What made these caps fail?

Thread Starter

strantor

Joined Oct 3, 2010
6,781
Was it the cold or the week of time off?

This evening I had the pleasure of replacing 9 VFDs in a conveyor panel because they all failed during our cold week. Ok, not all; only 6 out of 9, and a 7th one died while I was investigating. I thought it prudent to replace all of them.

These VFDs have been on (power applied, running or not) on average 99.9% of every day since 2011 and ran with load probably 75% of every day since then. Then there was a freak cold snap, temp down to 12F/-11C and blackout for 4 days. When it was time to go back to work they gave up the ghost. Capacitor schmoo leaking from them, literally dripping out the bottoms of the drives like honey. Strangely no input fuses blew on any of the drives.

I don't have any answers for those who keep asking. I have no idea how capacitors dry out from 4 days of disuse. I have no idea how 12f ambient could effect a capacitor when "cold" relative to capacitor specs is in the tens of degrees negative.

Help me out? WTH happened here? I've never seen anything like it.

Before install: the one in the middle died right after taking the picture:
20210222_130427.jpg

Drives out of the panel. The 3 on the right are the 3 that were lit up in the previous picture:
20210222_192923.jpg
20210222_192930.jpg

After install:
20210222_211930.jpg
 

SamR

Joined Mar 19, 2019
5,024
Wow! Just wow. What temp did the cabinet get to? With the blackout and no heat being generated in the cabinet I expect it got the full brunt of 12F. What was the temp @ startup?
 

Thread Starter

strantor

Joined Oct 3, 2010
6,781
Wow! Just wow. What temp did the cabinet get to? With the blackout and no heat being generated in the cabinet I expect it got the full brunt of 12F.
No idea. The cabinet is indoors in a warehouse. Not sure how cold it got in there but no pipes burst in the warehouse.

What was the temp @ startup?
I don't know. They didn't try to run the line until yesterday and it was probably 60f then, but if they left the power on (they probably did, they always do) then startup was whenever power came back and I'll probably never know when that was.
 

SamR

Joined Mar 19, 2019
5,024
I can't see how it's not environmental conditions somehow causing the problem. That many to fail at once does not make sense. I'd be on the phone with the rep who sold those to me and alerting purchasing about it. For industrial electrical equipment, it used to be a 35 year mean time before failure. I expected equipment I spec'd that unless it failed during project startup to still be there when I retired. We had equipment on the plant that was older than I was back then.

EDIT: Their shelf temp is what I would expect for any electronic device but their operating ambient temps lower limit is the freezing point. So sitting there blacked out they were fine temp-wise and you said the pipes didn't freeze so... Do you think they might have been hit by a power surge? Those caps may have been leaking for a while already. Did all of them fail on start up or run and then fail soon after?
 
Last edited:

Chris_E20

Joined Feb 17, 2021
3
That's a way to solve this problem, Vender can provide accuracy data for you it's undoubted.
sometime I look for and call vender to provide data, and highlight equipment issue warranty supply
anyway I hope you solve this problem faster.
 
Last edited:

kennybobby

Joined Mar 22, 2019
75
10 years is about the life of an electrolytic capacitor built in 2011. 10 yr is the rule of thumb i've always heard and this episode is more anecdotal evidence. i've seen datasheets showing expected lifetime as 20,000 hours; 10 years of 24/7 is close to 90,000 hours.

Look for bulging tops and other indications of swelling.
 

Thread Starter

strantor

Joined Oct 3, 2010
6,781
Do you think they might have been hit by a power surge?
Maybe, probably not. I can't think why there would have been a surge. There was no lighting. Perhaps all that switching during the rolling blackout phase generated a surge? I don't know why I that would have generated a surge but I don't know how it works either.

Those caps may have been leaking for a while already.
I think you're right. The capacitor snot had plenty of the ambient PVC dust embedded in it, which is generated from operation.
Did all of them fail on start up or run and then fail soon after?
Again, not sure. I showed up and 6 were dead. Tried cycling power a few times and on my 3rd power cycle the 7th never came back on.
 

Thread Starter

strantor

Joined Oct 3, 2010
6,781
10 years is about the life of an electrolytic capacitor built in 2011. 10 yr is the rule of thumb i've always heard and this episode is more anecdotal evidence. i've seen datasheets showing expected lifetime as 20,000 hours; 10 years of 24/7 is close to 90,000 hours.

Look for bulging tops and other indications of swelling.
I believe you, I've seen those numbers too. But I have come to expect at a lot more. I've seen them last 3-4x as long.

Or by "10 years is about the life of an electrolytic capacitor built in 2011" are you saying "they don't make them like they used to?" I wouldn't have any trouble believing that either, but the failure of 7/9 drives all at the same time? They would have had to put a timer in them to make that happen, and I don't trust that they could have planned it as well as it happened. If it was age related I would have expected to see more like flies dropping one at a time, all of the drives failing +/- 1 or 2 years of each other.
 
Last edited:

Thread Starter

strantor

Joined Oct 3, 2010
6,781
One other curious note: this panel is part of a larger system of conveyors. Another panel like this one plus a larger palletizer panel, both of which were commissioned at the same time, both containing identical drives, and neither of those panels suffered any losses. Not to mention all the other VFDs in all the other panels in all the other lines in all the other facilities in the area. There are at least 500 VFDs in the area that I would've been notified in the event they failed, and these were the only ones.

The more I think about it the less convinced I am that the temperature or time off was the cause. But it's still the only explanation I have.
 

Thread Starter

strantor

Joined Oct 3, 2010
6,781
In electrolytic capacitors it is common for the ESR to increase and temperature drops. Maybe the increased ESR killed your caps.

View attachment 231258
I could see the ESR being the problem if the drives were running (as in powering a motor with a load) at the time of the failure, but they weren't. They were either off, or on and just holding a charge from the rectifier. Is merely holding a charge enough action for ESR to come into the picture?
 

SamR

Joined Mar 19, 2019
5,024
Odds are greatly stacked against having that many VFDs fail and all in one cabinet out of 500... But strange things do happen. I'd still talk to someone at the distributor or A-B about it. Especially to get their take on capacitor life expectations.
 

shortbus

Joined Sep 30, 2009
10,045
Is there any type of serial or tracking numbers on the drives that failed? Compare those numbers to the ones that didn't. Sometimes components are "bad" right from the time they are made, and just decide to fail when under abnormal conditions.
 

Thread Starter

strantor

Joined Oct 3, 2010
6,781
Is there any type of serial or tracking numbers on the drives that failed? Compare those numbers to the ones that didn't. Sometimes components are "bad" right from the time they are made, and just decide to fail when under abnormal conditions.
Good call. The drives are all in the trash now but with any luck they won't have been emptied into the compactor.
 

Thread Starter

strantor

Joined Oct 3, 2010
6,781
How about temperature and humidity? ... couldn't those two factors compound the problem?
Who knows? It is very humid here but my understanding was that at temperatures below freezing not much humidity is possible. I think that 1/2" blanket of pseudo-snow was all the water that our air was holding.
 

DickCappels

Joined Aug 21, 2008
10,140
I could see the ESR being the problem if the drives were running (as in powering a motor with a load) at the time of the failure, but they weren't. They were either off, or on and just holding a charge from the rectifier. Is merely holding a charge enough action for ESR to come into the picture? emphasis added.
Not likely. ESR would only have an effect when current passes through the capacitor.
 
Top