What's new
What's new

RPC Problem - Wild leg really WILD. ( suddenly )

Zahnrad Kopf

Diamond
Joined
Apr 5, 2010
Location
Tropic of Milwaukee
15 year old shop built RPC, made with plug'n'play box** from one of the eBay sellers back then. ( ** - Buy box, connect motor here, here, and here, connect machines, start, enjoy. ) Has run well, run everything thrown at it during the last 15 years.

Until two days ago. :angry: :angry:

The most sensitive machine started throwing voltage/current errors yesterday. After some thought, I decided to check the transformer voltages. Sure as sunshine the wild leg was fairly high. ( from ground, 115V, 115, 208 ) ( same kind of behavior leg to leg ) This happens from time to time over the years due to Elec Co variances and always returns to normal within hours.

Now it's been two days.

Thinking it's a small issue with the PoCo, I altered the transformer taps to accommodate as I have in the past when I needed to deal with it right now. Worked great for a short period, and now voltage on wild leg is climbing some more again.

I am starting to think it's a RPC issue... But what would cause the wild leg to climb and climb like this? Obviously, I'm an electrical idiot, so use small words and napkin drawings.

Is this possibly the caps in the box getting old/going South? How do I test for this?

Thanks.

CORRECTIONS -

Leg to Ground - 115 W/Gnd, 209 V/Gnd, 115 U/Gnd

Leg to Leg - 232 W/U, 245 V/W, 232 U/V

Order - W, V, U on Transformer
 
Caps would be the first thing I would look at as well, knowing that there isn't much else, and they can fail progressively.

How to measure capacitance with a digital multimeter

Thanks. I thought better of my original post and updated the voltages measured. I'll climb up there and check the caps later today. Going to be a pain in the butt because the builder gooped them all the connections in RTV sillycone.

Keeping my fingers crossed that this is just a simple and quick repair...
 
Okay, well no love thus far.

I need some help with this. I'm completely ignorant here and need to learn quickly.

Run caps all appear to be the same. Start caps have four the same, one looks smaller.
Run caps all measure 60uF and in fact are (were) marked as such. Start caps vary a bit and are in sleeves marked TO-5 but I could not read the smaller one.

Machine does not like voltage variations nor more than 5% between legs, typically.

After the transformer ( headed for the machine ) I am getting -

W/U - 207
W/V - 222
V/U - 202

I have played with the transformer jumpers to no end, and this is the best I can get that way. So, I am thinking that my wild leg is simply too high. Is this correct? If so, how would I accomplish this? Add another cap in the RPC box?

Thanks for any assistance you can render.

EDIT - This thread seems to indicate the opposite of what I was thinking. THREAD LINK

Can someone verify or correct me? I may try removing one of the caps from the circuit and see what happens...
 
Well, I took a leap and a chance and went back to the shop, and disconnected one of the run caps from the wild leg.

Now getting 207, 206, and 204.8. I don't think I'm going to do any better than that... :)

Still throwing a machine error, which means I likely fried the servo drive card. :angry:

Dammit.
 
Confirms the idea that the run caps raise the voltage. They should.

Still does not explain what changed so that you don't need it any more......

Since the "balance/run" caps are load dependent, it's possible that the change is in the load machine, that it actually failed. In other words, that the RPC is fine, and the load machine failed and isn't drawing much on that line now. The high line is because that output is not getting loaded down.

That actually makes as much or more sense than the RPC "going wild" and killing the load machine. It still may have failed by high voltage, but due to some powerco issue and not the RPC's fault.
 
Confirms the idea that the run caps raise the voltage. They should.
Still does not explain what changed so that you don't need it any more......
Since the "balance/run" caps are load dependent, it's possible that the change is in the load machine, that it actually failed. In other words, that the RPC is fine, and the load machine failed and isn't drawing much on that line now. The high line is because that output is not getting loaded down.
That actually makes as much or more sense than the RPC "going wild" and killing the load machine. It still may have failed by high voltage, but due to some powerco issue and not the RPC's fault.

Hi JST,

I think it is a chicken and egg issue, but my gut agrees with your point about the RPC likely being okay. The one leg has always been a little high and I've never before understood the mechanics of the RPC components so didn't think or know to alter it.

