The NASA handbooks are good. I like the one for rework that shows neatly staked wires. There are a couple of subtleties you could miss on this one though:
1) On page three, that exposed end of the lead is a defect if you're NOT going to conformally coat the board. If you don't know what a conformal coating is, it's a defect.
2) On page eleven, flux residue. Sure, for NASA's applications. For most commercial applications, the IPC rule is "excessive flux residue". For prototypes, one-offs, and hobby projects, my rule of thumb is, if you can't measure the difference removing flux makes, leave it there. Incompletely cleaned flux can be more corrosive than undisturbed flux. Also, no-clean flux should be called "never clean" flux. Never clean it with ordinary methods because you will never get it clean.
That's fine, I like them that way too! But as a hobbyist, you risk leaving corrosive salts on the board or worse, sloshing them into electromechanical parts like switches, pots, unsealed relays, sounders, etc. Getting to "pristine" is harder than a lot of people think. There's a vicious cycle in production where assemblers will use too much flux, get flagged in an IPC-610 evaluation, try to clean the boards, and then get flagged for high ionic contamination in the next evaluation. You can't see the salts you leave behind from incomplete cleaning.
Beginners should not ever clean flux. Most hobbyists should never need to clean flux. Professionals should have data driving any decision to clean flux.
Fortunately, I rarely risk anything if the board fails. And if it does I always try to understand what caused the failure.
I never considered flux compatibility between my solder rosin core and the flux paste and never considered measuring salt residues on the board.
I almost exclusively use the same Kester sn64pb37 rosin core solder and MG Chemicals #8342 RA rosin flux paste. Both must be cleaned promptly. Now, I don't bother cleaning if I don't intend to use the board after I test it (like when I do prototyping) but anything else gets a thorough wash. I use various PCB cleaners, typically ethanol (not IPA!) based and then rinse boards in electronic water. I never saw any oxidation on any of the traces unless there was some external contamination.
Was curious what their treatment was going to be regarding tin whiskers. Not much there (of course this is point in time inspection so it kind of makes sense), but a quick google revealed quite a bit of study:
I thought they were just one of those low probability items until a Sunday morning when I was running live audio for a church service. Singer was mid verse and BOOOOM! Huge booming sound over speakers startling everyone into nervous laughter. I opened the cable up afterwards and sure enough found these alien crystalline structures growing between the conductors. It had shorted and even the little bit of current in the microphone vaporized about a mm of material.
Really really weird and I'm certain kept some folks up at night over the years.
Not a pilot or aircraft owner, but my understanding is that part of the problem is because certifying a new aero engine is insanely onerous (as is certifying anything that goes on an airplane), a lot of basically obsolete designs requiring leaded fuel were produced long after auto engines switched to not requiring leaded gas.
Why would an engine require leaded gas, you ask? Because it helps keep the valve seats from degrading. We have other anti-knock additives, but nobody has figured out a replacement for the other properties that lead has.
And since aircraft are usually extremely long-lived, we'll be stuck with this problem until a substitute is found. Merely certifying new engines that don't require lead won't solve the problem because of the remainder of the fleet that does require leaded gas.
It's pretty limited. Avgas doesn't mean jet fuel, as used by commercial aviation, it means the stuff you pump into piston engine light aircraft. Plus, at the risk of being obvious, the exhaust is either emitted way up in the air or, if at ground level, out in the middle of nowhere at an airport.
Use leaded solder and either inspect the joint for complete coverage or solder dip the part beforehand. Then, conformal coat the circuit card with sufficient thickness for tin whisker mitigation.
Conformal coat really is the saving grace for tin whiskers. It's not perfect, a whisker can push through sometimes but then the whisker has to push back through the conformal coat again which is much harder.
> Was curious what their treatment was going to be regarding tin whiskers. Not much there (of course this is point in time inspection so it kind of makes sense)
Did you read their specs?
There's a small part about whiskers that already convey enough meaning:
> UNACCEPTABLE WHISKER
> A whisker is a slender needle-shaped metallic growth between a conductor and a land. Typically the result of mechanical stresses in high tin compounds, it is a reliability concern."
'Please don't comment on whether someone read an article. "Did you even read the article? It mentions that" can be shortened to "The article mentions that."'
It doesn't add anything to the discussion. And in this case it was very clear that the OP read the article and just missed that part.
