almost anything, and they know it. The sellers at times appear to not know (or pretend not to know) what "Net weight" means (See, it's totally 50 grams of solder paste - he even shows that in the video, he puts the jar on the scale and it says 50g.), and sometimes it seems that "lead free" means "lead, free! (with the purchace of the rest of the solder)". Don't forget that this is the same marketplace where folks shamelessly undersize wire by as much as 6 AWG (sometimes right next to the accurate - or less-inaccurate - column showing the metric measurement... it pays to size check the english and metric sizes against eachother; If they're not equal, you can count on it being the smaller of the two. I've bought from several vendors on aliexpress and found the shortfall to range from 0 AWG (this is the norm for FEP insulated wire, which seems to cut out the riff-raff from the manufacturing chain. The PVC wire typically has false AWG specs printed on it, with extra thick insulation to make it look like the advertised gauge), to on one occasion as much as 8 AWG, which caused failures in he field. Some of this fraud is willful, but much of it is simply negligent; sellers are often unaware that the wire they got, which says 24 AWG, is actually 28 AWG with thick insulation, that their "Lead free" solder isn't, or wheather a batch of counterfeit Amtech flux (all the Amtech flux on Aliexpress is likely counterfeit) is any good or not (usually it's fine, but occasionally it is utter garbage).
If working with SMT parts (and, to a lesser extent, in general), you will sometimes need to add flux. I like Kester RF741 most, but... the difference between a $30 tube of that, and a $2 tube of random flux off aliexpress is pretty small. Compared to the differnce betweeen brands of solder paste, the differences are tiny. Watch out for solid rosin flux which looks in pictures like it could be a liquid, but is actually a rock hard solid - this comes in rectangular blocks and metal tins (paste may come in plastic tins or syringes, I recommend the latter). Whenever you use gel flux, even though it says no clean.... you're going to want to clean the boards. Use 99% IPA and alcohol on a toothbrush, wipe off with rag, then wipe with clean IPA.
Do not use any flux or solder paste that is not described as no clean. There exists "no clean water washable" solder paste from ChipQuik (it's below average flux, unfortunately, and the "water washable" claim is a stretch; I was profoundly disappointed) - that's not what I am warning against with this paragraph though. There exists flux (and solder paste) which must be cleaned and removed completely, otherwise it will corrode the board and ruin everything. This type is euphemistically called "water soluble" or "water washable" without the words "no clean". When they say water, they don't mean tap water without any additives, and when they say soluble, they don't mean readily, they mean in an ultrasonic cleaning bath. I recommend leaving that stuff to the pros (who almost universally use "no clean" as well, because having to remove every trace of flux from the soldering process raises costs, and modern solder, flux, and PCB surface treatments generally render those more aggressive fluxes unnecessary.
If you read the above, you're using "no clean flux" - so why would you care about removing the residue? In the case of the solder paste that is part of the solder, after running through a reflow process, further cleaning is rarely required - the flux is designed to mostly burn off during reflow and leave behind only inoffensive residue. However, other soldering processes (hot air or iron) leave more residue behind when used with solder paste, as does any additional flux applied during rework. To put it simply "no clean" means only that not removing the flux will not damage the board. It says nothing about any extrinsic issues that may result. As these fluxes advertise, they are tacky, they are not water soluble, and they don't readily dry out. The board will have sticky goo on it. I have seen boards which were not well cleaned before being placed into a static bag subsequently adhere to the bag and become difficult to remove. When you handle the boards, you will get sticky gunk all over your hands. And even if the board is not going to be touched, dust, fur from domestic animals, and so on will stick to the leftover flux. In addition to being unightly, this can indirectly contribute to board failure. So whenever you use solder paste outside of reflow conditions, or when flux is added for rework (which is nearly always necessary for reworking anything that failed to reflow), you should take the time to remove it. Note that this is not an issue with wire solder, which uses the equivalent of that solid rosin flux - it is neither sticky, nor is it easy to remove.
You can buy "Water for cleaning circuit boards" on aliexpress - but it is quite expensive after shipping. It's a very interesting kind of "water" indeed: it reeks of organic solvents, is extremely volatile, and as you probably suspect by now, highly flammable. That is to say, it doesn't contain water. Some of the brands which list their ingredients are simply denatured 95% ethanol that you could get for a fraction of the price in the paint thinner section of any hardware store, others are considerably less benign (I could never identify anything other than acetone, xylene, ethanol and IPA by odor - some of the brands are just IPA or ethanol, but others are something else that is none of those 4 solvents). Likewise, you can buy proprietary flux removing cleaners from western suppliers, at even higher prices. I don't recommend it, especially those "water based" ones - what they don't make clear is that the additives to make it dissolve the flux also make it highly toxic and corrosive to the skin, and need to be entirely washed from the board after they've removed the flux. Many of the organic solvent based ones are simply overpriced denatured alcohol or isopropanol. The claims of reusability are also deceptive: If they work, then the flux dissolves in the solvent. It doesn't take much dissolved flux to make whatever it's dissolved in leave leave a sticky film behind, which is exactly what you're trying to get rid of. I recommend using denatured alcohol (hardware store) or preferably 99% Isopropanol (IPA), available from many mailorder companies (including Amazon). In a pinch 91% can be used, but 99% is ever so slightly better, is priced similarly, and will retain effectiveness even after absorbing a few % water from the air, and won't leave moisture behind. I highly recommend buying a spray bottle of 91% IPA from the local drugstore, refilling your bottle of 91% from that and filling the sprayer with 99% (you can't use most spray bottles - the IPA breaks down the plastic). This can be used in conjunction with a "dirty" bottle used to dip a brush into for removing the bulk of flux, followed by spraying with the clean stuff and wiping it off with a clean rag or paper towel. (if you let it evaporate to dry instead of wiping it dry, you still end up with a sticky film on the board!).
Rosin flux as used in solder wire is not sticky - and is also very hard to remove without leaving behind white residue. Since it is not harmful to the board isn't sticky, that kind of flux residue can be left behind without concern in most cases.
I have been disappointed in liquid flux (from CML electronics and manufactuerd by the reputable Western company Kester). The flux is dissolved in alcohol, and the flux tends to form a film on top of the alcohol as it dries, so it takes a surprisignly long time to dry. If you don't let it dry, then it will splatter as the hot iron boils off the alcohol. And it has all the problems of gel flux, and you'll still want to remove it for the same reasons. So my advise it to just use gel/paste flux. It's not clear to me what advantage it is supposed to have over gel/paste flux (my understanding is that the main use of it industrially is for wave soldering)
Some people love flux pens. It may be okay if you don't do very much rework, but the price per unit of flux is exorbitant, and some are prone to clogging or drying up.
Just buy it by the 10-30cc syringe or 100-500G tub. The former is more convenient, while the latter is much cheaper - though often of inferior quality to what can be found in a syringe.
It's hard to tell the difference betweeen most brands of flux as long as they are neither flagrantly terrible or or notably outstanding. There is one general trend with flux that holds almost universally - it's based on the color of the flux. Yellow flux contains rosin. The rosin based flux usually works better. However, that rosin-heavy flux is harder to clean off the board. Flux that is more white in color is generally easier to clean - though also a little less effective. They also now make clear (Mechanic SD360) and nearly clear (UGAIN UG78) flux. Both of these are very easy to remove - and they don't turn into unsightly brown crap like some kinds of flux. They also generate less smoke than normal flux. They are markedly more expensive than "basic" paste or gel flux (though still cheaper than major Western brands), but they aren't quite as good at at cleaning off oxides and making disobedient solder behave as more aggressive rosin-based flux. On the plus side, you can more clearly see what you're doing, which you can't do if your flux was opaque to begin with and turned a dark brown during use. It's no fun to think you've fixed a board, then when you clean it off, find that you missed a bridge or created a new one.
That said, color most certainly isn't everything: all the fake Amtech flux looks the same, but I have gotten a bad batch ("Scamtech?") that had no fluxing activity whatsoever. I passed it along to another electronics enthusiast to test and he confirmed my assessment, it looked like flux, smoked like flux, and made a mess like flux, but didn't improve soldering, so it's not just me. And there are most definitely yellow fluxes that don't work any better than the rosin-free ones. They're still harder to clean up though.
That's rosin - it's mixed with somethign to soften it to make rosin-based gel flux, and it's what's inside rosin core solder wire. It's not the most useful form of flux. The tip of the iron will melt it, but this is rarely especially useful. I've also tried chipping off little pieces of it and applyign them where needed. It would be nice to have powdered rosin like this, instead of solid blocks. Crushing solid rosin flux actually sometimes works VERY well, being both highly effecive, surprisignly easy to clean with IPA, and
That's because it is a vape pen, being mis-sold to dodge the aliexpress ban on selling vape pens; I dont belive it to be of any use for soldering, unless you like to solder high (not receommended). This sales practice has become so widespread, though, that I think some sellers don't realize it.
Appendix A lists some flux brands and my comments on them.
A little bit of technique can make a big difference:
- Insert pin header into header of the opposite gender, double row on the side the pins will stick out from, to hold it in place while soldeering to ensure it comes out at the right angle. Knife tip can be used to .
- When using an iron to solder 2-terminal SMD passives, put a bump of solder on one pad. Solder the part to that and press it flush to the board with tweezers. Do this for all such parts, then go through them a second time to solder the other side.
- When using an iron to desolder 2-terminal SMD passives other than LEDs (and rarely LEDs), use a knife tip, and come in from the side, so you can touch both sides at once and easily remove the part.
- When using an iron to desolder 2-terminal SMD LED, accept that the LED is toast. The previous method doesn't work very well. What does work is to press a hot knife tip on the plastic part of the LED, which will quickly disintegate. With the iron on top of the part under the plastic you will now hit both ends at one and can easily remove what's left of the LED.
- Sometimes, particularly with almost-all-tin (traditional) lead free solder a component with large thermal mass, soldered to the ground plane (for example, an edge-launch SMA connector) you may have trouble getting enough heat into it with just an iron. A heatgun which might struggle to melt the solder alone works miracles for pre-heating the solder paste in combination with an iron that similiarly struggles.
This is the most important skill for soldering SMT components with more than 4 pins - from SOIC to UQFN. There are two challenges to all such parts, and one specific to QFN and parts with exposed pads. The first is the fine pitch of the pins, which may be as tight as 0.4mm, without forming bridges, (and of course making conect with every pin). The second is alignment of the pin and pads (this, surprisingly, is harder on TQFP than QFN, and larger QFN's are easier, while smaller QFP's are easier, all else being equal, due to the self-centering effect of the exposed pad on almost all QFN packages). The third, applicable only to parts with exposed pads, is getting heat to the solder under the part where it's needed to solder that center pad.
- Tinned face only bevel tips are ideal.
- Knife style tips are nearly as good.
- If neither of those is available, you want something with a broader tip, not the super fine point you would expect to want. Chisel tips are a usable substitute.
- Conical tips are the least suitable that will work.
- Fine point tips are not for this type of work and shoulld not be used.
- Clean the soldering iron tip on a damp sponge, or by wiping on a clean cotton rag. a. If you are, or may be (because you don't remember) using a different soldering alloy than you last used the tip with, if there's any chance that it may have been used with a bismuth-heavy solder when you're using leaded solder now, or vise-versa, melt some of the new solder onto the tip, wipe it off, and repeat. That ternary lead-bismuth-tin eutectic is a real bitch when it forms, and causes parts to fall off during rework of nearby parts or even during operation at higher temperatures.
- Secure the part in place. Do not worry about forming bridges yet. a. If using paste, hold it in place with tweezers after applying paste to the pads, and touch the iron to one corner. You're expecting to melt solder on several pads, which will likely be bridged b. If using wire solder on a largeish part, or any leaded one, hold it in place and tack-solder one corner. c. It is key that, with either method, you can easily melt the whole blob of solder you just put down without moving the iron.
