PowerFlex 4 Delayed Power Up Internal Issue

[jacob227]

Hello all! This is my first post!
Were I work we have 2 of the exact same machines. Each machine contains 2 PowerFlex 4's Cat# 22A-D4P0N104. They work as expected for the most part. But roughly about every 2 years. We will have one that will not power up as soon as the 480 VAC is applied.
If you disconnect the power for say, 8 hours then turn the power back on. It will take roughly 7 minutes for the troublesome PF 4 to come alive. "Digital display will come on." If you turned off power for 1 hour. It will take about 1 minute for the display to come on.
I realize it's an internal issue, but just wondering if anyone else has had the same issue. Or for the expert's, could we be doing something wrong for say like a bad parameter setting that could be causing the internal failure. Any feedback will be appreciated.

 

[OkiePC]

I have seen this before. The OEM paralleled the +24 from about 8 drives together, that connection then went through a 1746-OB16 card then back out to the drive command input terminals. After all sorts of failures and problems, I changed the scheme to eliminate dependence on the drives' on board power supplies and the problems went away. Two of the drives originally "wired goofy" were kept in the panel as emergency spares and one of them exhibited the behavior you describe. I nicknamed it the Lazarus drive. I suspect there is some failure in the VFD power supply.

 

[jacob227]

Thanks for the response! The Lazarus Drive! That's hilarious! I was curious so I cracked open my Lazarus Drive that I replaced with new. It looks like they take 2 leg's of the 480Vac, each leg has it's own step down transformer. I assume one leg steps down from 277Vac, to 24Vdc once rectified. The other probably steps down to whatever voltage they use to run the drive. From there probably some zener diode/capacitor/IC chip issue.
One thing I just thought of. The way these drives wired from OEM are a little off in my opinion. These drives are dead in the water until the machine control current circuit is true. When control current becomes true, it sends coil voltage to a starter, which sends the 480vac out to the drives. So every time a the control current drops out, so will the 480vac to the drives which shuts off the drives once they bleed out. This happens quite often daily. I wonder if the constant cycling off and on of the drives could be causing them to fail every other year. What do you think?

 

[DickDV]

Switching input power to a VFD is extremely bad practice and always leads to high drive failure rates. Powering up is one of the most stressful things a VFD does. VFD's are designed to stay powered with on/off being controlled thru the control inputs.

 

[ASF]

This is why safe stop/safe torque off VFD's are becomign more and more common

 

[OkiePC]

I agree. I have installed contactors between the VFD and motor successfully a number of times with a number of different brands. As long as you turn off the drive enable contact (and for good measure break that circuit with the contactor), they usually don't even fault, and the drive hardware "likes it better".

 

[jraef]

...
One thing I just thought of. The way these drives wired from OEM are a little off in my opinion. These drives are dead in the water until the machine control current circuit is true. When control current becomes true, it sends coil voltage to a starter, which sends the 480vac out to the drives. So every time a the control current drops out, so will the 480vac to the drives which shuts off the drives once they bleed out. This happens quite often daily. I wonder if the constant cycling off and on of the drives could be causing them to fail every other year. What do you think?

Absolutely a problem. To expand on what the others said, a VFD has capacitors on the bus and capacitors charge instantly by pulling all available current in a system, as in the Available Fault Current, albeit for only a fraction of a second. Still, that could be enough to damage the components. So VFDs must have what is called a "pre-charge" circuit in the power stream to deal with this every time the drive is energized from the line. There are several ways to do this and on small drives it usually involves a current limiting resistor that is in series with the diode bridge on the DC side, then once the caps are charged, the resistor is shorted out of the circuit. But EVERY time you first apply power to a VFD, you are stressing that resistor and whatever shorts it out (typically a small relay contact). The more often you do that, the sooner it fails.

Worse yet, smaller "Component Class" drives like the PF4 are not even going to use the relay to short it out, they use what's called an NTC resistor. NTC means Negative Temperature Coefficient of resistance, which means the hotter the resistor, the lower the resistance. So when you first energize it, the resistance is high and the resistor acts to limit current, but it quickly heats up and essentially takes itself out of the circuit. When you cycle that type of circuit on and off a lot, the failure probability increases dramatically. This type of circuit is less reliable, but saves on the real estate, allowing the drive to be slightly smaller, which is what OEMs want. They just need the drives to outlast the warranty, which they do, but it is the end user who suffers the long term cost of that.

So yes, bottom line, bad design. The PF4 is a low cost no-frills drive, no features like Safe Torque Off (STO) or even options for it. So the OEM was taking the cheapest way out of making the system safe by dropping line power, but in doing so they sacrificed the long term survivability of the components. At this point though, you would need to redesign the entire circuit and safety system if you change anything. If you want to take that on, I suggest going with a PF525 which has the STO feature built-in. Leave the drives powered up when the entire machine is powered, but have the machine control current energize a safety relay, which then enables the STO input of the drives. So if the control power is off, the drives are disabled and unable to have any output energy going to the motors in a manner that has been tested and verified by safety agencies (which is what STO does).

 

[jacob227]

Sorry for the delayed response. "I only work weekends." Thanks for the detailed response. Your explanation backs up my theory. I'll have to talk to Engineering to see what they want to do about this issue.