Your comment would have been just as informative and less snarky if it was simply:
There's a small part about whiskers that already convey enough meaning:
> UNACCEPTABLE WHISKER
> A whisker is a slender needle-shaped metallic growth between a conductor and a land. Typically the result of mechanical stresses in high tin compounds, it is a reliability concern."
> 'Please don't comment on whether someone read an article. "Did you even read the article? It mentions that" can be shortened to "The article mentions that."'
Not an article, but specs. You read specs carefully, you don't skim, especially to write long paragraphs about your personal experience
> And in this case it was very clear that the OP read the article
I disagree. Not read, skimmed at best.
> and just missed that part.
I agree: the most important part was indeed skipped.
I did see that part, which is why I said 'not much' instead of 'nothing'.
The issue with whiskers is that they grow over time (which is why I mentioned 'point in time'). I didn't see anything in there about solder alloy selection or any other mitigation.
These high standards aren't unique to NASA. These few days I've been taking a micro-soldering course taught by Hakko. They are teaching standards IEC61191/JISC61191 (Japan Welding Society standard) and a lot of details are similar.
I used to work with a developer whose second job was with HP back in the 70s. He was hired as a technician building lab equipment and had to be prequalified from an educational and trades/skill perspective to be hired.
Then they taught him to solder again. Their standards were apparently incredibly high.
This reminds me of my drafting classes (back when paper was used). I was re-taught how to make letters. One would think that it would not be necessary but for their standards, it was.
I had to fix the exhaust on this one too! This mower came with the property I bought this summer. You'd be hard pressed to find something less maintained than this mower. It was missing both exhaust bolts where the exhaust connects to the manifold. I had some left over "make-a-gasket" material so I used that and some old machine screws that I had available. Seriously, half of the exhaust was just banging around, exhaust leak city with it coming straight out of the manifold.
You are not alone.
I have a cheapo soldering iron with a roll of ... metal melty wire in a plastic bag that I pull out and struggle for 30mins to make a basic connection almost certainly the wrong way.
But I flick the little switch and the little fan comes on and I am happy.
But I should do a course. Or at least google some more. I don't even know many of the terms people are using here...
I want to share some of my thought on soldering:
The best way to understand soldering is to think primarily in terms of surface tension or capillary action and intermetallic compounds.
You set it up correctly and apply heat so that fluid dynamics can finish the job.
The atoms get to where they are supposed to be. There is no technique or dexterity required with the soldering iron itself.
I would also recommend that you think of solder wire as a 2in1.
The flux core in the solder wire is helpful, but not the endgame.
Use extra flux from a flux pen or flux bottle that is compatible with the flux core in the solder wire.
For example: You can put a glob of solder on the tip of your soldering iron and just bring it in to the surface or joint.
If you have pre-applied flux on a clean surface or joint, the liquid solder will wick off the soldering iron and onto the surface by itself. Perfect every time. (Surface tension, capillary action.)
For manual hand soldering, the flux management is the only complicated part.
What flux to use? Do you have to clean it? How to clean it? Saponifiers, or IPA good enough? Compatible with the flux in the solder wire?
You must be sure to reach the activation temperature of all the flux on the board or it can corrode metals over time.
"No Clean" fluxes are really hard to clean without special saponifiers, I would avoid these fluxes.
Is the flux residue on the board still inert after it gets wet by accident or at high humidity?
There are also problems with high frequency circuits and No Cleans.
Water soluble fluxes require excellent process control. Even tiny defects in the PCB can cause failures if you use water soluble flux.
Read the datasheets and recommendations from the flux manufacturer.
Some fluxes are more easy going:
For you Americans I would recommend Kester 186 Flux-Pen and Kester 285 Sn63Pb37 RMA Flux-Cored Wire.
Do not use more flux than necessary, do not make a mess.
Use eye protection for this flux from Kester!
Soldering gets even more interesting with gold embrittlement (gold plating thickness and migration to the center of the IMC), tin whiskers, silver whiskers, etc.
There is much more to all of this if you want really high reliability.
I went deep down this rabbit hole once. Not a professional soldering guy. Just obsessive compulsive about high reliability design.
This is enough text for a post on HN :)
These are all good rules. Sometimes it can be really hard to tell when a solder is solid. You can get hairline cracks that aren’t visible to the naked eye but cause intermittent problems.