- Check the alignment of ALL sides of the part, and ensure that pins are lined up with pads (they likely won't be). Melt the solder and nudge the part with tweezers to correct this. After doing so double check all other sides. A part with 4 sides with pins on them can have only one side misaligned: There are three axes in which the the package can be misaligned - it can be skewed in the X or Y direction (I've definitely had TQFP48 parts with the whole chip shifted one pin come out of the reflow oven. You can make excuses when using a reflow oven, but when soldering with wire, that just means you screwed up), but it can also be rotated, and the centerpoint around which it is rotated can be anywhere. It's that last point that makes it possible to have "one bad side" of a QFN/QFP chip: The center point of the rotation is near the opposite side.
- Once you've got it aligned perfectly, if using wire solder, apply tacky paste flux along all the pins, being careful not to move the part out of alignment. a. If you're using paste, it has enough flux in it for this step, though you're likely to need more later.
- Once this is done, starting from the opposite corner from the one you tacked down, holding the iron tip at a low angle to the board (bevel and knife tips mostly do this for you), drag the iron along the row of pins. You should not be applying pressure to the pins, as this can bend them. Still don't worry about bridges - however if you see bridges forming when using wire, reduce the rate at which you add wire.
- Repeat with the two other sides you didn't tack down. a. If a side came out completely bridged, you probably have too damned much solder on the iron - wipe most of it off and be more conservative with solder.
- On the side you tacked down, you don't have to add as much solder. Start from the end that you tacked down.
- Depending on the pitch of the part, the solder being used (and to a lesser extent the flux, but on boards and parts in good consition, it's hard to tell the difference between fluxes based on how well they "flux"), the part may now be well soldered along all the edges. Examine each edge in good light, so that you can see the solder fillet and any bridges. On your first few times drag soldering, you'll likely have a lot of bridges. Now we clear them. a. If you have a side with Way Too Much solder (eg, there's one giant bridge connecting every pin) add a bit of tacky flux, apply desoldering braid, and gently press with the iron - you should see the solder spreading into the braid. Remove the braid and iron (at the same time, otherwise you'll have soldered the braid to the row of pins). b. For smaller bridges, usually adding a small amount of flux if most of the original flux burned off, and dragging the tip of the iron either away from the chip or along the row of pins will clear bridges. If you drag away first, that clears bridges that aren't caused by an excess of solder. Dragging along the row can help by distributing excess solder, or will leave you with a group of pins in the corner that remains bridged. Use your desoldering braid and flux.
- You should soon have a properly soldered part, at least on the outer rows. If you are using a board that has a hole under the exposed pad, that can be used to heat the solder (preferably paste) there to solder that as well.
Important Do not drag desoldering braid along the surface of the board using the tip of the hot iron. You will scrape off solder mask!
The pitfalls:
- Difficulty gettting heat to the middle of the EP. This is a serious problem when not using reflow soldering method. a. If you're designing the board, try to use a footprint with pads that stick out further than usual, and are not more than half the pitch in width (this helps ensure the parts does't go on cockeyed by half-a-pin shorting every pin together). Like the ones on my breakout boards. This makes it easier to get heat to the point under the chip where it's needed - and also lets you drag excess soldeer onto that longer pad and reduce the need for braid. A PTH in the center of the exposed pad (large enough to not get tented) provides an excellent way to solder down the exposed pad in the middle of many QFN and similar parts. For parts with pins on only two sides, and an exposed stripe in the middle, make the stripe longer, allowing you to get heat under it by holding the iron against the stripe on the board where it sticks out from the end of the b. Failing that, a heatgun in conjunction with an iron makes a big difference (look closely at those youtube videos that make everything look easy - you often see the flux start to move before the iron touches the board - that's from the hot-air gun outside of the frame of the video and which they don't mention. c. Finally, or in combination with those methods, lower temperature solder can be used. Solder of good quality is is readily available in tradtional lead free (217-228C depending on specific alloy) - this can be challenging to work with, particularly where you don't have long pads to help conduct heat. 181-183C Tin-lead eutectic solder (traditional leaded) is much easier to work with (the 181C stuff has a touch of silver in it, which ever so slightly improves it's properties), and 138C tin-bismuth eutectic (138C is very low temperature - it's not appropriate for general use unless you know the board will neither generate nor be expposed to temperatures in excess of around 100C. From Chinese vendors, a variety intermediate temperature solders are available, though the quality varies widely. I also include the mysterious "183 C lead free" alloys in this category. They are discussed at length below.
- Alignment issues require either good hot air setup or a hotplate to rework in most cases. It sucks. You want to avoid it happening in the first place. the first thing to know is about: "self-centering". Surface tension will usually pull the pins onto the pads if ey're pretty close. A nice big exposed pad (relative the size of the pins is a huge boon). However, the more pins on the part and the closer together they are the smaller an accidental rotation needs to be in order to cause problems.
The specific alloy makes a big difference with self-centering, specifically it's surface tension. If using exotic low to intermediate temperature alloys, results may vary significantly between brands of solder, even with the supposedly identical alloy; I suspect not all of these are honestly marked.
Now the fun part - the kinds of solder!
There are two common forms of solder that will be encountered in hobby electronics. A third kind, "preforms" are rarely used outside industraial settings and is not economical for small runs.
This is the solder that that is in the shape of wire, and comes on a roll, designed for use with a hand-held solderign iron. Solder intended for electronics use typically has 2% rosin flux ("Rosin core solder") down the middle of it, and that's what makes all that smoke when you first melt it. The flux included in the solder cleans oxides off the surface of the parts being soldered, as well as the molten solder itself (which quickly oxidizes when exposed to air while molten). That's why if you go a while working with the solder in one area of the board (because it's being uncooperative) it will get more and more hostile towards your attempts to get the desired solder joints and only the desired ones. The classical quick fix is to add a bit of fresh solder. Sometimes that will just make the problem you're dealing with more of a problem (if it's that there's really too much solder, and you're having a hard time getting it out of there). A small amount of "no clean gel flux" commonly sold in syringes is an excellent solution for that sort of case, as well as an absolute necessity for drag soldering - (which is an incredibly useful technique: it makes soldering many SMD parts faster than soldering through-hole ones!)
Solder wire is available in tin/lead alloys, and the most common >95% tin lead-free solders. Bismuth-bearing alloys are rarely available as wire at all (Some of the stuff sold as Sn42Bi58 on aliexpress isn't)
Different brands of solder wire work better or worse than others, though for a given alloy, the difference is small, unless you get really shafted like I recently did, being promised SAC305, I got the most unruly solder ever. I confronted the seller. He admitted: Actually pure tin. Since chinese solder is really cheap, buy a couple of different brands, and see which one you like better. I haven't found a clear "favorite" brand./
- Rework is as much art as science. It helps a lot to have flux on hand, but if you don't have any, while you're waiting for your $2 tube of flux to arrive - you can use fresh solder: wire solder is usually 2% flux.
- Whenever solder is not behaving as it ought to (if you don't know what I mean, put some solder on a trash board, clean the flux off, and then try to rework it with an iron), you need a bit of flux.
- Reworking is better done with leaded solder (unless the board was soldered with low temp lead free. Traditional lead free can be reworked with leaded solder just fine.
- Do not exceed 700F/300C. With low temperature solder you can go significantly lower. The lower the temperature, the less risk of lifting pads.
- Often the most convenient way to remove excess solder, particularly when the solder sucker has proven ineffective is to melt the solder,and then whack the board against the table using a downwards motion. (Don't do this if you're wearing shorts, and really don't if you're not wearing pants...). It is especially effective at removing solder from through-holes in which it is not wanted, and which can be challenging to remove. Sometimes the solder suckers just don't do it.
- Do not ever slide desoldering brain along the board, especially with the low-quality copper-plated-steel (ie, magnetic) desoldering braid. It is one of the easiest ways to scrape off solder mask, so that the traces can form bridges to the ground plane
This is a greyish sticky (messy) sludge. It is normally used for reflow and hotplate PCB assembly. A "stencil" normally made of stainless steel is typically used, which has cutouts for each pad (made by laser cutting). It is positioned precisely, and a flexible spatula-like implement is used to spread solderpaste over it. Small scale operations typically use old gift cards as spreaders, and some use cheap laser cut polyimide stencils instead of stainless steel. Polyimide stencils are far from ideal, because they are too flexible, and as a result solder paste ends up getting under them, and makes a mess out of the whole thing. Cleanup is with IPA, (see below) and you want to wipe down both the desk and the stencils. Wipe the stencils clean with fresh IPA and paper towels as the final step. Clean them within a few hours of use, don't let it dry on them.
Once the solder paste is applied, parts are placed (with a pick-and-place machine, or by a patient person with steady hands and good eyesight ("Human Pick and Place" - it's not as awful as you'd think)), and the PCBs are placed in a reflow oven, or on a hotplate. Both methods work well. The hotplate method is much cheaper, though it doesn't scale well, and it's easy to overheat the boards and discolor the silkscreen (this would also damage sensitive parts); The cheaper hotplates always get much too hot - not to mention the fact that you have little control over the reflow profile.
Plan ahead!
Examine the board fully once it is cool, and figure out where there are things you need to fix. Make sure you have all your replacement parts at hand. Check for crappy looking solder joints. Look at it under different lighting angles. Apply flux to all the places you're going to be reworking. You will want to work as quickly as possible. Don't the crank temperature up to make the solder melt faster. Try to do one board at a time, not a whole bunch.
Yeah, no, it may seem like a good idea at the time, but it's usually not - unless you have a stencil (and if you do, where is your hot air, hotplate, or reflow oven?!), you're better off with wire solder. Of course, if not using the one of the big three classes of common alloys you have only one choice - paste). Paste with an iron tends to leave a lot of sticky flux behind.
- A bead of paste solder at room temperature can be applied to pin header after it's inserted in the holes and held straight with a jig. With board held vertically with the bead on the higher side of the pins, heat from the middle-lower side as you drag your "knife" soldering iron tip, (you do have one right?). Dragging the iron tip along the pins and melting the paste is faster than other soldering methods, as long as you avoid a puddle of solder forming between the plastic part of the pin header and the PCB, which happens if either the board is oriented wrong, such that solder under the influence of gravity wants to flow through the holes, or you try to go too fast). Surprisingly, the best fix for that is usually just leaving the iron in contact a little longer - the problem is caused by unmelted solder that became more liquid as it heated up, such that it flowed around the pin as solid solder and flux, and the heat never reached down there to melt it, which would cause surface tension to pull it back around the pin where it belongs.
- We've all gotten them: Poorly-made-in-china circuit boards with USB connectors on the board, and we made sure to get boards with the good kind of connector, with structural pins extending through the PCB. Almost invariably, these boards are poorly soldered - the structural pins are barely soldered if at all. I recommend solder paste rather than wire - I've gotten better results - though either way works. (Cheap USB hubs also suffer from similar problems; I have seen some where only 1 pin of each connector was soldered... and appeared to have been soldered by hand... I would use paste for the mechanical pins on connectors oriented horizontally on the board - but wire for connectors that stick up vertically)
You may have seen videos of someone printing paste from a syringe onto just the pads. That method does not work very well (if you were smushing the paste around before squeezing it through the syringe, it would work better, but you're not). I found that it sorta-kinda worked okay, as long as I kept the syringe perpendicular to the board, and let the solder paste come to room temp first... but the result was not really any better than smearing a slightly too-thin layer of paste smear over the board and placing the parts on that (works very well considering how halfassed it is - yeah, there might be a few more solder balls on the loose, but you were going to remove the flux with a toothbrush and 99% IPA anyway, since you'll inevitably have a few bridges that need to be reworked, which means using paste flux, which means the board will be sticky if not cleaned.
Only with solder paste, the differenmce is considerably larger, as the flux composition varies between brands, the care with which the paste has been handled will vary wildly, and the "mesh size" is not always specified (and sometimes may not be specified accurately). Basically all chinese solder paste claims to be #4, regardless of the actual particle size. Most any tin/lead paste or 99/0.3/0.7 or 96.5/3/0.5 lead free (assumimg your reflow oven can melt it) will work fine for 0805 passives and SOIC packages. Where it gets dicey is QFN, fine-pitch QFP, TSSOP, DFN, and similar fine pitch parts. With good solder paste, these can be done without pain. With some particularly disappointing solder pastes, you'd be better off not soldering a fine-pitch QFP or TSSOP in the reflow oven, and then coming back and drag soldering it with better solder afterwards.
