EEVblog Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: wbooth.clearscene on April 07, 2014, 11:15:13 am
-
Hi,
I'm currently repairing an 34401A which has an intermittent resetting fault. It appears to have an enormous amount of tin whisker growth. I've been using a jewellers loupe and a pin to remove them, but it's tough going.
I was wondering if there is a better way to remove these whiskers easily? I was thinking of feathering a smd hot air gun around all the joints, or even running a soldering iron over all the chip joins. Not sure if that would help? Maybe there is a better approach (than at least a loupe/pin)
Cheers for any suggestions.
-
Fix up a DeLorean, go back in time, smack EU politicians in the head, use leaded solder. :palm:
-
ipa & toothbrush?
maybe not. i guess you don't need them hanging around in random locations.
compressed air?
-
Somewhere on www there's a white paper on the results of an extensive 10(?) yr study in Europe on elimination of tin whiskers.
The ONLY solution the scientists (metallurgists etc) that solved it .. add a bit of lead !!!! I have a copy somewhere, I'll try to find it.
You need to resolder / re-wet the connections with good ol' 60/40. It's what I've been doing for years for failed vehicle ECUs.
-
I'm currently repairing an 34401A which has an intermittent resetting fault. It appears to have an enormous amount of tin whisker growth.
Can you take a picture of one? I've not seen a tin whisker.
Thanks!
edit: google images has enlightened me. Holy balls, those are small. Couldn't a conductive brush just brush them right off? canned air after to remove them? Can you just reflow the thing to remove them?
-
ONE example of studies / papers on the issue !! EVERY single study comes to the same conclusion - ADD lead.
It can't be fixed any other way. The whiskers will just regrow if you try to fix it any other way. That's why the list of ROHS exemptions
in Europe has been growing steadily every year.
www.theguardian.com/technology/2008/apr/03/research.engineering (http://www.theguardian.com/technology/2008/apr/03/research.engineering)
-
ONE example of studies / papers on the issue !! EVERY single study comes to the same conclusion - ADD lead.
It can't be fixed any other way. The whiskers will just regrow if you try to fix it any other way. That's why the list of ROHS exemptions
in Europe has been growing steadily every year.
www.theguardian.com/technology/2008/apr/03/research.engineering (http://www.theguardian.com/technology/2008/apr/03/research.engineering)
I understand the long-term fix is to add lead, but could you just point a hot-air gun at all the solder every few years and melt them back to their respective joints? they're very thin, and would reach melting temp long before the joints themselves would.
I'm not asking because I'm anti-lead, I'm asking in case I ever come across this and I need to fix it in the short-term.
-
I wonder about maybe using a fine brass brush and a gentle hand?
-
I understand the long-term fix is to add lead, but could you just point a hot-air gun at all the solder every few years and melt them back to their respective joints? they're very thin, and would reach melting temp long before the joints themselves would.
I'm not asking because I'm anti-lead, I'm asking in case I ever come across this and I need to fix it in the short-term.
MAYBE ! IF the whiskers have pushed themselves through the solder mask, or any other conformal coating, then you're SOL :-)
The whiskers can grow through anything. ALSO, sometimes they grow "into" the leads of the components, causing alloy breakdown.
I've had BOXES of Audi ECUs with whole capacitor batches "rotted" through. IF I get a ROHS repair job these days, I FIRST solder suck
a few different component types to see if there is any internal "corrosion".
IF you can vouch for whiskers not embedded in resist etc, then yes, a heat job will get you out of trouble for a while.
-
Are you sure it's tin whiskers? Can you post a good picture? I've seen insides of maybe 50 of those produced in different years and they all were made using leaded solder. IIRC test instruments are exempt from lead-free requirements for this very reason.
BTW, lead paint test kits sold in hardware stores work very well on PCBs too.
-
I have been in electronics design and manufacturing since 1978.
I think tin whiskers have been gradually made into a non issue by a combination of a number of factors
- lead pitch (on TSOPs etc) stopped getting smaller after about 0.5mm (BGAs took care of it)
- SAC solder (silver copper and tin) seems to be OK and does "just about work" albeit at a vastly higher price; SAC305 is the least bad
- consumer equipment gets chucked out after a few years
- whiskers grow mainly out of stressed/formed surfaces, and since IC leads have gone lead-free there were no more whiskers growing out of them
- loads of small firms used the ever so useful "control and monitoring" ROHS exemption (due to end c. 2017)
- military and medical have exemptions anyway
- there are expensive conformal coatings which stop the whiskers
But the business picture has been pretty bad. No new exemptions have come in. That article is wrong I am sure. Swatch didn't get an exemption. The EU told it to stick its business up its backside, or comply with the ruling.