In critical applications, much more than you'd imagine. Lots of defence and critical component suppliers still maintain a significant amount of hand work for low volume and critical areas.
Don't have a source unfortunately other than personal experience working as an engineer.
For low volume, it can cost more to set up the P&P machine, deal with alignments, check first articles, etc. than just soldering a few (dozen) board by hand. And there are a lot of small companies in niche markets whose annual production of a PC board might be in the single digits.
That and, people who hand solder stuff all day are fast.
Some fab I checked out use manual pick and place if the quantity of boards is less than 100 since it probably takes more time to set up the automated stuff at such low volumes.
I’ve see the videos on IG every now and then on their hand soldering skills. So fast!
Over the years I have seen couple of closeups of various boards on youtube. As electronics is my hobby, I tend to pay attention to this. They all looked hand soldered to me. And the soldering job always looked very, very high quality.
I've been told that in areas where lots of prototyping isgoing on, there is lots of scope for manual rework. (a pcb design only 95% correct needs modifications to figure out the right design)
We had a board where somehow the footprint of an 8 pin SOIC got flipped in the layout. One engineer was able to desolder it, flip it, and solder it on using wires so that it floated about 1-2 cm above the board upside down. I was impressed ;-)
Flux, practice, flux, good tools, flux, good lighting, flux.
Really, learning how to braze and weld steel helped me a lot. Heat control is extremely important. Heat flows, and you learn how to position your iron to get the heat where you need it. You can also tell how hot something is by how the puddle looks.
Basic soldering is easy to improve with practice. A couple of hours with a few hundred assorted components and some scrap PCBs will help to perfect the steps (tin the bit, hold the bit to the joint, apply the solder, pull the solder wire away, pull the bit away, inspect the joint). Things are a little bit trickier today because the alloy of the solder and the flux you use do make a difference, but lead-free solder today is very good.
Then practice with PVC insulated wire. The PVC insulation shrinks if it gets hot so this is good practice at speeding up.
I looked through the NASA document and most of it is not advance technique.
EDIT: people are saying "good tools" and I wanted to say a bit more about that. You can buy an iron for a few dollars from ebay. Don't use those. But almost any iron above that is going to be fine. I'd prefer to always be using Weller or Hako, but failing that any of the chinese no-name irons with temperature control are going to be good enough for most hobbyist projects. I love my Lindstrom and Bahco snips, but you can get good-enough cutters much cheaper. If this is something you're going to be doing all day every day you'll want to invest in the tools.
> Basic soldering is easy to improve with practice
If you're looking for a practical way to do this, consider hand-wiring your own keyboard. I did this a couple years back, with 75 keys or so. With the diodes on each key, it was about 225 solder joints plus whatever extra to land the matrix leads on the microcontroller.
Good variety, too. Some were wire-to-wire joints, others were wire wrapped around post, and a few where through-hole joints.
If I look at the underside of that keyboard, I can tell where I started and where I finished by how clean the solder joints look :)
I wouldn't give up my Metcal for anything else I've seen, but if I were broke, I'd buy one of the Chinese Hakko clones and eventually replace the tips with Hakko tips.
I found these videos from eevblog extremely helpful when I setup my home lab after working as an EE for many years but never really learning to solder.
Temperature control is so inexpensive these days you should not bother with heat control. Unless by heat control you mean the right size tip for the job, in which case we agree.
Not for you, but for those that might not get the point: the world of electronics assembly is really strange - it's like the whole past is present with us. It's 2021 and you can still buy a Weller soldering station with a knob for open-loop power (= heat) control for more than you'll pay for a cheapish station with closed-loop temperature control from any number of Chinese brands. (google "Hakko 936 clone"), Closed-loop temperature control is much superior to open-loop power control.
1) On page three, that exposed end of the lead is a defect if you're NOT going to conformally coat the board. If you don't know what a conformal coating is, it's a defect.
2) On page eleven, flux residue. Sure, for NASA's applications. For most commercial applications, the IPC rule is "excessive flux residue". For prototypes, one-offs, and hobby projects, my rule of thumb is, if you can't measure the difference removing flux makes, leave it there. Incompletely cleaned flux can be more corrosive than undisturbed flux. Also, no-clean flux should be called "never clean" flux. Never clean it with ordinary methods because you will never get it clean.