Some observations (See also tabel below)
- Conduction brand is poor in general, though usable for non-demanding parts.
- Mechanic seems to be available in 50 different shapes and sizes of container. I suspect they are being packed at different facilities with different levels of care, but I've never found any that I'd rate as above average, and few of them are even average. Steer clear of the common orange-and-blue jars in particular. Unlike basically all the other Mechanic pastes, they admit that that stuff is #3 paste, which is not suitable for fine pitch components. At least the label honestly states the specs. Their 148C and 158C solders are "mystery alloys" with no indication of whats in them, and the 148C stuff embrittles copper.
- Relife makes good tin-lead solder, and their ROHS 158C solder is also good. Unfortunately the traditional lead-free they sell is the low silver variety (0.3% silver) that melts at 228C instead of 217C. Aside from the irony of being named Relife while packing leaded solder in packages with green labels (they use other colors for their other alloys), I've found it to be the best of the non-silver bearing leaded solder pastes. They make paste in 4 compositions: 138C SnBi eutectic, with blue label, 183C tin-lead, with green label, and 227C lead-free (only 0.3% silver, making it poor solder that requires high temperatures and has poor properties) with a red label; I haven't tried the 227C stuff stuff. Finally, in collaboration with "GLON" they sell a claimed ROHS solder with a claimed melting point of 158C. Whatever alloy it is, it's quite good - it solders very well with hot air and reworks easily. Relife or their parent company may actually be the one manufacturing the first three, but it's implied that the 158C stuff (composition unmarked, unlike their others) is made by their partner company.
- KEK markets a wide variety of paste fomulations. They consistently mark the composition of the alloy, and labeling is professional, and the paste is of generally good quality. They make 138C, 151C (64.7% Sn, 35% Bi 0.3% Ag, quoting the solidus as the melting point), 181C (Sn62.8,Pb37,Ag0.2), 183C SnPb, 217C SAC-305, 227C Sn99/Ag0.3/Cu0.7, and highly specialized high temp 296C Pb92.5/Sn5/Ag2.5. I've tested the SAC-305 (It ain't as good as ChipQuik's SAC-305, but you pay less than half the price for about 4 times more paste. The 151C paste, however, was mediocre at best. I hope to compare the 181 and 183 paste in the future, and see what the difference in properties is while controlling for the manufactuerer.
- PPD's logo appears on Conduction's garbage pastes, but also on jars of higher quality solder paste without the Conduction branding, in 138C, 158C, and 183 versions, all claiming to be lead free. The 183C is definitely not plain old tin lead, either (though it seems to melt at less than 183C). I am waiting on a bottle of the 158.
- I have tested UGainGreat brand (as I like their flux) - I found the 158C solder to feel dry and come out looking like typical low-grade off-eutectic SnBi, with dull joints, difficult reworking, and little to recommend it. I don't think I'm going to try their other products.
- MaiJing manufactuers 148 and "189" solder pastes. The "189" stuff melts at a lower temperature than claimed but higher than most of the off-eutectic SnBiAg alloys. Joints were still rather dull.
- Amaoe sells 138C, tin/lead with and without 0.2% silver, and traditional lead free pastes. I gave not tested any of their pastes for a few reasons. First, their lead free is SAC105, inferior to SAC305. Almost every listing also shows a picture with a red flag: The composition of their SnPb solder is listed as Sn63 Pd37 Pd? No, they most definitely mean Pb. They are not selling solder paste made with 37% palladium (besides that it couldn't be a solder - Tin does not readily dissolve palladium, which is unsurprising since it has a melting point in excess of 1500C and very different chemistry), it would never be used at 37% in something that sells for $5 or even $500 for a 30-50g jar. Pd exorbitantly expensive, being a platinum group metal with a market price of $60-70/g), which if I'm not mistaken is the third highest
*price of any naturally occuring**metal. The fact that they so prominantly display a composition with such a basic error is a sign of sloppiness and ignorance on the part of the manufacturer. - WNB has a different approach that takes laziness in labeling to a new extreme: They sell 148, 158, 183, and 217C solder paste, all stating the alloy as Sn/Pb/Ag. Their official store doesn't seem to have a clue what the actual composition is either. It's no secret that the companies packaging solder paste in small tubes don't actually make them paste, but rather buy it in bulk from industrial suppliers and repackage it; I suspect that this is the case here, and that they know little about what they're selling.
- JPT's SAC305 paste looked great in photos, in on examination of the paste, and the labeling was well done and professional - in fact it seemed perfect until I tried to use it. The flux lacks activity, the melting point seems too high for SAC305 and it was a huge disappointment.
- I have 158 C and 183C "Laster" (which is sold at a very attractive price point) and 158C WL2003 Wylie brand pastes coming for test.
* Based on july 2022 spot prices, Palladium is behind only Iridium ($150/g, used for ultra high temperature crucibles for making single crystaline silicon for semiconductors) and the incredibly rare Rhodium (> $400/g, mainly used in catalysis), which is highly sought after for it's catalytic properties, with a mindboggling spot price of around $450/g. Gold is #4 at around $53/g followed by the rest of the platinum group and the most sought after rare earths. most of which are FAR less expensive. With the limited supply and varying demand, these metals are highly volatile, and an examination of spot price history shows wild swings in in the prices of platinum group metals, especially those other than platinum (which is used as a store of value, which forms a bit of a cushion, since people will buy when it looks cheap in hopes of selling when prices rise) and palladium (which is as espensive as it is largely due to it's mandated use in automotive catalytic converters, and there is a robust recycling chain for those - indeed, people sometimes go so far as to remove them from cars they don't own by sawing them out of the exhaust system of parked cars for to sell to a recycler (the car owner, of course, then has to pay for repairs, and a new catalytic converter, and will end up paying several times what the crook got for the old one, making it a particularly irresponsible crime, similar to punching a hole in the gastank of parked cars to siphon out the gas, and stealing installed copper pipe or wire that is still being used to sell for scrap). In any event, the platinum-group metals market is extremely volatile, similar to indium: Iridium prices fell 50% before rising 25%, and the price of Rhodium at it's peak was nearly twice that at it's lowest price... and that's just in the the last 12 months.
** Naturally occuring as opposed to either isotopically pure (via enrichment) or radioactive materials synthesized through nuclear reactions, which are a totally different ball game, with prices hundreds to thousands of times higher due to the enormous cost of production and purification to yield small amounts of unstable material which requires special handling. Thankfully the only applications for them are in tiny quantities in nuclear medicine and as tracers, and in nuclear weapons, where the high expense and difficulty of production are desirable (from the perspective of the bulk of humanity, at least). Production of the synthetic plutonium and enriched uranium is said to be 84% of the cost of the Manhattan Project during WWII.
I use a reflow oven (T962/T962A) withs solder paste applied using stainless steel stencils (the polyimide ones suck - they flex and stretch too readily, and solderpaste gets underneath them, so the first board you put paste down onto comes out well, but after that the quality of paste deposition falls if the underside of the stencil is not wiped, particularly in the vicinity of large exposed pads, and more boards have bridges and other defects) for assembly. Solder joints are inspected visually in good light to find bridges and poor solder joints (the same method is used in automated production, and many automotive parts are made with slightly modified packages featuring "wettable flanks" which make it easier for computer vision to detect bad joints. Parts with soldering flaws get fixed where practical.
Solder bridges are cleared with a Weller WSD81 iron with an LTKN or LTCP22 tip, and flux (Max360 or UG78 for most applications as it cleans easily and is transparent, or Kester RF741 when a more potent flux is required). When there is excess solder present (this often is seen on QFP packages a solder bridge spanning all the pins on a side, or bridges reforming at the last point that the iron is pulled away) the use of desoldering braid is indicated. Some braid somes with effective rosin flux on it, but often a small amount of additional flux is needed. I usually use cheap generic flux for this purpose, as high activity is not required. That serves to effectively fix bridges and incomplete solder joints.
If a part must be removed entirely (like a part with 1 year+ lead times soldered to a board found to have a fatal design flaw; this happnened to me with a batch of 3227's soldered onto the Rev. C breakout board, which had omitted the power trace), or is misaliggned, a hotplate is used if the solder is traditional lead-free - I have had little success with hot air in that case. For solder that melts below tin-lead solder, hot air alone is sufficient, however, standard leaderd solder may reqire both hot air and and iron (again, knife tip is a must here. In my experience, fine pitch QFP packages with large numbers of pins, and small fine pitch QFN parts with smaller numbers of pins are most vulnerable to misalignment (not counting mistakes in which a part is correctly soldered in the wrong orientation, which depends more on the visibility of the orientation markers on the part and the quality of the silkscreen on the board).
All soldering using 217C lead-free solder is currently done with a hotplate, as I haven't been able to push the T962A to a high enough temperature, despite manufacturer claims that it can be done with the unmodified oven. I have only applied the "basic" mods to my ovens (improved firmware and replacement of the inappropriate tape they use with more suitable Kapton (the stock tape emits foul smelling fumes when heated; not using kapton probably saved them only a few cents), and installation of screws into the slots on the tray the boards rest on; that makes the contact area between the tray and board smaller, thus reducing heat loss to the tray. Since the heat is applied from above via IR tubes, reducing thermal conduction from the bottom of the board to the tray made a large difference). A full elecronic overhaul, including new SSRs and control boards might bring it up to a point where it could be used with 217C lead-free solder, but the modifications are extremely expensive, being in excess of half the cost of the whole oven. I have high hopes that one of these intermediate temperature lead free solders will be determined to be both effective and actually lead free; the Relife/GLON 158C and PPD "lead free 183C" are the best candidate thus far, though I have yet to verify that it is ACTUALY free of lead, since you can't trust the chinese solder vendors, and the manufacturer has been unhelpful.
These two groups comprise the vast majority of the alloys used in electronic solder. "Eutectic" solder alloys mean that there is one composition that melts at a single temperature. In non-eutectic alloys, there is a range of temperatures - the liquidus and solidus - Between these two temperatures, the solder takes on a slushy consistency as it forms a mixture of a liquid and solid phase. Generally in this range it won't wet the workpiece well, but it also won't hold parts in place. Eutectic alloys are preferable for that reason.
Classic solder is 60:40 or 63:37 Tin:Lead. People claim that the 63:37 is better. I could never tell the difference. Leaded solder is not used as much in commercial assembly (due to environmental regulations) but remains very popular in less regulated regions, and among hobbyists. It's not that the fans of SnPb solder like being environmentally irresponsible; leaded solder didn't become popular because of it's toxicity. It was and is popular because it makes excellent solder, and despite having the whole periodic table to work with, there aren't really any alternatives that compare favorably to it. The 183 degree melting temperature is spot on perfect - high enough that it's well outside of operational conditions, but cool enough that you don't have to worry too much about damaging parts, lifting traces, or turning the nice white silkscreen a sickly salmon color. The tin-lead eutectic has amazing physical properties, too - as liquid, it wets copper well, but doesn't dissolve it very well, it's surface tension helps it "flow" well and it is generally agreed to take the least skill to make a great solder joint with it. It's high ductility is amenable to making solder wire - it is soft and pliable rather than brittle. Other than the toxicity, you couldn't ask for much better.
The composition is typically between 60 and 63% tin, sometimes with a small amount of silver, and the balance being lead. They all behave very similarly in practice. Different batches seem to make brighter or duller joints, but I could never correlate that to the composition (or rather, the listed composition) of the solder. Unlike lead-free solder, leaded solder works fine without silver (though the tiny amounts of silver used, rarely more than 0.4%, does help - while you don't need the silver, silver bearing solder is virtually always ever so slightly better than unsilvered.