What has happened is that entire product ranges have been scrapped, at a massive cost, because some component could not be obtained in an ROHS version.
Later, the multinationals have aggressively demanded total ROHS and REACH compliance, sending out spreadsheet lists with about 200 substances and demanding a signed certificate that none of these are present, thereby rendering any exemptions useless (unless one lies, which most people did, but eventually people go lead-free and accept a much higher SMT reject rate). The big US companies have been especially aggressive, and now they are trying to screw their suppliers with the Conflict Minerals garbage, where one is supposed to identify the smelter that produces the metals one is using!
And it's no use telling your customer that you are compliant via an exemption. They will simply not understand. The US based multinationals I sell to use remotely located offices (Philippines, eg. which can barely speak English) to screw their suppliers on compliance and if you return the questionnaire with some words like "exemption" all hell will break loose.
For small scale stuff, repair/service work, hobby soldering, just use tin/lead solder. It works much better :)
Also if you sell direct to small end users, off a website, etc, nobody is going to question what solder you use. The only time you have to comply in practice is when selling to big companies who demand the various statements of conformity.
-
Are you sure it's tin whiskers? Can you post a good picture? I've seen insides of maybe 50 of those produced in different years and they all were made using leaded solder. IIRC test instruments are exempt from lead-free requirements for this very reason.
I'll hopefully grab a "fresh" one this weekend. The units I've repaired were definitely RoHS solder. I took a sample to a University
I mentor at, because they were interested. They're always looking for projects. I admit, the results weren't conclusive, and the "active"
alloy mix was likely mostly responsible, but the best guess was that the stress "mechanism" of whiskers definitely assisted in whatever
"corrosion" process eventuated. Sometimes it's difficult getting students excited, or keeping an attention span :-)
-
Toyota's Sudden Acceleration Problem May Have Been Triggered By Tin Whiskers
http://www.huffingtonpost.com/2012/01/21/toyota-sudden-acceleration-tin-whiskers_n_1221076.html (http://www.huffingtonpost.com/2012/01/21/toyota-sudden-acceleration-tin-whiskers_n_1221076.html)
Can we sue the EU pls?
-
Toyota's Sudden Acceleration Problem May Have Been Triggered By Tin Whiskers
If you do a "Tin whiskers failure" search, you'll find literally 100s of famous failures due to whiskers.
And naturally, many many are kept low key. Wouldn't want to upset the RoHS folks :-)
-
Tin whiskers are not new what is new is the very low current devices which cannot source enough current blow the whiskers away.
-
I forgot to add another factor:
Whiskers tend to grow out of shiny surfaces. A matt plated surface is unlikely to generate them.
In the old days one did hot tin dipping or bright tin plating which was bad. For many years now, ICs which have tin plated legs have a matt surface on them.
"what is new is the very low current devices which cannot source enough current blow the whiskers away."
I don't think so... a whisker may need a good fraction of an amp to blow it.
The EU is full of **** and full of highly paid people who cannot do a real job who love these regs because they provide an easy way to make money.
But I really don't think that whiskers are an issue in current production methods (SAC305 solder). I think the industry has just got lucky on this topic... I use contract assembly houses who can speak off the record. Their business has moved from ~50% tin/lead (5 years ago, with lots of people using the "control and monitoring" exemption) to ~90% lead-free. The move has been driven purely by pressure from multinationals' compliance officers screwing the supply chain.
Anybody who sells direct to small end users would be stupid to use lead-free because it costs more, the yield is less, you get a lot more bad joints which look good on a visual inspection, and the joints are brittle so any vibration/movement is going to cause an intermittent failure.
-
I have been in electronics design and manufacturing since 1978.
