RCD on load side of VSD

Hi all,

I have an application which requires an RCD on the load side of a VSD (PowerFlex 525, 400V 3 phase, 1.5kW). I can find plenty of information about which RCD to use on the line side of a VSD, but next to nothing about using them on the load side.

Does anyone have experience with this sort of thing, or has anyone seen any documentation I can use to justify my RCD selection choice if someone comes asking?

Thanks!

Registering my interest. I would have thought a standard 3ph RCD would work here. Also, what would register leakage first; the RCD or the drive?

I have seen some old school Clipsal Type A Super-Immunised RCD's work in this arrangement, but everything I'm reading says you should use Type B. Except the accounts department, who have just seen the price of a type B and nearly had an aneurysm.

I'm not sure either about which would pick it up first either - but it's a fair bet that an RCD is going to trip quicker than a drive is going to ramp down. Whether the RCD trips quick enough to prevent the drive from actually faulting would be an interesting question.

I thought Type B was recommended for the Line side? I guess it depends on the sensitivity of the drive output to detect leakage or a ground fault. I imagine the drive would fault anyway, and not ramp down.

Well, yes, Type B is recommended for the line side. The problem I'm having is that as soon as I ask any technical support people the question, it goes something like:
"We don't recommend you use an RCD on a VSD."
"Yes, but I need to."
"Well, if you must, use a Type B."
"Can I use that on the load side?"
"Well, we suggest using it on the line side."
"Yes. But can I use it on the load side?"
"Well, we suggest using it on the line side."
"What about the load side?"
"Well, we suggest using it on the line side."
"And if I can't use it on the line side, and have to use it on the load side?"
"Well, we suggest using it on the line side."
"...is there another way I can ask this question that won't result in you giving me the same answer?"
"Well, we suggest using it on the line side."
"...are you sure you're not a chatbot?"

In any case, using an RCD on the line side of a VSD doesn't seem to me like it'd do anything at all, except protect the VSD itself from earth leakage. Once the three phases get taken into the VSD, turned into DC, and then inverted back to AC, I wouldn't think that any earth leakage would track back through the DC bus and trip the upstream RCD in any event. Certainly not at 30mA!

ASF said:

In any case, using an RCD on the line side of a VSD doesn't seem to me like it'd do anything at all, except protect the VSD itself from earth leakage.

Click to expand...

Hi, I would register my interest in this thread too.

We have used RCD's on the line side of equipment with VSD's as the equipment is portable and plugs into a three phase outlet (as per AS3000).
Our experience is that we have to disable any RFI filters to prevent nuisance tripping (which you probably already know)

However, recently one of our customers had an incident whereby the VSD supplied a socket outlet and when the appliance was plugged in a shock was received. This was on an older piece of equipment - not something that had been installed recently - and we have been asked to provide a RCD for this socket (downstream of the VSD).
We are still looking for a suitable RCD

I don't know if this is of any interest.
http://www.austmine.com.au/News/art...-and-Ampcontrol-apply-brakes-to-earth-leakage

Steve

Mmmmm - have never used one on the load side - do lots of swimming pools and we use RCDs on everything going outside the switch board - the drives are mounted along side the motors.
We use Zener drives and they have a solution that will not trip a standard 30ma AC RCD - low loss filters. No need to disable anything in the drive.
We have used standard AC RCDs on the back of Schneider MCCBs at 30ma up to a 75kW drive without issue.
That being said the motors are a different kettle of fish! Cheap Chinese motors - 6 in a line - 5 OK - sixth one trips the RCD through the drive - capacitance of the motors all over the place and that was the issue. Changed out the motor and all was well again.
By the way if you test 30ma RCDs I bet you will find they all trip at or very near too 22ma - have done the tests.

Schneider Electric have some interesting stuff on RCDs and fault paths. It is just to big to upload but here is a link:

http://eduscol.education.fr/sti/sit...es/ressources/techniques/3374/3374-ect204.pdf

or search for ect204.pdf

I had an enquiry on RCD protection of motors a while ago, and the only thing I could find at the time was this mob. Have a look/ask and see if they have what you are after.

http://www.bender.org/

ianingram said:

We have used RCD's on the line side of equipment with VSD's)...however, recently one of our customers had an incident whereby the VSD supplied a socket outlet and when the appliance was plugged in a shock was received.

Click to expand...

This is what I'm concerned about. Everyone I speak to says that a regular RCD won't detect all earth leakage off frequency inverters.

nehpets said:

I don't know if this is of any interest.
http://www.austmine.com.au/News/art...-and-Ampcontrol-apply-brakes-to-earth-leakage

Click to expand...

That says a lot on it's own. The fact that they've gone to so much trouble and expense to develop and test something like that hints toward RCD's just not being able to cut it.

BryanG said:

Schneider Electric have some interesting stuff on RCDs and fault paths. It is just to big to upload but here is a link:

http://eduscol.education.fr/sti/sit...es/ressources/techniques/3374/3374-ect204.pdf

Click to expand...

That did have some interesting stuff. It goes and contradicts everyone else by saying in relation to earth leakage faults downstream of a VSD:

The complex shape of the fault current requires the use of a type A RCD (see fig. 41 ).

Click to expand...

It does go on to say later that a type B RCD is required if the braking resistor is accessible, as only the type B will detect the DC leakage. So we're back to type B?

Jeev said:

I had an enquiry on RCD protection of motors a while ago, and the only thing I could find at the time was this mob. Have a look/ask and see if they have what you are after.

http://www.bender.org/

Click to expand...

Good tip! I have had a look and spoken to their Australian distributor. He tells me that a standard RCD will not detect all earth leakage faults on frequency inverters, and they do have a couple of products available that do. Of course, they're well over $1k, and they don't actually trip anything - you have to use the relay output of the device to shunt trip your breaker. So by the time you get the detection relay, toroidal transformer, and a shunt trip for your 10A circuit breaker (which I didn't even know you could get until just now), it's a $2,000 solution at minimum. Interestingly, the manuals for these devices show installation UPSTREAM of the VSD. The rep assures me that they will still detect faults downstream of the VSD, and also mentioned that I CAN install it on the load side if I wish, but it will only protect the motor cable and motor, not the VSD any more (which is ultimately all I care about in this case).

So, in summary:
- I should use a Type A RCD
- I should use a Type B RCD
- No RCD will work in this application
- Spend more money!

Everyone else as confused as I am?

ASF said:

So by the time you get the detection relay, toroidal transformer, and a shunt trip for your 10A circuit breaker (which I didn't even know you could get until just now), it's a $2,000 solution at minimum.

So, in summary:
- I should use a Type A RCD
- I should use a Type B RCD
- No RCD will work in this application
- Spend more money!

Everyone else as confused as I am?

Click to expand...

It probably wouldn't meet the trip time requirements of an RCD protected system either by then. If this is for an inverter to socket application, perhaps using a socket that detects the removal of the plug and isolates the supply? Either that or make a system that does that yourself.

It doesn't sound like there's a standard (Or common) industry solution for this The comment about the complex shape of the fault current makes me think it would be an electronic solution if anything.

Jeev said:

It probably wouldn't meet the trip time requirements of an RCD protected system either by then.

Click to expand...

I'm inclined to agree.

Jeev said:

If this is for an inverter to socket application, perhaps using a socket that detects the removal of the plug and isolates the supply? Either that or make a system that does that yourself.

Click to expand...

That's already in place, but it doesn't protect against earth leakage once it's plugged in.

Jeev said:

It doesn't sound like there's a standard (Or common) industry solution for this The comment about the complex shape of the fault current makes me think it would be an electronic solution if anything.

Click to expand...

I think you're right.