-
2UUL markets Sn62 Pb36.6 Ag 0.4% solder, claiming an anomalously high 189C melting point. This is often sold claiming to be be lead free, which it clearly is not - but it does make good solder.
-
Other sources give Sn 62.8%, Pb37.8%, Ag 0.4% as the "right" formula for silver-bearing leaded solder, and quote temperatures as low as 181C for the melting point. Which one is right matters less than the professionalism (or rather lack of it) which these sort of desparities demonstrate in the Chinese solder market.
This is mostly tin, typically with a small amount of silver (which greatly improves it's physical properties) and copper (because straight tin dissolves copper far too well). There are a few variants on it commonly found. They all melt at between 217 to 227 degrees, and this can make it hard to solder in many cheap reflow ovens, and increase the risk of lifting the pads through excessive heat when hand soldering. In the latter case, the problem is compounded by the fact that unleaded solder just isn't as cooperative, in terms of joining surfaces when you want it to, and not forming unwanted bridges... leaded solder is better at wetting the pads and leads, has a higher surface tension, helping keep the solder where you want. Bridges form more easily and are harder to correct using unleaded solder and it just doesn't flow as well.
The situation can be made worse by the specific alloy you end up with if you aren't careful:
96.5 Sn/3.0 Ag/0.5% Cu (SAC305) is well regarded, melting at 217, and is eutectic. When bought direct from china, the quality of supposed SAC305 varies considerably, unfortunately. Note also that some sellers are dishonest in their listing titles. It is a far more common "mistake" to see a solder clearly marked in the photos as 0.3% silver as 3% silver in the listing title. Naturally, this "typo" never gets changed, even after they are called out on it. I do not recommend using any other type of Sn-Ag-Cu solder. The savings isn't worth the pain of working with the crap.
98.5 Sn/1.0 Ag/0.5 Cu (for example AMOME brand) is a compromise between the more costly SAC305, and the 0.3% silver stuff. As you might expect, it's worse than SAC305.
99.0 Sn/0.3 Ag/0.7 Cu is a much cheaper solder, and the predomenant alloy in chinese lead-free solder (wire and paste). It is markedly worse than SAC305.
99.3 Sn/0.7 Cu - The cheapest of all. The 7 ppt copper prevents it from dissolving copper traces, but everything else about it is sucks. Don't use this solder.
< 96.5% Sn - solders with more silver and less tin exist. They're not really any better, and are no longer eutectic, and hence are likely worse on some metrics. They are often marketed at eye-watering prices (with some brands being sold by the meter of wire at prices approaching that of a whole roll of normal SAC305) to the audiophile crowd who ascribe all sorts of unquantifiable benefits to silver. Audiophile means "lover of audio" in theory. As far as vendors are concerned, it means something very different: "easy mark". "You're using plain old copper for the cable between your turntable and your tube amp? You call yourself an audiophile? That must sound terrible [nothing to do with the low fidelity of turntables or distortion from the tube amps]. You need to replace it with this super cable! It's got oxygen-free copper to eight 9's (99.999999%), and it's plated with 99.9998% silver! Anyone who knows audio will hear the richer tones!". Talk like that is utter hogwash, meant to dupe people into buying more expensive wire that likely doesn't even meet the claimed specs (if it's from China at least). The same is true of solder. I've seen 4% silver solder wire on aliexpress being sold for around $5-6.... per meter.
Melting at 139C or 138C (depending on the source), this extremely low temperature solder. Tbe proximity of the melting point of very low temperature solder to the upper end of the temperature range of some parts you may be using should give one pause. If the regulator with thermal cutout at 150C die temperature overheats, and ambient conditions are warm, will you run the risk of the connection simply melting?
It is lead free, but it suffers from a number of problems. First, it approaches the ternary alloy from the bismuth rich direction. Woe betide he who let leaketh lead unto his SnBb solder. You thought the low temperature of Sn42Bi58 was too low? Try the 96C melting point that can be achieved by an unfortunate combination including a few % lead. Nothing good will come of that, and remember that most HASL board treatment is with leaded solder, and most wire solder in many countries is leaded. SnBi wire solder is rare and expensive. What's more, this solder forms an alarmingly weak bond with copper. In my tests, the surface did't wet nearly as readily as any of the other alloys... and afterwards, the metal above the tinned area could be easily removed, leaving a layer with some copper-tin-bismuth intermetallic with very little strength. It may be highly dependent on the surface treatment of the PCB
It is often used on cellphones, flexible circuit boards, and so on, where lead is not an option due to regulations, but the parts would not survive traditional lead-free temperatures.
There are a few other bismuth based alloys, sometimes d:ith a touch of silver. These are aurellard seen offered for sale outside china./ hat are available generally from China only (they are almost unknown from western brands).
I have seen and tried the following (most of them don't have the alloy listed. The sellers frequently don't know what they're selling, but that's okay, because some of the ones that do lie anyway.
- 148C (Mechanic brand) Presumably something just off of the eutectic - the solidus and liquidus are very sensitve to even a fraction of a percent silver here. But there may be more to this, because of the reason I found it profoundly disappointing: It seems to either dissolve or embrittle copper. Try putting this on wire, and then rub it with the iron. Normal solder doesn't do that! That says "Indium" but the price says "hell no not indium"
- 148C 2UUL - Slightly better flow and behavior compared to the Mechanic brand - but what really sets this apart is that they finally spilled the beans on the composition 45:55 Sn Bi. This means they're measuring the liquidus, of course. More experimentation needed to see it it attacks copper like Mechananic.
- 151C KEK brand See Sn64Bi35 below - this actually isn't all that bad though, though it does make rework unpleasant. After reflow it comes out fine.
- 158C - there appear to be two or more alloys marketed as having this melting point: One is leaded, consisting of Sn43 Pb43 Bi14. The others claim to be lead free, but the composition is not disclosed for any of them. I'm hoping to confirm that one or more of them is lead free. The Relife SP-X (G-LON Sunshine X) 158C paste seems to be very good. It has an ROHS logo on the label. But is it lead free? Forgive my lack of faith. It works quite well, though. Testing on other 158C solder pastes has had mixed results, with the two non-lead-free brands being mediocre, and the Ugain Great being something very much different from "great". Both PPD and "XG" or "laster make renditions ofthis that are not total garbage. XG/Laster also makes a claimed 183C lead free paste that works suspiciously well for lead free paste. But if it is lead free, between the low price and quality, it just may take the cake.
- "183C lead-free" Conduction brand. See Sn64Bi35 below - except this stuff is flagrantly terrible. I think it's made with inferior/inappropriate flux, or maybe they handled it with the same level of care and attention to detail as their labeling (note the numerous misspellings), and the solder had spoiled by the time I got it. It didn't melt right, didn't flow right, left a terrible finish.... and the test boards that I then tried to use had an astonishingly high failure rate over timescale of months. Don't buy any conduction brand solder paste, period. They also make 183C leaded, 138C lead-free and 217C lead free. All are sub-par.
- "172C SD520" See Sn64Bi35 below.
- "189C" 2UUL It is claimed in many listings that this is lead free. It performs suspiciously well, much more like, oh, silver-bearing tin-lead solder... and oh look, right there on the label, it says that's what it is. Very good leaded solder though...
- "183C PPD Pro Paste S600 Lead Free". No idea what the alloy is. One seller told me that it wasn't lead-free. And yet, it sure doesn't behave like standard tin/lead solder. It feels like it's got too much surface tension for normal leaded paste, and the and the solder joints are very bright, bizarrely so. It also appears to melt at a slightly lower temperature than normal leaded solder paste, and the flux smoke smells very distinctive. I don't know WTF it is, but it's not normal. Whatever it is has strong adhesion to copper, unlke straight SnBi eutectic, but so do most of the other solders here. Lead swab test appeared to pass, but I need to repeat after electrochemically corroding the solder to make it more accessible to the reagent. Also passed the crude bismuth test, wherein it is mixed with 138C bismuth solder in the liquid state, and the melting point again approximately measured - the presence of lead should be accompanied by a significant drop which was not observed. The dramatically different surface tension of this solder may have significant impacts on the soldering effectiveness. PPD Pro Paste also makes a 158C solder that appears to be be based on SnBiAg. The listing showed S240 as the model number, but what arrived was marked D50. No composition given, but it's prperties scream SnBiAg. This and Majing are the best of the SnBiAg solders that look like SnBiAg.
- "Meijing 189C" is another supposedly lead free solder with that claimed melting point and composition But the claimed melting point 17C above the liquidus of the claimed alloy makes one really wonder. They're something very suspicious about these SnBi alloys with anomalously high claimed meltig points. But it does appear to be both different from PPD (very different surface tension, probably better wetting) It seems to melt a little lower than claimed but it definitely looks and works like bismuth-heavy solder paste, so it's plausible.
- The melting point of traditional lead-free solder is usually advertised between 217 and 228 (which it is - the eutectic is 217, as is the solidus of the near eutectic alloys, while the liquidus ranges up to 227, though listings of "260" (which is clearly inaccurate for that composition) are not unheardof (!!). Regardless of the number they claim, it's going to melt below 230 and above 217, a good reflow oven can melt them, and the cheap ones struggle. Only the precise details depend on the exact alloy at hand.
Notice how all solder pastes advertise a specific melting point expressed as one number; that cannot possibly represent the melting point of the solder unless the alloy is eutectic. If the solder is close to a eutectic point, and hence the solidus and liquidus are very close, that's a fine approximation. This is the case with tin-lead solder, SAC305, and 138C bismuth solder. That is not always the case, though - and many brands only advertise the one number that they think the customer wants to hear - though they don't always agree on what that is. Some of the numbers claimed for SnBi and SnBiAg are impossible to reconcile with the phase diagrams of that mixture
If you just look at the liquidus of 172C, this looks like it should make a great replacement for SnPb solder, and if you look only at the solidus, 151C, it looks like good low temperature solder, but the combination is not so great - there are those 21 degrees where there's a liquid and a solid portion of the melt - it means the solder is like slush at those temperatures. It isn't well melted enough to wet the things you're soldering (which typically requires temperatures above the liquidus to ensure), but it's also not a rigid solid, so there's a longer portion of the cooling phase of reflow during which vibrations can knock parts out of position.
This is available with essentially the saem composition but with a wide variety of claimed melting points. Part of this is the broad range between the liquidus and solidus, but products claiming melting points as much as 17C above the liquidus are on the market. How is this possible? Likely thanks to the widespread dishonesty in this business. KEK uses the solidus (151C - though other brands claim that ), SD520 uses the liquidus. A few of the melting point claims are straight-up false, examples of willful mismarking of the product. Certainly that is the case for Conduction (183C) and Maijing (189C) unless they're lying about the alloy instead. It's also unclear what it is, specifically, about that ratio that makes it "special" and more appropriate for for solder than some other ratio; the phase diagram implies that incrementally more tin should yield higher melting
And only one of them says what it is. The Sn64.x/Bi35/Ag system could be called 158C, since manufacturers seem to think it's okay to pick any number in the melting range. The four I have tested all behaved very differently, but none were flagrantly bad. Well, maybe the UGainGreat brand wasn't so great.... The Relife XP-S/G-LON Sunshine X paste performed particularly well.
Generally tin-lead with extra lead. These are maybe useful for ultra-high temperature applications, but it's unclear what parts could survive their use. The very-high-lead solders are exempt from ROHS, surprisingly, because there are apparently use cases for which no alternatives exist.
I have done my best to identifty the alloys used by brands of aliexpress solder paste, as that stuff is far more affordable than the alternatives, and because that's the only place to get many interesting solders - western suppliers just don't seem to have the same variety of solders (at least some of them are so terribly no western company would want to be associated with them). However, some remain unknown (since some sellers/manufacturers are cagey about the details of solder compositions, others are are flat out dishonest, and many sellers others are apparently either clueless, or have such poor english skills that they might as well be. This list includes only stuff that can be had cheaply in consumer quantities. From Western suppliers, the only lead free solder you're gonna find is SAC305, 99/0.3/0.7, and the tin/bismuth eutectic, sometimes with 1% silver. All the leaded compositions around the eutectic point with and without silver are also available. All of these are only marginally better than the best stuff of the same claimed formulation from aliexpress, and much more expensive. However, the worst stuff from aliexpress is unspeakably horrifically bad and is entirely unusable. Makes a good "gift" to electronics enthusiasts who you don't like.
