Author Topic: Component Failures.  (Read 7550 times)

0 Members and 1 Guest are viewing this topic.

Offline szaraTopic starter

  • Newbie
  • Posts: 9
Component Failures.
« on: September 07, 2014, 12:15:03 pm »
Hi, this is my 1st post on the EEVBlog forum.  I love the enthusiasm that Dave brings to the lab bench!

Question: Can anyone help me understand why my original Heathkit test equipment (scope function generator, curve tracer, etc.) that was built with the utmost of care, and stored in an unfinished basement after a move (for around 15 years) have failed?  I find many open resistors, shorted tantalums, and even dual-in-line resistor networks that are partially ruined?  These were stored for this long term wrapped with bubble wrap or newspapers in cardboard boxes that were taped shut for all this time.

Thanks for your insight.
Steve
 

Offline wagon

  • Frequent Contributor
  • **
  • Posts: 289
  • Country: au
Re: Component Failures.
« Reply #1 on: September 07, 2014, 12:46:15 pm »
Was there a bit of moisture in the air in the room where it was stored?
Hiding from the missus, she doesn't understand.
 

Offline amyk

  • Super Contributor
  • ***
  • Posts: 8270
Re: Component Failures.
« Reply #2 on: September 07, 2014, 01:08:51 pm »
Those are typical components that go bad with time - as mentioned above, carbon resistors are particularly sensitive to moisture. Tantalums rely on electrolytically-deposited and very thin oxide dielectric, which is also quite sensitive to overvoltage and possibly may have degraded over time.

On the other hand, the semiconductors are probably still OK.
 

Offline wagon

  • Frequent Contributor
  • **
  • Posts: 289
  • Country: au
Re: Component Failures.
« Reply #3 on: September 07, 2014, 01:42:31 pm »
Does the instrument smell 'musty' or 'mouldy'?
Hiding from the missus, she doesn't understand.
 

Offline szaraTopic starter

  • Newbie
  • Posts: 9
Re: Component Failures.
« Reply #4 on: September 07, 2014, 10:35:07 pm »
Thanks for the replies.  None of the equipment smells musty or is rusted or anything.  As a matter of fact, they all look as new as the day I completed the kit, and all the solder joints look as shiny as day one.  The only thing that I see that is a cosmetic flaw is a very thin, white coating on the red knobs(almost like a thin layer of spray paint), but it takes a bit of a strong rub with a dry cloth and it comes off.  For example, the IG-1271 waveform generator's 15VDC regulator had zero output, but when I probed the surrounding resistors, caps, and even one of the dual-in-line resistor network chips I found that some resistors are open, and some caps are shorted.  What was puzzling, some of the resistors in the r-network were still fine.

Because everything seems to be so clean (mint), I am going to replace most resistors and caps, then continue troubleshooting to see if I can fix the thing.  These have sentimental value to me because I used to just love building Heathkits and had learned a lot from them.

BTW, I always had a dehumidifier (or two) running constantly, and have no water or heavy moisture issues in the basement.

Could it be the thermal cycling (I mean cold winters and hot summers) even though the equipment had been off and packed in boxes?

Lesson learned, though.  No more electronics test equipment down there until I finish the basement and have a more uniform climate (my new scope or programmable power supply, etc.)
 

Offline David_AVD

  • Super Contributor
  • ***
  • Posts: 2806
  • Country: au
Re: Component Failures.
« Reply #5 on: September 07, 2014, 11:17:33 pm »
Old tantalum caps going short circuit in storage is reasonably common.  That in turn can take out resistors in their current path.
 

Offline szaraTopic starter

  • Newbie
  • Posts: 9
Re: Component Failures.
« Reply #6 on: September 08, 2014, 04:27:17 am »
David_AVD:  Thanks, didn't know that, but that sequence of events makes sense.  What do think if I were to replace all resistors with metal film, instead carbon film?  I'd be putting in 1% resistors.  Would metal film work same/better/or worse as substitutes?
 

Offline poorchava

  • Super Contributor
  • ***
  • Posts: 1672
  • Country: pl
  • Troll Cave Electronics!
Re: Component Failures.
« Reply #7 on: September 08, 2014, 04:38:31 am »
Actually the oxide on Tantalum caps tends to be produced thermally,  not electrochemically.

Anyway -  I have some carbon resistors in my drawer that are over 50 years old and they are all fine despite having been stored in a basement. The only potential problem seems to be oxidation of lead plating.
I love the smell of FR4 in the morning!
 

Offline szaraTopic starter

  • Newbie
  • Posts: 9
Re: Component Failures.
« Reply #8 on: September 08, 2014, 04:47:38 am »
What I meant is, I can't seem to find 1% 1/4W Carbon resistors on eBay, I can find plenty of 1% 1/4W Metal Film.  Would it be okay to substitute these for the carbons?  Also, can Tantalums be replaced with Mica or ceramic with no issues (keeping the voltage rating the same or higher).

What you said about the way this failure might've occurred got me thinking that if I had checked for short caps BEFORE 1st turning power on (on an old, long-stored piece of equipment), I might've spared the resistors from being fried.  Does that make sense?
 

