Powerflex 525 dies every 18-24 months

[lmackey91]

First time here
I have a cabinet with a PowerFlex 525 that starts and stops about 60-80 times per minute, it is one of 10 in the cabinet and the only one that has needed replaced in the 8 years I have worked on it.

The problem is when it gets the signal to start, it does nothing for about 5-10 seconds and then return to normal for a while, it will do this about 2-3 times a day at first and after about another month of dealing with, it will get as bad as once every 5-10 minutes and once I replace it everything is good for another year or so.

It's a 3 Phase PowerFlex 525 2HP, Ramp to stop:0.1s, external brake on motor, no resistor brake, start source ethernet, no fault codes.

Tried swapping the start source to the terminal block with no success, old drive will bench test for hours without fail.

Just wondering if anyone has had a similar application and problem and has seen one of these drives slowly die like this.

I don't know enough technical details of these drives as far as stopping/braking, could it be stopping too fast and getting power fed back into the drive and stressing/damaging the internals, I've thought about either installing a resistor brake or maybe switching it to coast to stop so the mechanical brake does the stopping power.

I just replace the drive 3 hours ago for the 4th time and haven't had any issues since. just like every time I've changed it before.

[jstolaruk]

I've "fixed" VFD or servo weird behaviors with line reactors. So, I now routinely design them in or at least leave panel space for them in case I run into weird problems at commissioning time.

[parky]

Different application but we had a system that had 8 525's on them, these failed at an alarming rate compared to other drives on other systems, although there was no braking, the 3 that failed the most were ones that started & stopped at about 3 second intervals, one was also reversing i.e. toggling between forward & reverse with a delay to allow natural stop.
due to the number of drives failing at one point we had run out of spares, as a temporary measure a Mitsubishi drive was fitted, as usual, if it don't break leave it, this turned out to be reliable never had one go faulty, eventually replaced all of them, appart from some idiot spraying water in the panel never replaced one again.
We have no idea if making changes i.e. filters and such would have aliviated the problem or if it was a bad batch of drives so can only assume that other makes of drives are either better quality or will put up with other influences better in hard working conditions. We did have a system of replacing 525's with a number of other makes & I can honestly say we had fewer problems, however, even before this our policy was to try & standardise on fewer makes
like Mitsubishi and Danfoss, the latter being on the surface the most reliable.

[lmackey91]

Forgot to mention that these do have line reactors.

ill have to bring up the idea of trying a few Mitsubishi drives, over the past few years we have swapped a bunch of PowerFlex 4 and 40s for 525s and they have seemed to be a bit more troublesome. Might end up swapping it for a used 40 next time and be better off.

thanks for the input.

[parky]

The ones we replaced were originally installed in early this millenium, we put up with having to replace them like for like until about 5 years later, we also had problems with one part of a two part machine regarding the HMI, this was a PV600+ I think, again had frequent problems of them dying, the cost of repair was edging on the cost of a new one with our discount so our engineering buyer would rather scrap them rather send them back for repair. interestingly the two machines worked together one in a low risk manufacturing area feeding the other in a high risk area, these were manufactured by companies in two different countries, in the hi risk machine the VFD's were Mitsubishi, the HMI was also a PV, but this one never failed, turns out the one on the low risk side when an E-stop was pressed it dropped the power off everything, we modified it so the HMI & PLC (appart from outputs) was left on, it certainly reduced the amount of HMI's replaced.
So, again, the same HMI one powered down more frequently due to e-stop & one rarely powered down made us come to the conclusion that RW equipment at that time had issues either with component quality or design.
Nothing against RW equipment but we had to start standardising & the results seem to favour other manufacturers due to number of units already in place, ease of procuring the equipment.

[Ken Roach]

This might not be a physical or electrical failure of the drive.

Examine your start/stop logic very closely to see if the STOP bit is set false in the same rung as the START bit is set true, or in closely adjacent rungs.

Because of the way asynchronous I/O works in ControlLogix, if the EtherNet/IP connection RPI timer expires in between the evaluation of those instructions, the STOP and START bit will be true in the same Output assembly packet transmitted to the PowerFlex drive.

Because A-B drives require the STOP command to be false at the time that the START command goes from false to true, the result is that the drive appears to "ignore the Start command", but there will be no errors or faults indicated.

Systems that have such a program logic defect but start and stop infrequently might never see such an occurrence. Systems that start and stop very frequently, like yours, will have that failure by coincidence more frequently.

I can't explain the exact circumstances of why this occurs on your system: there may be intermittent network issues that introduce some lost packets and timing differences, or there may be elements of the control program that are not being executed all the time that change the timing of the drive control logic relative to the comms subsystem.

The standard way to prevent this in logic is to hold the STOP bit true only until the drive indicates with its status bits that it has stopped (or at least that it is decelerating).

This is very difficult to get a confirming diagnostic on (I've done it, but I have network skills and tools), but I have definitely seen it on several occasions with high-speed packaging machines. It's not unique to PowerFlex 525 or EtherNet/IP: the first time I saw it was with PowerFlex 40P drives on ControlNet.