VFD Overvoltage Trip - what exactly happens?

[SCOC]

Greetings,

I am wondering what exactly happens in an IGBT driven VFD drive when it trips due to overvoltage. I presume that in all cases there is voltage detection on the DC bus. Once it goes above a certain threshold, the drive "trips". What exactly does the drive do to protect from the overvoltage? It stops the IGBT switching obviously, but does anything get electrically disconnected or switched off to prevent damage from the overvoltage condition?

 

[jraef]

It depends.

If the OV is occurring as a result of regeneration of a spinning load, the drive will turn off the IGBT firing so as to prevent that regen power from happening. But if the OV condition is the result of a high incoming line on a standard 6 pulse inverter, there is nothing the drive can do about it, there is no "off" with regard to the line power. If however it is an AFE (Active Front End) drive, aka a "Line Regen" drive, some of them have full control of the incoming AC line power and can turn off.

"Will work for (the memory of) salami"

 

[Skogsgurra]

And, of course, if there is a contactor or breaker that is controlled by the VFD itself, it is common that it opens. Not that it helps much (too late) but still.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[Marke]

Or if it has a braking resistor fitted, that will be dumped across the DC bus to try to pull the bus down.

Mark Empson
Advanced Motor Control Ltd

 

[SCOC]

Thanks for the replies.

So in the case of an AC line surge, the only real protection for the IGBTs a drive has is its DC link inductor/caps acting as a low pass filter, and external protection like SPDs and line reactors. Of course there may be MOVs internal to the drive too. And as Mark mentioned, a brake resistor will discharge some energy off the bus.

I've done some research on transient voltages ocurring from fault clearing of other equipment on the bus (<1000V), and several thousand volts can be generated for relatively short periods of time (several microseconds), but a substantial surge of 1000V or more could exist for a few mS after large fault clearing which is a lot of energy to deal with, especially on a low impedance bus. How often have people experienced IGBT drive failure after surge voltages, resulting from fault clearing or other, and what do you do the protect against it? My thoughts are to use an SPD device on the supply bus, and a line reactor to help protect the drive. Any other techniques used, internal or external to the drive?

 

[djs]

In my experience the most common reason for a VFD to trip from over voltage is when the incoming nominal voltage is at or greater than the nominal rated voltage of the VFD. For example a VFD that is listed as 460 VAC is connected to a supply that is nominally 480 VAC but is actually 490 or 500 volts. It will work, but usually during low load periods the drive will trip.

 

[mikekilroy]

I've done some research on transient voltages ocurring from fault clearing of other equipment on the bus (<1000V), and several thousand volts can be generated for relatively short periods of time (several microseconds), but a substantial surge of 1000V or more could exist for a few mS after large fault clearing which is a lot of energy to deal with, especially on a low impedance bus. How often have people experienced IGBT drive failure after surge voltages, resulting from fault clearing or other, and what do you do the protect against it? My thoughts are to use an SPD device on the supply bus, and a line reactor to help protect the drive. Any other techniques used, internal or external to the drive?

I think if you continue your research you will find that it is not the output IGBTs that typically fai from these input voltage spikes but rather the input rectifier diodes.

In addition, if you are in a location where you cannot stop damaging hi voltage input spikes on your 460v lines, then you can consider 2 other things: 1) make sure you have a wye power source: if delta, then you are much more prone to these spikes since there is no low Z path to hold each leg down, 2) you can install input hi joule TVS devices to help take the spikes out at the input of the machine: we have some aerospace customers in GA & KS who have problems with lightening strikes damaging their drives on regular basis. Although we liked the monthly repeat repair business, we developed a 460v 3ph 50,000 joule TVS device we sold them for $ 500 each to put at input distribution block of each machine that has nearly totally eliminated the damage.

 

[John2025]

We lost a VFD drive for a compressor once from probable ligtning surge. No line reactors. How much, if any, would a line reactor have helped for those kinds of surges? Someday we'll get around to surge protection, for now we shut off the disconnect when the unit's off. Cuts the odds in half since we run 12/day.