Note that there are many other exotic alloys beyond the scope of this chart due to cost, temperature, and applicability. Sn20Au80 (not a typo, that's 80% gold - it's a eutectic alloy used for connection gold bonding wires to the die of a chip, and is part of why you can run chips through a hammer mill, separate by density, and then extract gold from the heavier fractions of the resulting dust. While not very economical on modern parts, some old and long obsolete parts were apparently quite rich in gold, such that it's worth doing so, provided one is in the business and has the required equipment. A few of the more plausible solder alloys for general use (not die-bonding ones) are listed in the final table below.
These are available from many manufacturers. Some quality variation can be seen between manufacturers, ratings given are for typical examples. Low grade brands
| Common name | Melttng Point | Eutectic? | Tin (Sn) | Lead (Pb) | Bismuth (Bi) | Silver (Ag) | Copper (Cu) | Lead Swab Test | Rating | Wire? |
|---|---|---|---|---|---|---|---|---|---|---|
| Lead free | 227 | Yes | 99.3 | - | - | 0.0 | 0.7% | Untested | poor * | Yes |
| Lead free | 217-225C | No | 99.0 | - | - | 0.3 | 0.7% | Pass | mediocre * | Yes |
| SAC305 | 217-219C | Almost | 96.5 | - | - | 3.0 | 0.5% | Pass | sometimes good * | Yes |
| SAC387 | 217C | Yes | 95.5 | - | - | 3.8 | 0.7% | Untested/Pass | likely good * | Yes |
| SAC405 | 217-225C | Almost | 95.5 | - | - | 4.0 | 0.5% | Untested/Pass | likely good * | Yes |
| SAC405 | 217-225C | Almost | 95.5 | - | - | 4.0 | 0.5% | Untested/Pass | likely good * | Yes |
| SN100C ### | 217-227C | No | 95.5 | - | - | - | 0.7% | Untested/Pass | likely okay * | Yes |
| 60/40 | 183-190C | Almost | 60.0 | 40.0 | - | - | - | Fail | v. good ** | Yes |
| 63/37 | 183C | Yes | 63.0 | 37.0 | - | - | - | Fail | v. good ** | Yes |
| 62/36/2 | 179C | Yes | 62.0 | 36.0 | - | 2.0 | - | Fail | great ** | Yes |
| Leaded silver | 181C approx | No | 62.8 | 36.8 | - | 0.4 | - | Fail | great ** | Yes |
| 138C low temp | 138C | Yes | 42% | - | 58.0 | - | - | Pass | mediocre @ | Yes |
| 137C low temp | 137C | Almost | 42% | - | 57.0-57.6 | 0.4-1.0% | - | Pass | mediocre @ ## | Yes |
Many of these are available from Chipquik, Kester, or other Western suppliers.
These solder pastes are available on Aliexpress and, except where qualified with "likely", have all been tested. The alloy composition is rarely advertised, though in some cases the composition is listed or at least the constituents listed on the packaging (typically carefully obscured from view in photos in the listings).
| Common name | Melttng Point | Eutectic? | Tin (Sn) | Lead (Pb) | Bismuth (Bi) | Silver (Ag) | Copper (Cu) | Lead Swab Test | Rating | Alloy known? |
|---|---|---|---|---|---|---|---|---|---|---|
| Lead free AMOME | 217-227 | No | 98.5 | - | - | 1.0 | 0.5% | Untested/Pass | likely mediocre * | Yes, SAC105 |
| Chipquik SAC305 | 217C | Nearly | 96.5 | - | - | 3.0 | 0.5 | Pass | Good, expensive | SAC305 |
| JPT SAC305 t4 | "217C" my ass | No clue | >95.0 | - | - | ? | ? | Pass | Terrible! | Sure doesn't seem to be SAC305. |
| KEK SAC305, #4 | 217C | Nearly | 96.5 | - | - | 3.0 | 0.5 | Pass | Good | SAC305 for real |
| AMOME M13 | "190C" | No | apx. 63 | apx. 37 | - | 2.0 or less | - | Untested/Fail | likely v. good ** | Alloy unstated |
| 189C 2UUL | "189C" | No | 62.6? | 37.0? | - | 0.4? | - | Fail | Excellent ** | Composition marked on label. Often mis-sold as lead free. Excellent leaded solder |
| Mijing 189C | "189C" | Not close | 64.7? | 35.0? | 0.3? | - | Pass? | Pending # | Mysterious melting point claim, testing is required | |
| PPD S600 leadfree | "183C" | Unlikely | likely | unknown | unknown | unknown | Unlikely | Ambiguious | good, with caveat# | Claimed "SnBi", but some sellers have said otherwise |
| Laster 183C ROHS | "183C" | Unlikely | Unknown | Unknown | Unknown | Unknown | Unknown | Untested | Good first impsn. | Alloy unstated. seems very similar to PPD S600. |
| 64/35/1"Conduction | "183" (a lie) | Not close | 64 | - | 35.0 | 1.0 | - | Untested/Pass | awful *** | |
| 64.7/35/0.3 SD520 | 172C (liquidus | Not close | 64.7 | - | 35.0 | 0.3 | - | Untested/Pass | Not tested | I am pessimistic about the prospects for this alloy |
| Relife/GLON 158C | "158C" | No | ???????? | ????????? | ???????? | ???????? | Unlikely | Ambiguous | good ** @ | ROHS claimed, but unclear what it is made from |
| MaAnt "158C" | "158C" | No | 43? | 43? | 14? | 0 or ?? | 0 or ?? | Fail | good ** @ | Alloy unstated. If leaded, as they claim, there's not much else it could be besides 43:43:14 |
| Mechanic NS58 | "158C" | No | 43? | 43? | 14? | 0 or ?? | 0 or ?? | Ambiguous | passable ** @ | Alloy unstated. I suspect, 43:43:14. Definitely doesn't feel like SnBi, and they don't claim ROHS. |
| UGAIN 158C | "158C" | No | apx. 65? | 0 | apx. 35 | 0.0-1.0? | Unlikely | Untested | passable | Alloy unstated, but undoubtedly SnBi or SnBiAg, and below average at that. |
| PPD 158C | "158C" | Unlikely | Unknown | Unknown | Unknown | Unknown | Unknown | Untested | Untested+ @ | Alloy unstated. Looks like an above average SnBiAg alloy. |
| Laster 158C ROHS | "158C" | No | apx. 65? | 0 | apx. 35 | 0.0-1.0? | Unlikely | Untested | Passable+ @ | Alloy unstated. Very cheap. Looks like an above average SnBiAg alloy. |
| Wylie WL203 | "158C" | Unlikely | Unknown | Unknown | Unknown | Unknown | Unknown | Untested | Untested @ | Alloy unstated. On order. |
| 64.7/35/0.3 "KEK" | 151C (solidus) | Not close | 64.7 | - | 35.0 | 0.3 | - | Ambiguious | passable | Dull joints other typical defects of SnBiAg. |
148C 2uuL | "148C" | No | 45% | No | 55% | No? | No? | | Not good | But neither is it terrible, better than mechanic. Suspect mechanic 148 is sme alloy, or nearly so. 148C (Mechanic) | "148C" | Unlikely | Yes | Unlikely | Likely | Unknown | Unlikely | Pass | v. poor # @ @@ | "Lead free" |
Notes:
* Traditional Lead-free solder is often challenging to reflow in cheap reflow ovens.
** Contains, or is belived to contain, toxic lead. This is an envuironmental/disoposal hazard only, not an assembly.handling problem.
*** Abysmal mechanical properties. Conduction's take on the apr. 2:1 Sn:Bi+Ag alloy is particularly bad, and the claimed melting point is flagrantly inaccurate. Actually, Conduction is an awful brand in general. I got 3 kinds of their solder paste, and even the tin/lead solder paste sucked!
# Mysterious claimed melting point raises questions about composition.
## The added silver makes SnBi near-eutectic less brittle than eutectic SnBi.
### SN100C is SnCu 99.3/0.7 with 0.05% Ni and 0.006% Ge doping, which is said to improve many of the qualtiies that are lacking in 99.3/0.7 lead free solder, including poor wetting, whisker growth, dull joints and so on. It is said to be very good solder, possibly better than SAC305. Sometimes called Germanium doped solder. Scarse and exorbitantly expensive, until recently it was patent encumbered, and the name is still trademarked, so every second source manufactuerer has to call it something different. This has likely significantly hindered uptake of what should be a much cheaper lead free solder Unfortunately they don't address the most pressing problem for most hobbyists of the melting point of lead free solder being too bloody high, and I haven't been successful at finding it in paste form - not that I could reflow it anyway.
@may become liquid or partially liquid dangerously close to operating conditions! 158C solder performance will vary dramatically between brands, as there are a considerable number of alloys that it could be. Some is Sn43/Pb43/Bi14 (144-163C melting range) - this is in danger of forming a ternary bismuth-rich liquid phase whose eutectic melting point is a mindbogglingly low 95C. Presumably - barring contamination with additional bismuth, this doesn't form under real world conditions (and the most likely sources of contamination will shift the composition away from that, for example contamination from either leaded or lead free HASL)
@@ In my tests, this either embrittled or straight up dissolved copper wire (couldn't tell which)
- For home rework and repair, probably tin/lead 60/40 or 63/37. SAC305 is also suitable, though harder to work with (higher temperature needed). You may not have a choice though, due to environmetal laws
- For home rework and repair of fragile temperature senstive things, eutectic Sn/Bi
- For reflow, if your oven and boards and components can take SAC305:
- If you are selling the boards, SAC305.
- If you're doing it on large scale, the board is simple, and you're really cheap, Sn 99-Sn 0.3-Cu 0.7.
- If you're not forced to adhere to ROHS - Tin/lead eutectic, or SnPbAg alloy.
- If your reflow oven can't really do SAC305 but you want lead-free, I've yet to find a suitable replacement that is verifiably lead free. Candidates are:
- Maijing 189C - So far, so good, and definitely doesn't feel like tin-lead
- PPD S600 183C - If it's lead free, this is acceptable (but rework is particularly hard)
- L
- Relife/GLON 158C - Very good solder. Could it actually be lead free? It stands out for it's ease of rework
- PPD 340 158C - Untested, claims lead free.
- Wylie W203 158C - Untested, claims lead free
- Laster 158C - Untested, claims lead free
- Tin-bismuth is lead free, but otherwise worse and limits maximum operating temperature, and cannot be used with non-lead-free HASL.
- Tin-lead contains lead, but is excellent solder.
- If reflowing somehing very fragile, which has no leaded solder on it currently, SnBi eutectic or SnBiAg (which has improved properties - seems to only be available from western suppliers.
- Otherwise tin/lead - it works very well in almost all situations.
- 158C 43:43:14 solder paste is probably not very useful. Has possible application where the workpiece is fragie and lead is acceptable - as long as it's not being mixed with existing bismuth based solder, like the 138C eutectic. (provided that solder doesn't contain more than 14% bismuth). The properties are generally inferior to eutectic tin-lead solder, and it comes with all the caveats of bismuth bearing solder as well as the toxicity of lead, all for 25C lower melting point.
- Avoid JPT305T4 "SAC305" solder paste. It claims SAC 305. It looks lovely. I hotplated boards with it until the silkscreen was brown, and the solder still hadn't fully melted. This was the least successful assembly run I have ever done, since I began reflowing boards over 5 years ago. This solder paste is terrible. There is virtually no flux activity. Negative 5 stars, this stuff is unfit for any purpose whatsoever.