Offline szaraTopic starter

  • Newbie
  • Posts: 9
Re: Component Failures.
« Reply #9 on: September 08, 2014, 05:00:37 am »
poorchava: Thanks for your input on the Tantalum failures.

David_AVD: That last question about checking the caps on an old, stored piece of equipment before powering it on was meant for you as a result of your suggestion of how the failure in the resistors might've been caused in the 1st place.

Thank you both for this helpful insight.
 

Offline David_AVD

  • Super Contributor
  • ***
  • Posts: 2806
  • Country: au
Re: Component Failures.
« Reply #10 on: September 08, 2014, 05:02:13 am »
Maybe.  The problem is that sometimes old caps seem to go S/C at the moment of power up.

Replacing them before powering up after a long time in storage may be the only way to stop more damage.
 

Offline Legit-Design

  • Frequent Contributor
  • **
  • Posts: 562
Re: Component Failures.
« Reply #11 on: September 08, 2014, 09:49:05 am »
Maybe.  The problem is that sometimes old caps seem to go S/C at the moment of power up.
Replacing them before powering up after a long time in storage may be the only way to stop more damage.
One variable frequency drive manufacturer had a note about slowly building up the oxide layer inside electrolytics by slowly increasing the current and voltage after long periods of storage. The need for this went up as storage period was increased. I'll have to see if I can find the paper somewhere...
 

Offline wagon

  • Frequent Contributor
  • **
  • Posts: 289
  • Country: au
Re: Component Failures.
« Reply #12 on: September 08, 2014, 11:12:23 am »
I get high-voltage capacitors in electric fence energisers fail after being stored for a year or more.  Most of these parts are 8 to 30uF and rated at 900 to 1200VDC. 
Hiding from the missus, she doesn't understand.
 

Online tautech

  • Super Contributor
  • ***
  • Posts: 28371
  • Country: nz
  • Taupaki Technologies Ltd. Siglent Distributor NZ.
    • Taupaki Technologies Ltd.
Re: Component Failures.
« Reply #13 on: September 08, 2014, 11:15:34 am »
I get high-voltage capacitors in electric fence energisers fail after being stored for a year or more.  Most of these parts are 8 to 30uF and rated at 900 to 1200VDC.
@ wagon
Have you had any success in re-forming them?
Avid Rabid Hobbyist
Siglent Youtube channel: https://www.youtube.com/@SiglentVideo/videos
 

Offline wagon

  • Frequent Contributor
  • **
  • Posts: 289
  • Country: au
Re: Component Failures.
« Reply #14 on: September 08, 2014, 11:39:12 am »
I get high-voltage capacitors in electric fence energisers fail after being stored for a year or more.  Most of these parts are 8 to 30uF and rated at 900 to 1200VDC.
@ wagon
Have you had any success in re-forming them?
In the bin, and replace.  I don't want the callbacks. (they aren't expensive really, and are often 10years+ old)
Hiding from the missus, she doesn't understand.
 

Offline Legit-Design

  • Frequent Contributor
  • **
  • Posts: 562
Re: Component Failures.
« Reply #15 on: September 08, 2014, 11:53:18 am »
Attached some note from ABB. ABB should be trust worthy source, I think Chris Gammell worked for them and talks about them on The Amp Hour. Reforming current is 500mA MAX. and reforming time is in hours for every year stored, so 3 years of storage means 3 hours of reforming.
 

Online Dr. Frank

  • Super Contributor
  • ***
  • Posts: 2384
  • Country: de
Re: Component Failures.
« Reply #16 on: September 08, 2014, 01:54:16 pm »
Hello,
if properly soldered 15 years ago, your complete electronic circuit is much more durable compared to the storage of the single components.
(I have many measurements instruments here, which are 10, 20, even 40 years old and still working very well)

That especially goes for the solder joints which may oxidize on their surface only, which would prevent a good soldering to-day, but in a soldered state, the bulk of the joint is obviously ok, and that counts.
These solder junctions on "stiff" SMD packages may deteriorate only by strong temperature shock of -40/+60°C, > 500 cycles , (as in a car over the years) , but definitely not by storage in a basement of a house. 
Also, vibration of several g or bending of the PCB or so might crack the solder joints, or the components themselves.

There are some classes of electronic components,  passive components, which do not withstand storage so long, especially when the circuitry is left unpowered.
All semiconductors normally survive a very long time.

These sensitive components are especially wet electrolytic capacitors, which either dry out, or their "forming" decreases. That is their inner Al2O3 dielectric film on the Al electrodes.
If you power your circuitry up, they draw a lot of current, due to the re-forming process, and then may blow up, or may damage the surrounding circuitry.

Old circuits have to be powered up very slowly!
(How did you do it, after 15 years??)

Tantalums are not prone to that effect, because they are "dry" electrolytics.
But those old ones may be produced badly, and in the circuitry charging current limiting may not have been foreseen, so they also may burn.
If you want to replace them by other technologies, you have to pay attention concerning low ESR value, if they sit around linear stabilizers, otherwise latter one may start oscillating.

From what you describe, that you stored the electronics in a closed (taped) box, you may have enclosed a lot of humidity inside the box.
Especially paper attracts humidity from the environment and creates a fine humid atmosphere around the PCB.

