failed circuit board...no apparent cause

[wirenut1980]

I have a customer with a laser cutting machine from Germany, who after a momentary outage on the 69 kV transmission system claimed they had a 24 volt circuit board damaged. The customer is fed off a 12.47 kV distribution circuit fed by the transmission system, their service is 480 V ungrounded delta. The machine in question is fed from a 480 V delta to 480Y/277 V transformer. Grounding appears to be correct based on visual inspection only. The machine has several transformers within it to feed PLC's, relays, and circuit boards. No other boards in the same machine were damaged, no boards in other machines in the shop were damaged. This board's power supply is connected to the 480 V bus (480 V input, 24 V output). I thought I had read somewhere that power supplies fed from 480 V were more susceptible to damage than if they were fed from 120 V source. Any reliable information to back that up out there? Otherwise, I cannot come up with a good explanation as to why the board failed. Maybe a bad component that failed due to the stress of the quick off-on?

Thanks for any insight!

 

[Smart $]

Otherwise, I cannot come up with a good explanation as to why the board failed. Maybe a bad component that failed due to the stress of the quick off-on?

Exactly, inadequate surge protection.

 

[petersonra]

based solely on my very limited anecdotal experience, its my impression that 480VAC input power supplies seem to have a higher rate of failure than 120/240V input units, and that they seem to throw off nastier waveforms.

 

[charlietuna]

I seriously question that this type of equipment's 24 volt control power is transformed directly off the 480 volt supply to the machine. If it is that critical the manufacturer should have buffered via a 480/120 volt transformer feeding a 120 volt ups power conditioner. Might have increased the cost of the machine by $200.00 ??????????????

 

[LarryFine]

It may be that the spikes were generated by the machine itself; when magnetic fields collapse, the back-EMF can be many times the steady voltage.

It sounds like the single board damage was a lucky break.

 

[charlietuna]

I think most of these problems are caused by elevated voltages instantaniously applied to the ground effecting electronic components before they can bleed off to a zero state??? We have isolated sensitive controls using small UPS units to prevent this type of damage.....

 

[wirenut1980]

I was wondering if a UPS with surge suppression in front of circuit boards would help. Or just installing a 480-120 V transformer might do the trick, or maybe both. They say it cost about $10K to replace the board, so the economic justification should not be too hard.

 

[iaov]

Unless you can look at fried components on a board it may be almost impossible to say why it died. It is possible it had nothing to do with losing power, although in this case it seems likely. Expensive board may very well justify an expensive UPS system.

 

[charlietuna]

Also understand that in Germany their operating voltages are different than ours along with their phase colors... We have solved many control problems by installing small inexpensive UPS units in equipment experiencing board failure problems...

 

[hurk27]

Without knowing what is also connected to this board, this would be a very loose guess, but if this board has connections that has hard connection to other boards that are supplied buy other sources, I.e.. transformers then there could have been a back fed current through a common signal ground/bond between the boards/sources this can happen when the voltages decay from the power loss, some of the capacitors or another board don't decay as fast and can back feed through the signal I/O's... poor design if this is the case, but I have seen it when multiple powersources (transformers/power supplies) are used with equipment sharing a common signal ground between the I/O's.

In this case, isolation would be the key, or a common supply for all the boards.

(hard connection) = non isolated type connection.

 

[charlietuna]

The Green color code in Germany is a hot -- I believe!!

 

[wirenut1980]

Without knowing what is also connected to this board, this would be a very loose guess, but if this board has connections that has hard connection to other boards that are supplied buy other sources, I.e.. transformers then there could have been a back fed current through a common signal ground/bond between the boards/sources this can happen when the voltages decay from the power loss, some of the capacitors or another board don't decay as fast and can back feed through the signal I/O's... poor design if this is the case, but I have seen it when multiple powersources (transformers/power supplies) are used with equipment sharing a common signal ground between the I/O's.

In this case, isolation would be the key, or a common supply for all the boards.

(hard connection) = non isolated type connection.

This is entirely possible. The machine has at least two transformers inside plus a couple power supplies. I won't be going in that far to check, as I am just a simple utility employee. I will keep everyone posted if a cause is identified.

Out of curiosity, how would one isolate the common signal grounds?

 

[robbietan]

Exactly, inadequate surge protection.

ac to dc conversion boards acts like motors where they draw more current when the voltage supply is lower. during a voltage dip (sag in our area), these power supply boards tend to overdraw current just like an induction motor. with slow acting fuses or breakers, you get a burnt board