- Never mix any bismuth bearing solder with anything not lead free including HASL boards with leaded solder
- Always examine the label of dubious chinese solder (all solder from aliexpress and other market sites should be considerered dubious until tested) before buying (to the extent that you can). Failing that, before using it on anything that matters do some test runs. Even if there is no composition listed, you can use some heuristics without even opening the container; this applies to most anything else, too: Assume they put the same degree of care into the label and the product. Are the english words misspelled? Is it "Made in chian"? (ie, they didn't even spell their own country's name right)? You probably don't want that brand. Garbage labels like that are found on garbage product. Good labels sometimes are placed on low quality product, but manufacturers very rarely put a terrible label onto high quality products; It isn't that expensive to get a label that looks decent, and print it on glossy paper like everyone in the west does, certainly not compared to what it costs to make good solder paste instead of bad solderpaste.
- When in the position to choose between someting that looks like it was designed to impress individuals vs something designed for indstrial use, you've got better odds with the industrial looking one. This goes for everything, no matter where it's made. Every manufacturer in the world knows that "individual consumers" have virtually no leverage, are usually too unsophisticated to recognize how crappy the a low quality product is, and rarely have the time or patience to make a stink (especially on AliExpress) and can generally be taken for a ride without fear of repercussions. In buying solder, you'll notice two styles of marketing. You will immediately know what I mean when you see the first green, blue, or white jar a part number, alloy, and manufacturerer clearly marked and minimal brand-rubbish and compare it the label of Conduction (printed on dull paper, rife with misspellings, surrounding a plastic syringe full of unusually poor solder paste for any given alloy), or Mechanic (glossy labels and usually passable solder for a given alloy), KEK (glossy labels, but with less branding rubbish and more specs - and what do you know, it's generally better than Mechanic's rendition of the same alloy, and always better than Conduction's). An example of an encourging label on a crap product is the JPT SAC305 solder, as noted above.
- Avoid burning your boards by overheating them. A particular problem in hotplate rework. White silkscreen and solder mask turns a sickly salmon, and then brown, and everyone can see that you overheated your boards, and held the chips at peak reflow temps for too long. Hotplate temps should be tested to find the minimum needed to effectively melt the solder and permit the required rework.
| Product | Approx. Price | Cleaning difficulty | Color | Flux performance | Notes |
|---|---|---|---|---|---|
| Kester RF741 | ~ $30/15g, $65/100g | Moderate | Deep yellow | Excellent | Sold in the west, not from china. Note the higher price. While harder to clean, it is very very good flux. |
| Chipquik SMD4300TF | ~ $20/10cc | Average | White | Below averaqge | Water washable? Didn't seem to be! |
| NC-559 | ~ $5/2x10cc | Average | Cream | Unremarkable | Counterfeit Amtech paste. |
| UGAIN UG78 | ~ $8/10cc $24/100g | Below average | Clear | Unremarkable | Suitable for almost all tasks calling for paste flux, cleans up real good, and sold in decentsize jars. |
| Mechanic Icing SD360 | ~ $9/10cc | Below average | Clear | Unremarkable | Similar to UG78. Maybe a little easier to clean, and maybe not as good as flux? Works fine. |
| ZJ-18 Mild Rosin | < $2/10g tin | Moderate | Deep yellow | Below average | Occasionally, you get what you pay for; this stuff is cheap and crappy. Plus the tin is almost impossible to open. |
My favorites are the UGain UG78, unless I need something more aggressive, in which case I go for Kester RF741.
Sometimes while solering and cleaning boards with an iron, you will damage the soldermask. Clean the effected area thoroughly, be sure to sever any bridges to the exposed copper, and and apply (preferably color matched) UV curing solder mask.
I use dirt cheap disposable plastic paintbrushes to apply it. Try to stick to the board, but it isn't a disaster if some gets on top of solder if that solder isn't forming a bridge. with a black transparency, either printed or or made with tape on transparent film (that's how it was originally done). Use the kind of transparent film used for oldschool overhead projectors. Unless you want to go whole-hog and get stainless steel stencils made with (for example you're making a lot of boards and need two colors of solder mask - imagine an H-bridge controller where the high half was red and the low half was black).
Solder mask is commonly available in 6 colors - Red, Green, Blue, Yellow, Black, and White. Other colors are exotic (please let me know if you find cheap sources for any of them, ESPECIALLY clear).
A tube of mechanic soldermask costs a few dollars. Larger bottles can be had for under 10. Wrap the bottles in foil to be sure that no light will get to them. Mechanic brand is fine for this purpose, unlike their solder. Once the repairs are made, place the board in full sun for 15-20 minutes,
While (with very few exceptions) the traces on a PCB are made of copper, if the copper was used bare, it would corrode in air, giving a board a very short shelf life. This, in fact, is phenomenon known to anyone who has tried etching or milling their own PCBs Unprotected, the copper soon turns dull and becomes difficult to solder to. Eventually this corrosion will result in board failure (especially since such boards rarely have solder mask applied, and so the bare copper is exposed to air; you may have noticed a shiny uneaven coating on some amature-milled/etched boards; usually, the unnevenness is due to the brush strokes since they paint the board with clear nail polish as an ersatz solder mask and marginally effective conformal coating; the boards are still relatively unreliable). Of course, making PCBs at home is not advisable nowadays, what with low cost board houses producing excellent boards at prices lower than what an individual would spend on consumables to make their own, and the home process is laborious, requires handling corrosive and toxic solutions or an expensive CNC router to mill out the copper. Both methods require PCBs to be designed specifically to account for the artifacts of the patterning due to the crude methods generally employed and their single-layer nature * (which in turn leads to using lots of 0 ohm resistors as jumpers over traces). Circling back to my original subject, many home made boards are then treated with "immersion tin" which produces a layer of tin plating on the surface, which improves solderability and the lifetime of the solder-mask-less board, though when conducted at home, the quality of this coating is often very poor, limiting it's effectivness, and even applied commercially, it is no longer used without further treatment, if at all (I do not know if it is used as an intermediate step in some surface treatments or not).
The main surface treatments used today are described below.
Solder (either eutectic tin/lead, or lead-free) is applied to the boards as a liquid and spread using forced hot air to a uniformm, thin thickness. This is the cheapest one commonly used on boards sold into the hobby market. Leaded HASL is still quite common where it is not prohibited by environmental regulations. One telltale trait of HASL boards is that the solder on one side of the board is thicker than the other, particularly noticable on large pads: that side was facing down during the HASL process. That fact is occasionally annoying, but those occasions are rare.
One of the biggest practical issues with HASL boards is that they have solder on them; if you're using the same kind of solder with them, then you're happy. If not, you get a mixture of the two types of solder - in the case of regular leaded solder on unleaded HASL, the concentration will be shifted in the direction of tin, keeping the solidus unchainged but raising the liquidus a few degrees. Not ideal, but not disasterous - provided that the reflow profile gets sufficiently hot that the HASL solder has melted fully. Otherwise, you'll have a poor quality joint, which may only intemittently make contact (head-in-pillow defect or cold solder joint). In the other direction, using lead-free solder on a leaded HASL board will still require the full reflow temperature of lead-free solder to melt properly, and will still be full of lead - combining the worst property of leaded and lead free solder. Much more problematic is the incompatibility between lead and SnBi solders, due to the ternary eutectic that melts at 96C. Since leaded HASL will introduce lead to the melt, you can't use it with bismuth-heavy solder (the leaded 158C stuff with only 14% bismuth, shouldn't have this problem - the leaded solder already on the board shifts the compostion further away from the bismuth-heavy side, where that low melting phase is found), rather than towards it like the 151-172C Sn64.x/Bi35/Ag or 138C Sn:Bi eutectic. Those solders are not recommended for leaded HASL boards; the same problem occurs with indium-heavy solders.
As you can see, there's no upside to using a HASL board that doesn't match the type of solder you're using.
Shelf life of properly stored boards is measured in years, though on very long timescales, solder will slowly oxidize and become more difficult to work with. Prototyping board manufactured in the '80s for TTL, with leaded HASL, noticably harder to work with than new, and more concerningly, the adhesive between the traces and the circuit board has degraded, causing pads to lift more easily - but that's like 40 years old, so issues like that are only relevant to repairing vintage equipment.
This is the cheapest, and is extremely widespread in large scale production - it just consists of some organic chemicals that form a protective layer over the copper but which will be burnt off during reflow by the heat and/or flux. This coating is rather fragile (not resistant to abrasion) and has a fairly short shelf life, so it's not as appropriate where the boards are not going to be laoded within days or weeks, or will be handled extensively. Rarely even offered as a choice by the small scale prototypeing board houses.
This is the usual gold surface treatment you see on PCBs. exposed pads are first treated with a chemical that deposits nickel onto the copper, and the gold is then desposited by dipping in an appropriate gold-bearing solution. The process conditions have to be controlled carefully, and the nickel has significant amounts of other elements in it. Improper conditions will cause the dreaded "black spot" phenomenon where after reflow soldering, some parts of the gold pads would have instead turned to a plack patch that solder doesn't wet - the gold was dissolved, revealing the corroded nickel layer below; the corrosion occurs due to an interaction with the immersion gold and a nickel layer with excess impurities. Thankfully, this is rare now that ROHS has been effect in some major markets for over a decade. ENIG has become the surface treatment of choice for lead-free applications, being relatively low cost (hardly any gold is involved) though it works fine with leaded solder too. One surprising thing is that the gold layer is so thin that it actually dissolves in the solder! It's there only to prevent the nickel from oxidizing and preserve solderability. The nickle, meanwhile, serves as a barrier between the solder and the copper, which is felt to be desirable.
This is much like ENIG, only with an added layer of palladium deposited on the nickel before the immersion gold. It prevents black spot and increases corrosion resistance. Additionally, it allows a thinner layer of gold and the extremely thin palladium layer would theoretically allow lower overall production cost - though the lower volume means that it usually ends up costing more in practice. ENEPIG has been found to form brittle intermetaliics at the interface if leaded solder is used, hence ENEPIG should be used only for lead-free solder. ENEPIG has been shown in published papers to do better than ENIG with SAC lead free. I am aware of no research on it's compatibility with other alloys.
Rarely seen - gold plating applied directly to the copper. Has been shown to be compatible with both leaded and unleaded solder. The thickness of the gold layer is comparable to that of ENIG, so it will fully dissolve. Hence it is unsuitable for solders that dissolve or embrittle copper.
* You can't make doublesided boards as easily. First, the images on the two sides of the board need to be aligned very carefully (this is hard). Secondly, plating the inside of the holes and vias has always been tricky - but the clasical process is not practical for hobbyists in the US (and possibly elsewhere) because of the unavailability of a key chemical required by one of the steps; it was made very hard for individuals to obtain because it could be used in production of a major illegal drug. The restrictions had little impact on the drug chemists, because it turned out to be possible to produce that chemical in-situ from over-the-counter ingredients when making drugs with it; the drug in question, when made in domestic clandestine labs, is now made using an even cheaper and easer process, which in turn has faded from prominance and been replaced with industrial scale production in countries with poor rule of law, allowing cartels to leverage their existing expertise in smuggling across international borders, which they are better at than organic chemistry. So as it stands today, that key chemical, even if it was unrestricted, would be unlikely to see criminal use - but the restriction remains. The phrase "This is why we can't have nice things" comes to mind. That left people who needed two sides no option other than running a very thin wirehrough the hole and soldering it in place, which is as timeconsuming as it sounds. And no matter what, you'd still need to be able to accurately drill tiny holes with fragile drills that break frequently; Fiberglass is not an easy material to drill.
- Tin (Sn) is the base of most solder. Tin-free solder is extremely uncommon, and most solder is between 40% and 99.3% tin. Tin is very good at wetting other metals and getting small amounts of the other metal to dissolve in the solder. Alone, tin is horrible solder because the solublility of copper causes it to dissolve terminations. But it is the basis for almost all solder alloys.
