another soft start quandry

[wirenut1980]

Hello, I would like to resurrect this thread, because I am seeing similar problems with a solid state soft start on a 500 hp compressor for a freezer.

The soft start limits the starting current for the first 5 seconds to approximately 750 amps to 900 amps, but then jumps to approximately across the line starting current for about 0.5 seconds before settling down to running current, which is well below full load amps. Attached is a cycle by cycle snapshot of the RMS voltage and current during the start, as well as a couple waveform captures. I suspect maybe the motor is not up to at least 80% speed as Jraef previously noted, but has anyone come across a standard that talks about this?

The customer is not having a problem with overloads, but the customer is intermittently blowing main 3000 amps fuses that are feeding about 6 of these compressors. The individual fuses feeding each compressor are not blowing. There is the possibility of more than one compressor starting near the same time or at the same time...i.e. no interlocks.

No offense because this site is a great resource for me, but I probably need to go to the customer with something better than "I read it on Mike Holt's site.":smile:

Thanks!

 

[gar]

080506-1435 EST

wirenut1980:

I can only offer questions and generalized comments.

Is the scale on the left side of the first plot voltage? Yes is my guess.
Is this the input voltage to the soft start control? Again my guess is yes.

It appears I have to make a rough estimate on the current assuming 900 A for the 5 seconds and that makes the peak about 3000+. What is the peak and how does it vary from one start to another?

Do you have a way to measure RPM?

Can you adjust the the 5 second period?

Can you measure the motor voltage, and how does it vary?

Are the main fuses fast blow, or what?

What is the approximate period from one turn on to the next?

If you look at a probability distribution curve of the period between turn ons there will a mean value with some variance about the mean. Then as ambient temperature varies or some other factor the mean and variance will both change. The three pumps are independent to some extent from one another but their turn times are not totally independent. You should expect to see these turn on times drift relative to each other and it is not too unlikely to get coincidence of turn of two or more.

You might find it useful to monitor the current to all 6 and see if this helps to identify the likelyhood of coincidence.

Can you lengthen the slow start duration without hurting the motor to see if you get to a higher speed? It could be you need more slow start voltage. Can you ramp up the slow start voltage?

With your current recorder how long can you run, and is that duration at all long enough to capture a fuse blowing? If long enough, then monitoring the total load would allow you to see the event and maybe identify the cause.

.

 

[GeorgeB]

The soft start limits the starting current for the first 5 seconds to approximately 750 amps to 900 amps, but then jumps to approximately across the line starting current for about 0.5 seconds before settling down to running current, which is well below full load amps.

It appears that you are still accelerating when you go to "full speed" or on-line as the current has not dropped towards running levels. I believe the 5 seconds is too little time to get this rotor to the speed where transition from soft start occurs.

IMO, you either need to give a little more starting current or a little more time. I'd expect to see the 850A/750A (by eye) value drop to perhaps 400A, maybe lower, before transition.

The few soft start with bypass contactor systems I've personally been involved with actually reached "full" speed before the bypass contactor closed. It that happened, you would see current at unloaded run levels, about 200A on your plot.