My belief is that the PwrCo fluctuated again with the result being that the leg-to-leg went higher yet, resulting in the blown component on the servo amp. Common on this type of amp. ( Mitsubishi Meldas control and MSR type servo drive ) Last time it was a $.59 opto. I'm praying it is the same culprit again...
 
As you found out, the high leg is affected by the "balance" (PF correction) capacitors. Maybe you just leave that one out and there will be a bit less overvoltage on the "wild leg".

It would be relatively rare for an opto to fail from mains overvoltage and nothing else was bad.

Did you measure the fault on the old one when you found it?

Where was the failure?

Line-related overvoltage is typically going to stress the isolation barrier of the opto. That is not a common failure cause and is usually damaging to other components as well. The optos in industrial equipment are typically good for at least 5000V to 8000V of that "common mode" voltage (rated at 3500 to 5000V isolation). You can imagine what a transient of that amplitude could do to 5V logic circuits

A failure in the LED is not affected by mains voltage, there are power supplies and voltage regulators in between. It would be due to circuit issues, designed with higher current drive, or just plain failed from heat etc. Can be open or shorted, usually shorted.

A failure in the receiver side (usually an optotransitor) is likewise not affected by mains voltage, for the same reasons. Heat-related or lifetime related failures are the usual. Typically the output will be "stuck" either high or low.
 
As you found out, the high leg is affected by the "balance" (PF correction) capacitors. Maybe you just leave that one out and there will be a bit less overvoltage on the "wild leg".

It would be relatively rare for an opto to fail from mains overvoltage and nothing else was bad.

Did you measure the fault on the old one when you found it?

Where was the failure?

Line-related overvoltage is typically going to stress the isolation barrier of the opto. That is not a common failure cause and is usually damaging to other components as well. The optos in industrial equipment are typically good for at least 5000V to 8000V of that "common mode" voltage (rated at 3500 to 5000V isolation). You can imagine what a transient of that amplitude could do to 5V logic circuits

A failure in the LED is not affected by mains voltage, there are power supplies and voltage regulators in between. It would be due to circuit issues, designed with higher current drive, or just plain failed from heat etc. Can be open or shorted, usually shorted.

A failure in the receiver side (usually an optotransitor) is likewise not affected by mains voltage, for the same reasons. Heat-related or lifetime related failures are the usual. Typically the output will be "stuck" either high or low.

JST - I don't have any of those answers for you, but am happy to report back once I do.

The bottom line is that I am an electronic ignoramus, bordering on flat out retarded, but have an EE friend that is BRILLIANT and also tolerates my intrusions while being magnanimous enough to troubleshoot and repair the electronics that I destroy from time to time.

Yeah... that pretty much sums it up... :)
 
JST - my friend has the amp and is looking at it now. From what he has told me, he suspects a combination of heat and "the transients" from some of the relays and transformers caused by the imbalanced RPC more than actual high voltage. I did not understand what that meant, but when I asked the answer started out well enough but VERY quickly degenerated into the "WahWaaahh" of Charlie Brown's teacher so I hope that you understand what he meant more than I. ( I may or may not have gotten some of that incorrectly ) :)
 
"Transients" are the "spike voltages" which come in on the power line, or are are created locally when motors, transformers and other "inductive" devices are turned off. Inductive devices are ones that involve coils of wire.

The spark you see in a switch when the current to a motor etc is interrupted is caused by a "transient" voltage. Transient meaning short duration. The "transient" only exists until the stored energy from teh "inductive device" is used up, but the voltage can be quite high while it lasts.

Automotive ignition uses that same principle. A high voltage occurs when the points open, and the coil is a transformer that steps it up to many thousands of volts.

The primary of the ignition coil is only operating on 12 volts, but can have over 300V on it when the points open. Likewise, a 230V line might produce a transient of as much as 5000 volts or more, depending.

An unbalanced RPC is unlikely to cause any transients just from being unbalanced, it just makes slightly higher (or lower) voltage. The effect of slightly higher voltage on transients is just that, "slight".

You generally need to have switches open and close to get transients. There aren't many switches in the RPC, but perhaps when it is shut down there is a spike. And there are the ones that come in on the power line, which you just have to deal with, because they are made elsewhere.

Use of some "MOVs" (voltage limiting devices) on the input of the damaged machine should cut down the problem. Your friend ought to know how to do that. I'm surprised it does not already have them.
 








 
Back
Top