- Lead (Pb) lowers the melting point of pure tin greatly, and also reduces the solubility of copper greatly in SnPb solder; far from the eutectic point on the lead-rich side, the melting point rises dramatically 90+% lead solder is used in high temperature applications and exempt from ROHS as no suitable alternatives are available (though of course, normal ratios of tin to lead are not exempt, just the strange high temp stuff)
- Copper (Cu) is present in most high-tin lead-free solder at 0.5% to 0.7%. Why this specific percentage? That's the solubility of copper in tin unless temperatures are very high. Slightly less is used in silver-bearing alloys, because the silver also lowers the melting point of the copper. Copper's solubility in tin-lead and tin-bismuth alloys is lower, and so it is not added there.
- Silver (Ag) added to solders at between 0.2% and 3% (typically) results in an improvement to a variety of physical properties, better wetting, and a slight reduction in melting point, particularly pronounced in solder that is otherwise 95+% tin. Silver is extremely expensive relative to most other solder-metals, and as such the concentrations used are frequently lower than the ideal ones. It appears to also slightly lower the solubility of copper in the tin, as evidenced by SAC305, (0.5% Cu, vs the cheap stuff with 0.0-0.3% silver which always has 0.7% copper in it. )
- Bismuth (Bi) is a non-toxic alternative to lead in order to lower the melting point of tin - but it is unfortunately rather too good at that - 138C is about the lower limit of melting point that can be tolerated in solder for any large scale use, while automotive applications require higher temperature alloys. Bismuth-heavy solder also generally has overall inferior properties.
- Zinc (Zn) is rarely seen in electronics solders, though some alloys with a near perfect melting point containing small amounts of zinc exist and are listed below. Unfortunately, it oxidizes and contributes to formation of dross, and it's not clear if any amount of zinc can be tolerated in solder used for electronics.
Unless one of the branded pastes is - in fact - one of these, these aren't readily available, or are available only with MOQ in excess of what an individual would want and/or (usually and) are too expensive for most hobbyists. Though their melting point = especially when inaccurately presented as a single number (as is commonly done), suggest that they might make good solder, some of them are lead-free, and they have been included in lists of solder alloys or investigations of their use as solder are documented, in most cases, there is a good reason they are not used. These, where known to me, are listed.
They are ordered by liquidus, descending - that's because this document is targeted at hobbyists. For hobbyists who for one reason or another would prefer not to use or cannot get* leaded solder, the main reason for not using "traditional" lead free solder alloys like SAC305 for reflow soldering is that they do not have a good reflow oven (which can be quite expensive) and are instead using either a modified toaster oven, or a cheap reflow oven (like the Puhui T962, which is quite popular) have difficulty reaching high enough temperatures; often they will not completely reflow the joint, or when pushed that hard will generate an uneven temperature distribution that successfully solders only boards in a small fraction of the oven area, and/or has high defect rates. Hotplates can be used, and sidestep that problem, and while highly effective at reflow soldering with lead-free solder, they can easily overheat the board. This results in white silkscreen and solder mask taking on a sickly salmon color (it is rare for components to be damaged by this, especially since - unlike most reflow ovens - hotplates heat from the bottom). Hence, for hobbyists, a lead-free solder that melts at a temperature lower than 217C is desirable Obviously, the solidus is also very important, as you need to keep operating and storage temperatures a minimum of 20C below that (preferably more), as is the difference between the two, which determines how long the board spends cooling with the parts not rigidly attached, (20C is the recommended maximum, according to literature). The joints formed by non-eutectic alloys are frequently inferior - consider that as the solder cools and solidifies, the composition of the liquid phase will change, leading to solder with a variety of solid grains. This effect is why eutectic alloys tend to produce "brighter" solder joints: the surface, rather than being a uniform composition, is a mosaic of small grains of different compositions.
* I will mention that if you can't get it due to legal restrictions on lead-containing alloys, but wish you could use it, you're unliekly to have any difficulty finding an overseas vendor (probably in China) who would be happy to ship you leaded solder or solder paste. Such activity is rarely the target of enforcement action when not done on an industrial scale - there aren't environmental police coming to individuals houses to test their solder for lead - and Customs has much higher priorities (drugs, weapons, and components thereof) than nabbing a roll of leaded solder. Not that I condone such willful disregard for the law, not to mention the environment, of course).
Notice that several alloys consisting of metals which are sometimes used in the eutectic composition are used in other ratios: While the liquidus (and thus the maximum temperatures required for reflow soldering) increases, the solidus remains at the melting point of the eutectic. This makes for lousy solder. Once the alloy is far enough from that eutectic point, this ceases to be the case (this can be seen most clearly with the SnBi and SnPb systems; recall that the solidus of SnBi 65:35 is 151C). This phenomenon is almost universal with binary alloys, and can be seen in any phase diagram. In alloys with more than 2 components more complicated behavior is seen, and this effect, while it still occurs, manifests instead as a small change in the solidus and a larger change in the liquidus.
| Name (if any) | Melitng Point | Eutectic? | Tin (Sn) | Lead (Pb) | Bismuth (Bi) | Silver (Ag) | Copper (Cu) | Indium (In) | Other Elements | Issues | Looks good? |
|---|
-
| 287-347C | No | 5 | 92.5 | - | 2.5 | - | - | - | Leaded, 5, 8 | No, MP far too high | -
| 296-301C | No | 5 | 93.5 | - | 1.5 | - | - | - | Leaded, 5, 8 | No, MP far too high | -
| 268-291C | No | 10 | 88 | - | 2.0 | - | - | - | Leaded, 5, 8 | No, MP far too high |
Bismuth metal | 271C | N/A ** | - | - | 100 | - | - | - | - | 7, 8 | No, MP too high | Babbit | 237-253C | No | 88.9 | - | - | - | 3.6 | - | Sb 7.5 Ni 0.07 | 6, 8. 3, 7 | No, MP too high | Babbit + 3% Bi | 321-241C | No | 85.9 | - | 3.0 | - | 3.6 | - | Sb 7.5 Ni 0.07 | 6, 8. 3, 7 | No, MP too high |
-
| 220-240C | No | 95 | - | - | - | - | - | Sb 5 | 6 | No, MP too high |
Babbit + 2% Ag | 225-238C | No | 85.9 | - | - | 2.0 | 3.6 | - | Sb 7.5 Ni 0.07 | 6, 3, 7 | No, MP too high | Sn:Pb 40:60 | 183-238C | No | 40 | 60 | - | - | - | - | - | Leaded, 3 | No, much-worse tin-lead |
-
| 220-234C | No | 97 | - | - | 0.2 | 2.0 | - | Sb 0.8 | 6 | No, MP too high |
CASTIN | 219-230C | No | 96.3 | - | - | 2.5 | 0.7 | - | Sb 0.5 | 5, 6, 10 | Not really | SAC-I | 212-229C | No | 92.5 | - | 3 | 3.8 | 0.7 | - | Sb 1.4 Ni 0.15 | 9, 6, 5 | Not really | Sn:Pb 50:50 | 183-216C | No | 50 | 50 | - | - | - | - | - | Leaded, 3 | No, just worse tin-lead |
-
| 200-205C | No | 84.4 | - | - | 5 | - | 8.6 | Sb 2 | 1, 5 | Probably, except for the cost. | -
| 190-195C | No | 89 | - | 3 | - | - | - | Zn 8 | 11 | MP looks good, but zinc said to corrode. | -
| 175-187C | No | 77.2 | - | - | 2.8 | - | 20.0 | - | 1, 2, 5 | No, see issue 2 | -
| 175-187C | No | 86.9 | - | - | 3.1 | - | 10.0 | - | 1, 5 | Probably, except for the cost. | -
| 174-186C | No | 86.5 | - | 3.5 | - | - | 4.5 | Zn 5.5 | 1, 11 | MP looks good, but again the zinc corrosion issue |
SnBi 60:40 | 138-170C | No | 60.0 | - | 40.0 | - | - | - | - | 3 | No, slightly higher tin alloys may be better | Indium metal | 157C | N/A ** | - | - | - | - | - | 100 | - | 1, see also 2| No, too expensive |
CASTIN is a registered trademark of AIM.
* Assuming you call the nickel and germanium "dopants" not constituents of the alloy.
** As a pure metal, it has a single melting point like a eutectic, but as it is not a mixture, that term is not appropriate.
Name is omitted even where soemtimes given (every metals company has their own trademarked names for alloys, which are often identical to what other companies call other things.)
- Indium prices are both volatile and high. The main source of demand is in production of Indium-tin oxide (ITO). While alternatives are an active area of research, ITO is currently the only viable option for a material that is both transparent and electrically conductive. Constructing an LCD or OLED screen requires a grid of elctrodes, one being placed on either side of the active element. A screen is not very useful if you can't see through at least one side of it, so every LCD and OLED screen requires some indium. The amount of indium is very small - around half a gram for a 15" screen (ITO is used in a very thin film, with as low a ratio of indium to tin as they can manage without degrading the optical properties). This is small enough that it is not economical to recycle, as there is no easy way to separate it from the rest of the screen, even ignoring the need to separate the screens from the rest of the device. Indium is also used in blue (and thus white) LEDs (as InGaN) and other semiconductors (GaInAs. GaInAsP, and as a dopant), some types of solar cells, and in a recently commercialized inorganic pigment (as Yttriuum-Indium-Manganese oxide, "YInMn blue", vibrant blue in the visible spectrum, and highly reflective in near-IR, which unlike other blue pigments is not particularly toxic) - all uses of it which are increasing. This wouldn't be a problem, as the element isn't particularly rare (similar to bismuth, mercury and silver) but it doesn't form specific mineral that can be concentrated through geologic processes, so you can't go find a rich vein of "indium ore" and set up an indium mine. It is instead extracted as a byproduct of mining other metals (mainly Zinc) and forms a tiny fraction of mine output. For much of the 20th century it was worth about as much as silver, and production of Indium was far lower than production of silver, mercury, or bismuth - it was not worth the effort of extracting it from mine waste. As LCD screens became practical, that changed: prices jumped from around $200/kg in 1995 to 500/kg in 1998, before falling back to below $200 in 2002, then jumping even more dramatically to over $1000 in 2006. This valatility has continued - it had fallen below $300 in the 2009 recession, before jumoping back up to almost $900 in the mid 2010s, though it has fallen back to around $250/kg as of 2022. To put that in perspective, it's peak was close to triple the price of silver at the time - though silver has steadily increased in price, and today indium is less than half the price of silver. The incredible volatility resulted from a combination of relatively inelastic supply and the demand increasing rather erratically as technology has progressed - and whereas silver is held in large amounts as an investment, serving to buffer the price, indium is not. Silver is also mined in significantly larger quantities, with electronics being a relatively minor application. And whereas silver works it's magic in alloys at a few percent, in alloys that make good solders, indium is needed at much higher proportions.
- Has problems above 100C, becase of an SnIn eutectic phase can form, with a melting point around 120C. This does not form at 10% indium. The same issue would be be problematic for pure indium even if it wasn't prohibitively expensive as presense of tin would cause problems. Indium and Indium-heavy alloys also encounter the same problem with lead or bismuth for the same reason. Using these as general purpose solders, techicians who tried to rework the board without knowing what kind of soldeer was originally used could form alloys that would melt at operating temperature... not to mention the fact that tin plating is used on terminations of viruallty all electronic components, which would make a transition to pure indium solder very difficult for the industry. (See Appendix C). If Indium bearing alloys have direct contact with copper, it will form a very brittle intermetallic, making those less suitable.
- Huge gap betwee liquidus and solidus, > 30C - the solidus is only 138C. So it combines the downside of a lower melting eutectic with the need to heat the solder more to fully melt it and
- Copper is soluble in this solder up to 0.7%. This can cause problems with the copper traces literally dissolving into the solder.
- Silver is expensive.
- Antimony is toxic, but is nowhere near as large of a concern as lead, though it doesn't bioaccumulate to anywhere near the degree that heavy metals do, nor does it oxidize particularly readily (which would generate more toxic forms - in addition to being directly undesirable in solder, since it would mean the solder was corroding).