Don't you believe that? Those paper packing cases which are left in the basement, begin to smell heavily after a few months already, guess why.
If you have left dehumidifiers in the basement all the time, the whole room very obviously has problems with high humidity..
 
You did not add  a hygroscopic salt bag inside the box?

Well, then several of your components definitely suffer from rust , which probably can be seen under a microscope:

If you initially used cheap components, the protective coating may have had microscopic pin holes, where the humidity entered and destroyed the resistive layer. Especially thin film technology is sensitive to that effect. carbon resistor maybe also effected , but you can easily replace these by the much better thin film technology, if you also have through hole components on your PCB.

Then, if you used solder with a lot of flux, and did not properly remove that, and did not protect by plastic spray or so, this flux may have aggregated humidity even at lower level of air humidity, and then a chemical process using salts may have destroyed the solder joints, or the PCB traces.

Another effect are whiskers, if you used bad solder alloys. They may cause shorts, and are barely visible.

Also, you may have introduced ESD failures 15 years ago, which now show up.


High resolution, sharp, macro photos of your PCB would help.

(I'm working in the development of an automotive electronics supplier, being sort of a component technology expert, and may tell a lot about long term storage and component failures)

Frank
« Last Edit: September 08, 2014, 02:43:53 pm by Dr. Frank »
 

Offline szaraTopic starter

  • Newbie
  • Posts: 9
Re: Component Failures.
« Reply #17 on: September 09, 2014, 12:34:37 am »
Dr. Frank: The dehumidifiers were used as a preventive measure so that I don't get a musty basement during the summer months and when we had frequent or heavy rains during the spring months.

How would you turn on old equipment very slowly?  I have a curve tracer that I haven't powered on yet after that long storage.

I like your other suggestions, like the detrimental effects of the paper in boxes and not using the salt bags, etc.

Thanks for all the insight.  Too bad I did not know this, I would've powered the equipment frequently.

Best regards,
Steve
 

Offline szaraTopic starter

  • Newbie
  • Posts: 9
Re: Component Failures.
« Reply #18 on: September 09, 2014, 12:53:24 am »
By powering slowly, do you mean applying incremental AC volts to the mains with a rheostat transformer?  Or is there another methodology?
 

Online tautech

  • Super Contributor
  • ***
  • Posts: 28371
  • Country: nz
  • Taupaki Technologies Ltd. Siglent Distributor NZ.
    • Taupaki Technologies Ltd.
Re: Component Failures.
« Reply #19 on: September 09, 2014, 01:06:12 am »
By powering slowly, do you mean applying incremental AC volts to the mains with a rheostat transformer?  Or is there another methodology?
That's normally the accepted procedure.
Without a Variac or similar, I would use in-series domestic incandescent light bulbs, starting at say 15 W and slowly increase the wattage in a manner similar to the methodology outlined in the pdf posted by Legit-Design.
A handful of bulbs has to be a much cheaper option IMO.
Also gives you a good range to use in a dim bulb tester that you might build for this purpose and keep for gentle power-up after appliance/instrument repair.
Avid Rabid Hobbyist
Siglent Youtube channel: https://www.youtube.com/@SiglentVideo/videos
 

Offline szaraTopic starter

  • Newbie
  • Posts: 9
Re: Component Failures.
« Reply #20 on: September 09, 2014, 01:41:27 am »
Thank you for these very useful tips.  I really appreciate it.

BTW, I just want to make a small comment:

Besides being a retired engineer and teacher, I am a professional wedding photographer.  I have frequented forums and read comments on websites like www.nikonrumors.com and www.dpreview.com, and I have to tell you guys, in my opinion, engineers and techies are more courteous, helpful, polite, and professional than many individuals (by all means, not all) on those and other forums, where insults and disrespect are sometimes tactics employed by some who disagree with you, or feel that they are smarter than you.  We all learn a little from each other, no matter what our experience level is.

I already love the feeling I get from my 1st post here.  Power to you guys (I mean boys and girls)  :)
 

Offline Seekonk

  • Super Contributor
  • ***
  • Posts: 1938
  • Country: us
Re: Component Failures.
« Reply #21 on: September 11, 2014, 05:55:19 pm »
I read an EICO ad in 1964.  It said the only piece of test equipment needed to test their transistorized amplifier kit was a lamp.  It has been standard equipment on my bench ever since.  It sure has saved a lot of transistors in amplifiers on the first power on after a repair.  (I put myself through college repairing high end audio.  Even had dinner with Saul Marantz.)  If I missed one of the components in the bias circuit the lamp would just glow.   So reform with a lamp.  As the leakage slows the lamp will increase reforming voltage.
 

Offline szaraTopic starter

  • Newbie
  • Posts: 9
Re: Component Failures.
« Reply #22 on: September 16, 2014, 02:01:29 am »
Seekonk, Would you be kind enough to provide a link to some article/how-to for specific details on how to use the lamp to gently/safely turn power onto a piece of equipment?  Should the lamp be connected in series with the equipment or a different setup, and what wattage should the bulb be?  Thanks
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf