How to build a reliable PCB

[williamson]

Since I cannot recommend lead-free soldering for hand-soldering, I wanted to post that I support "green". I put less than 2000 miles a year on my ICE car, and I am building an electric car.

[williamson]

Topic: Lead free soldering iron:The new Weller model WESD51 is rated for esd protection. Lead-free requires that the soldering iron be adjusted for 20 deg. C hotter than for lead. This requires an iron with a potentiometer and a digital readout. The most popular of these irons were the EC2001M and 2002.
Topic: heat damage: It is possible to damage an led by lingering with the iron, while soldering very close to the body. The safe way is to hold a pair of needle-nose on the leads, up against the body. This requires a helper, unless you have 3 hands.

[williamson]

Topic: learn to solder: Electronic Kits offers a kit called "learn to solder" kit # AK-100, $19.95. soldering will be discussed here, but it is only one part of the required knowledge you want.
Topic: Static electricity: Integrated circuits (dead bugs) especially CMOS IC's, are susceptable to static elecricity. This may not come as news to you. CMOS IC's are identified by part #'s starting with "MC14xxxB" or "CD4xxxB", or similar. But there is a safer path: parts suppliers ship their parts in static-safe bags. Therefor, keep all parts in the bag they come in. DON'T open the bags. Only open the bag the minute before you will install that part. The second reason is that moisture in the air, or oxygen, cause a very slow corrosion of part leads. Therefor keep the bags sealed as shipped. When you remove partial contents, reseal the bag--as best as is easy. Static-free bags are identified by the slight tint or color-any color. Keep some of these bags. You will want to place any PCB with ANY parts on it, into one of these bags, between work sessions, or when the board is built-that is, up to the minute the board is installed!

[jfitzpat]

Originally Posted by williamson

Mr. Smalls is absolutely right. We have 100 years of technicians soldering, and I've never heard of one getting lead poisoning.

I'm sorry, but that doesn't sound like the sort of argument an engineer makes. Perhaps you are speaking from the point of view of a manufacturing technician?

The primary problem with leaded solders is measurable, direct risk to workers in manufacturing, not secondary risk from waste. That latter exists, but exists with many, many compounds, which is why waste management engineering has to take other steps to prevent exposure and contamination of groundwater, etc.

The danger to workers is measurable, and still of concern in exempted areas of aerospace and the military (ex. see the recent UC Davis study and the work of the JG-PP (see below)). That is, folks working with leaded solder are at statistically higher risks for lead related health problems.

In design engineering, we look at larger statistical samples. That is, making one run on the bench is not the same thing as predicting reliability of 1,000,000 in the field. Also, you are required to account for spectral factors. Severe heavy metal poisoning results in some very obvious symptoms, incoherence, impaired memory, neuromuscular problems, etc. But smaller doses often result in statistical correlation to more subtle problems, or serious problems over a much longer period of time.

This form of statistical-universe-of-one-as-'obvious' argument is a very human reaction. I'm tempted to go over a photo someone sent me recently of a book in a snowdrift, but that is too politically charged, so let's take alum. Aluminum salts are the most common metallic element on earth - the third most common element in the earth's crust overall, so there is no particular reason to suppose that we are not exceedingly well adjusted to it. But alum toxicity, as well as causal relationships to serious long term ailments like alzheimer's disease are quite real. In fact, we add small amounts of alum to some vaccines to trigger a severe immune system response in the human body.

But how should we react to this knowledge? Humans are, instinctively, highly reactionary and subjective creatures. So, the most two common reactions are: 1) panic and be terrified of alum. cookware and products like deodorants (both of which the FDA has deemed to generally safe) or 2) interpret the singular absence of immediate and obvious harm as a definitive proof of 'safe' (ex. "Well I eat Jiffy Pop and I don't become incontinent and yell at clouds, so it must be a load of hooey...") Either way, the instinctive, or emotional reaction has no logical relationship to objective risk assessment and management.

Originally Posted by williamson

Each person viewing this site will have to decide his own priorities, when it comes to lead-free.

If they live here in CA with me, I sincerely hope not. I don't get to pour mercury or MTBE down my ground drain and I don't have to look much beyond the crowd at the airport or current primetime TV to make the case that vast numbers of my fellow citizens are not properly equipped to make intelligent decisions about public heath and environmental toxicology on an individual basis.

Originally Posted by williamson

This site is called "a reliable". As stated earlier, the military doesn't allow it because of "wetting". So if you are inexperienced in soldering, you should take into account that lead-free may not "wet" as well.

The DoD and NASA both participate in something called JG-PP, the Joint Group from Pollution Prevention. JG-PP and JPL both have done studies on the implications of lead free versus leaded solder. There is also a report from the JC-AA (Joint Council on Aging Aircraft), but that is really just a summary of the JG-PP '06 and '07 study reports.

The super short summary is that solder joint reliability turns out to be a lot more complicated than leaded vs. unleaded solder. Lead free compounds appear to be just as reliable in many applications. In those that the joints fail, it isn't generally a deficiency in the solder compound so much as a systemic (design or total material) or a manufacturing problem (remember, it is generally unleaded solder being directly substituted into existing designs and manufacturing processes).

The JPL study is interesting in that it didn't just look at subsystem reliability, but failure based on application technology. Hand soldering had the smallest discernible difference in reliability (essentially zero in industrial temperature range applications). On the flip side, hand soldering represents the highest exposure and statistical risk to the workers.

So, with no disrespect to your personal experience, I can't find any evidence from the military and aerospace studies to support the idea that hand soldered lead free boards are not reliable in consumer automotive applications.

Remember, the military definition of 'reliable' and the common understanding are two different things. A brand new commercial airliner is not reliable enough for military applications, even though it uses lighter and stronger alloys than a mil spec aircraft. The modern nav system in the plane I most often fly is lead free, as are the ECUs and brake module in my daughter's car, and I'll never lose a moments sleep about either.

Consumer electronics are highly sensitive to changes in reliability. Because of the margins, a warrantee repair is catastrophic to the bottom line. Following the work and articles on statistical quality control, consumer manufacturers have, by and large, found the move to lead free fairly smooth.

Automotive applications have some other challenges, but I'd readily bet good money that, say, Toyota's current woes have everything to do with designs and firmware, and nothing to do with the type of solder.

-jjf

[MPaulHolmes]

I tried lead free solder once, and I thought it was a little harder. I knew almost nothing about soldering at the time, though. I think williamson is just saying that if someone is going to make a little control board for this motor controller (one time solder thing), it might be a little easier to use lead solder instead of lead free, and the one time use of lead solder will almost certainly not have any negative health consequences. At least that's how I'm thinking about it.

Lots of people that have ordered the control board from me have had no soldering experience whatsoever, and have no interest in continuing to solder after the controller is done. They just want to get their car on the road without paying $2000 for a freeway capable controller.

However, I'm going to go out and get some lead free solder and give it another try. I don't want to do any behavior that causes a statistical increase in health problems over the long term.

[williamson]

jfitzpat raises important points that we should address. People should get some leadfree solder, and try it for themselves. Maybe someone with lead-free solder, (that they no longer need) could drop about 12 inches in an envelope to any one requesting some. A roll is about $30.00 Secondly, as I think I mentioned, we should never discard a lead-soldered PCB- or any electronics, in any way except approved re-cycling for lead-soldered products. So, if you decide on lead solder, you commit to responsible re-cycling.

[jfitzpat]

Originally Posted by williamson

jfitzpat raises important points that we should address. People should get some leadfree solder, and try it for themselves. Maybe someone with lead-free solder, (that they no longer need) could drop about 12 inches in an envelope to any one requesting some. A roll is about $30.00 Secondly, as I think I mentioned, we should never discard a lead-soldered PCB- or any electronics, in any way except approved re-cycling for lead-soldered products. So, if you decide on lead solder, you commit to responsible re-cycling.

When lead poisoning used to be called "plumber's cough" and "plumbism", your 100 years of leaded solder quip seemed clueless. But this new tact is just silly.

1.2 oz of .032" diameter SAC305 solder (with a 'no clean' flux core) is about $4.50 at Radio Shack. Probably less from Digikey. If the handy little plastic dispenser isn't enough, a .25 lb. roll is $12-$14.

Compliance in CA has been since '07, so most of us engineering types here have been using lead free to hand solder surface mount parts for 3.5-4 years. The bottom line? It is all about a more heat and getting used to it. SAC305 (which I'd recommend for the casual electrical connections) melts at ~217C. SN100 (which you'll see in some manufacturing) melts at ~228C. A 63/37 tin/lead alloy melts at ~183C.

As a practical matter it means that you need to a) be more careful about picking solder diameter b) you need to be more careful about keeping your tip clean and properly 'tinned' and c) you want to be careful about mixing and matching.

That is, 'this tub of flux I've always used' may not be high temp and may turn black at the higher temps involved. Also, the alloys don't flow well on each other. That part is of note because lead free parts, which are becoming the norm, are harder to solder reliably with leaded solder than with the lead free alloys they are designed for.

We could pretend that we are planning an arctic expedition by dogsled, or face the reality that one of the largest densities of electrical hand solders in the country went through this, largely without pain, years ago. Clean and re-tin your tip (or, if you are lazy, buy a new tip), practice a few joints, and get on with your life.

-jjf

[williamson]

jfitzpat is right: there are smaller quantities of solder available, and I should have mentioned them. Because of the higher temperature he points out, you can see that you really need the temperature controlled iron I recommended for lead-free work.

[jfitzpat]

Originally Posted by williamson

jBecause of the higher temperature he points out, you can see that you really need the temperature controlled iron I recommended for lead-free work.

I have the same Weller WP25 in my field kit that has been bouncing in there for years. The tip is 750 degF, which comes out to about 400 degC. Again, it is about being a bit more clean and precise and smart about heat management.

If this one takes a walk and I run out again (which is how this one got in my kit), I'd probably pick up the WP35, which is 100 degF (about 50 degC) hotter. Not so much because of lead free, but because I inevitably end up soldering heavier wires together, not just board work, and the WP25 is underpowered for doing that quickly regardless of the solder.

For bench board work, you want an adjustable iron regardless, really two - for even basic repairs. It's now a surface mount world.

-jjf

[MPaulHolmes]

Maybe there should be 2 threads: One for people that solder for a living, and this one, which is basically in the open source dc motor controller forum because it's here to help people reliably solder a single thru-hole control board for this motor controller. I know that "How to build a reliable PCB" sounds pretty broad, but it's intended scope really wasn't meant to be the final say on industry solder policy. There have been a couple problems with solder not sticking on this control board, or use of too much solder, crossing the isolation rings, from people that have never soldered in their life (and many happen to live in one of the 49 states where lead solder is allowed).