- Poor wetting properties.
- Melting point is too high.
- Not ductile enough to make solder wire with, which also implies that the resulting joint would be vulnerable to failure from brittle fracture.
- Patent-encumbered, as SN100C was until recently. In fact, many of these alloys probably are patented. They certainly were all patented at the time that they were first developed. CASTIN and SN100C were patented by the same company, actually - AIM. While patents seek to increase innovation by ensuring that there is an incentive for inventors to invent, they also form a barrier to widespread adoption if the patent holder is unable to move a newly patented invention from the lab to widespread use, for example because they are not able convince investors to provide sufficient funding, or are uninterested in doing so and sell the patent to a company which has no intention of commercializing the technology, instead profiting from licensing fee and judgements in lawsuits against others who try to bring that invention to the world. Such "non-practing entities", particularly those with overly broad patents, are of course rather controversial, as they have the effect of hampering innovation, since the patent is being used to hinder innovation for profit instead of profit from making use of the innovation. Even when the patent is being licensed or produced directly by the patent holder, the pricing is often high enough to inhibit market adoption (like SN100C solder). Often technologies take off only after certain key patents expire, because the patent holder was enforcing the patent against all, but only developing products to cater to the high end of the market (this happened with several 3d-printing technologies). Trademarking the name of something is less harmful to innnovation, but may hinder uptake long after the patent expires. This is the case with SN100C (A registered trademark of AIM), if you decide you want some, well, how do you find other suppliers who might charge a less stratospheric price, if they're not allowed to call it that?
- Zinc has problems with dross/corrosion. Pity that, is, as the melting point of Sn89 Bi3 Zn8 looks just about perfect!
This section serves to clear up to most obvious messages claimed by the above. The mixing of different solders forming lower melting alloys can be a big problem, since unless far from the eutectic point, the mixed solders will start to melt at around that point. The SnBiIn eutectic, known as Field's metal, after it's inventor, melts at 62C and none of its components are toxic. However the effect of a 62C melting eutectic forming because you mixed indium and bismuth solders is definitely toxic to the circuit board's reliability, and the price of indium makes any indium-heavy solder highly toxic to the wallet. That is particularly true if one hopes to use this as a low melting alloy for some other application - the composition is 51% In, 32.6% Bi and just 13.3% tin, bringing it to somewhere around $140/kg on a metals basis. It is the lowest metling alloy not based on gallium that contains no toxic elements and doesn't react violently with water (the elements in the first column of the periodic table are all either liquid at room temperature (Cs, Rb), or can alloy with eachother to form a an alloy that is (NaK or LiNaK), but they also react rapidly with air forming first an oxide, and which hydrolyzes to the hydroxide (lye), which is deliquescent in air (will pull moisture out of the air, speeding the reaction of any remaining metal, after which it will continue absorbing moisture until it dissolves into a puddle of corrosive liquid - which then absorbs CO2 from the air, is no longer as hydroscopic, and you're left with carbonates), and reacts explosively with water. It is generally stored under oil, these alloys are only interesting for very specialized uses, or for throwing into bodies of water to watch the fireworks). Adding the last of the major solder elements, lead, to the SnBiIn eutectic can result in a SnBiPbIn eutectic that melts at 58C (136F), well below the boiling point of water (but still too hot to touch at the melting point). The eutectic alloy only has 21% indium in it, making it less eye-wateringly expensive if you're in the market for fusible alloys, but likewise totally unacceptable as solder.
In any event, the message here is that two solder alloys, which might be great solders on their own, if mixed, can form something that is thoroughly unsuitable as solder by virtue of it's low melting point, and which will fail at temperatures that can be realistically achieved during opperation even in a mundane environment.
The following isn't relevant to soldering (thankfully - as you'll soon see, these are not alloys you want to be exposed to) there exist two eutectics containing those four metals plus an additional metal or two, which melt at an even lower temperature. Adding cadmium (and changing the ratios of the others; adding another major constituent always significantly shifts the eutectic composition) forms a eutectic alloy that melts at a temperature low enough that you can cast your fingerprint in it, though it is uncomfortably hot. Unfortunately cadmium is considerably more toxic than lead and a known a carcinogen. (it's toxic enough that I didn't bother including alloys containing it in the table above; there are a few that might be viable solders were they not so poisonous. Cadmium has a non-negligible vapor pressure in the liquid form (as mercury does, and which is the main reason that mercury is as hazardous to work with as it is*), so using Cd-bearing solder would generate toxic vapor - normal soldering fumes aren't all that bad..., even from leaded solder). Is there anything we could add to the SnBiInPbCd system to push the melting point down even further, maybe so we could touch it without discomfort at the melting point? Yes, in fact there is a 6-component eutectic that builds on that with the addition of a sixth metal, and it will melt on a hot summer day in in many parts of the world (41.5C or 107F mp at eutectic), so you could cast your fingerprint in it without it being uncomfortably hot. Alas, the sixth metal, one row down on the periodic table from relatively non-hazardous indium, is thallium which is far from benign. While I have made the 4-and-5-component eutectics (starting from the metals, it's quite easy to melt them down to get an alloy, since indium melts at such a low temperature, and as soon as it melts, the other metals start dissolving in it - as they do so, the melting point decreases further, so you don't need to get it anywhere near the melting point of bismuth or lead to melt it all), I never had the guts to open the vial of thallium in our elements collection to make the 6-componnt alloy. Thallium is is a heavy metal, and one of the nastiest of them (excluding synthetic radioactive isotopes and metals).
Thallium makes lead look downright healhful. before being banned for this purpose in many countries and abandoned for that purpose where not explicitly banned, it was used in rat poison (and basically nothing else) making it readily available, and it became a popular murder weapon. Unlike cadmium and lead where the concern is predominantly of chronic toxicity (the body has great difficulty excreting lead. cadmium and to a lesser extent, thallium) and handling solid lead or even cadmium isn't that scary - and metallic lead oxidizes only very slowly, and isn't very bioavailable **). Most lead and cadmium poisoning results from repeated small exposures slowly building up to toxic levels, resulting in a variety of deleterious but not immediately fatal symptoms. In children, lead is much more harmful than it is in adults, as it interferes with brain development, and it is this that is the main impetus to remove lead from things). In contrast thallium has extremely high acute toxicity- the LD50 of thallium in humans turned out to be just 10-15 mg/kg, in the same neighborhood of the estimated 2-20mg/kg for the much more famous arsenic (meanwhile for the rats it was intended to kill the LD50 was around 30-40mg/kg). The soluble salts - it's most toxic form, and the one used in rat poison - are tasteless, and once in the body, it is mistaken by cells for potassium and absorbed readily, However, having very different chemistry, it disrupts nearly system in the body through interactions with sulfur containing amino-acids. Thankfully thallium is almost never encountered (despite being around 14 times more common than Indium, bismuth, silver or mercury). Demand for the metal is tiny, and a mere 10 tons or so are extracted annually (as a byproduct of extracting more desirable metals) - with it's use as a pesticide discontinued, it finds use only as a dopant in specialty semiconductors, high diffraction glass (being non-water-soluble, thallium glass is not particularly hazardous) and in scientific research, in basic chemistry and physics. There are not many pure elements more toxic than thallium - arsenic competes for that position, but I can't think of anything else quite as nasty.
* Many dental fillings are made from an amalgam of mercury and nobel metals - these do not leech mercury, surprisingly. One one occasion, police entered a home for a wellness check on a funeral home employee who had not been seen for some time. They discovered him dead in the basement next to his illicit operation (sometimes, occasionally, bad people get what they deserve). He had expired from mercury poisoning, you see. To make some extra money, he had been removing the fillings from bodies before burying them. But no precious metals dealer is going to buy gold fillings from someone - they'd have to be a moron to not realize that the fillings were obtained illicitly. So he had to melt it down and get rid of that damned mercury. So he simply heated them, volatizing the mercury and leaving behind the precious metals that constitute the rest of the alloy, which he could then sell off for handsome profits. Unfortunately for this dispicable individual, he did it apparently without proper ventilation. The entire house was found to be heavily contaminated with mercury. Interestingly, however bioavailable mercury vapor may be, it's nothing compared to the bioavailability of organo-mercury compounds (same goes for organolead compounds). One researcher, spilling a drop of dimethylmercury on her glove, immediately washing hands, removing glove and so on, she expired of mercury poisoning within 6 months, despite aggressive chelation therapy.
Gallium metal is cool stuff. It's a relatively benign metal biologically, with a bizzarely low melting point of 80 F (thoroughly useless as solder, obviously - at least to humans; it might be a component of solders used by some alien race on a much colder planet with oceans of ammonia or methane. If you hold it in your hand, it will melt. Why This is all really cool. And it's more or less non-toxic. But it is otherwise quite evil. Aluminum that has been abraded with gallium is frequently destroyed (how severely depends on the specific aluminum alloy, but the worst piece I tried it on, 24 hours after exposure of a small area to gallium, the whole thing had the structural integrety a potato chip. That's why it's not supposed to be shipped by air (of course guess how the vendors on Aliexpress ship it?). So anyway, I found a solid puddle of what I belived to be low melting solder set aside, and decided to melt it by placing it on an index card and poking at it with the soldering iron, in hopes of insight into what it might be. Whatever it was it sure did melt fast. And didn't solidify (gallium has a remarkable tendency to cool below it's freezing point while remaining liquid - those aforementioned ammonia-drinking aliens would need to alloy it with something). The next day I noticed that it was still liquid (it was a hot summer day)... "Huh", I thought "One of my gallium vials must have spilled at some point" and marked a vial "dirty gallium" (since it was now contaminated with tin and probably lead from now) and poured it in. Later that night, I tried to use my soldering iron. The tip would not wet. It looked bright and shiny, but solder would just bead up and roll off it. Tip tinner did not help, gently scraping the tip did not help. I thought back to the ways in which I had abused that soldering iron tip since I last knew it worked, and this is the only thing out of the ordinary. That was LT-KN tip, $17.30 from McMaster Carr, goddamnit.
So uh, I guess don't use a soldering iron to melt gallium? Hot water works better anyway.... At least not your soldering iron, or mine - It strikes me that this would be a wonderful awful thing to do to an electronics geek as a practical joke or act of sabotage....
** To make lead a serious poisoning hazard (as the metallic form is not harmful, being almost totaly non-bioavaialble), you'd have to do something crazy - like if you could get people to cover their walls with lead oxide, leave it there until it started falling off, and then people chose to to play in the resulting dust or eat the paint chips... or if you were to build large numbers of machines that would burn an organo-lead compound such that they would spew very fine lead aerosols, and then got people to handle this organic lead compound and locate those machines in populated areas.
Which of course, is exactly what we did. Lead oxide made a good white pigment, and lead paint was widely used until it was displaced with titanium dioxde, which was also a better pigment. And lead salts, for some perverse reason, taste sweet, so small children. who are most vulnerable to lead poisoning and least liekly to know better, apparently sometimes ate lead paint, in addition to having more exposure to dust in the home in the course of play. And for decades, we put tetraethyl lead into gasoline used by our cars, whose exhaust would then contain lead particles small enough to be absorbed through the lungs) - at the time, petrochemical technology was relatively crude, and producing fuel for cars that didn't cause "knocking" (premature ignition from the heat of compressing the mixture of fuel and air) was expensive. The organic lead somehow inhibited this, allowing the cheaper fuel to be used, without knocking, which damaged the engine. Initially, car owners added the tetraethyl lead themselves, though that was banned after people accidentally poisoned themselves when a bottle spilled in the enclosed, poorly ventilated garageor they were otherwise exopsed to it. But instead of banning it entirely, they just cerntralized the addition of it to the fuel, eliminating the acute exposure to huge doses of lead that came with untrained persons without protective equipment handling it.
Those two sources of lead poisoning have contributed to a perception that lead is more toxic than it is, though its harm to children in particular should be taken seriously. I could name a dozen compounds off the top of my head that are far more lethal than lead, both organic and inorganic compounds.