I think tin whiskers have been gradually made into a non issue by a combination of a number of factors
- lead pitch (on TSOPs etc) stopped getting smaller after about 0.5mm (BGAs took care of it)
- there are expensive conformal coatings which stop the whiskers
Which conformal coatings stop whiskers? I've seen them grow through urethane based, silicone based and even parylene. And as for the pitch, there are plenty of examples around of whiskers growing 50mm in length along boards in studies I've seen. I agree that it's less of a problem, but I think it's far from solved
-
I found some conformal coatings when I googled on whiskers a year or so ago. Here are some of the links I found in Usenet
http://www.national.com/packaging/leadfree/tin_whiskers.html (http://www.national.com/packaging/leadfree/tin_whiskers.html)
http://www.ewh.ieee.org/soc/cpmt/newsletter/200506/tinwork.html (http://www.ewh.ieee.org/soc/cpmt/newsletter/200506/tinwork.html)
http://www.microchip.com/stellent/groups/environmentalhealthsafety_sg/documents/devicedoc/en025487.pdf (http://www.microchip.com/stellent/groups/environmentalhealthsafety_sg/documents/devicedoc/en025487.pdf)
http://nepp.nasa.gov/whisker/background/index.htm (http://nepp.nasa.gov/whisker/background/index.htm)
http://nepp.nasa.gov/whisker/reference/tech_papers/kadesch2000-article-effect-of-conformal-coat-on-tin-whisker.pdf (http://nepp.nasa.gov/whisker/reference/tech_papers/kadesch2000-article-effect-of-conformal-coat-on-tin-whisker.pdf)
http://www.calce.umd.edu/lead-free/tin-whiskers/WoodrowConfCoatPaper.pdf (http://www.calce.umd.edu/lead-free/tin-whiskers/WoodrowConfCoatPaper.pdf)
Also google "tin whisker" "conformal coating"
The stuff I found was American aerospace stuff and appropriately hard to buy in small quantity...
It seems that to get whiskers to grow readily from non-stressed surfaces you need elevated temperatures and elevated humidity, and the two are rarely present concurrently.
I am not saying the problem has gone away. I am just saying that a coincidence of various factors has made it a non-issue for most electronics - evidently. I say "evidently" because e.g. the Japanese went lead-free with huge enthusiasm maybe 10 years ago, and you see very fine pitch components (TSOPs down to ~ 0.35mm) on the controller board for hard drives, etc.
If I was making industrial gear which needs to work at say +60C or more and high humidity then I would definitely use a top grade coating.
And tin/lead solder, with the C&M exemption - if somebody asks whether you are ROHS compliant you can truthfully say YES because the exemption IS compliance. What you should not do is actually state you do not use lead...
However high humidity will eat your PCB anyway...
And as for the pitch, there are plenty of examples around of whiskers growing 50mm in length along boards in studies I've seen
If that was happening under anything resembling normal circumstances, electronics as we know it would be unworkable.
-
Whiskers? The zener in the centre is growing a beard! It should look more like the one on the right - although this too is showing signs of bubbling-up. Both zeners presently work, but it can only be a matter of time ...
-
Yes - they are coming out of bright tin surfaces. Look up some of the links I posted above. They are full of pics of metal can transistors which are full of whiskers inside. These metal cans were bright tin plated as sheet and then pressed (deep drawn) out of the sheet.
Looking at those components, that PCB goes back to the 1970s and possibly earlier. A lot of my products have been working for 20 years now, 24/7 (although to be fair ALL done with tin/lead solder) but if I was making stuff which had to work for 40 years I would absolutely not use unleaded solder.
Nor would I use any trimpots especially those cheap and nasty carbon ones ;)
-
Looking at those components, that PCB goes back to the 1970s and possibly earlier.
Spot on. It's the power supply section from a Marconi TF2000 oscillator. The design is very late sixties. The components in this unit are dated 1970/71
Nor would I use any trimpots especially those cheap and nasty carbon ones ;)
Better than those open-frame types, surely?
-
Open frame cermets are better than any carbon pots.
But any type of pot will give trouble eventually, via oxidation of the resistive element.
The least bad option is an immersion sealed (washable) multiturn pot. There are some sealed single turn pots but not many. Sealing works a lot better on a multiturn one because you have just one little o-ring.
In the 1970s I used what was then a very popular high quality Allen-Bradley trimpot. Google/images fails to find it now but it looked very close to this one
http://i.imgur.com/y9EzvRE.png (http://i.imgur.com/y9EzvRE.png)
Those were very good. Nowadays they come with a plastic case to make them look like they are enclosed...