Wiring of a Non-Isolation Transformer.

[ChristianFletcher]

[img]http://i.**********/o5mPbwd.png[/img]


Hello everyone. I am looking for some advice on resolving a problem with some equipment I have purchased recently. This is a single phase milling machine that has a variable frequency drive. The equipment is supplied with a 13 amp UK plug. The machine is located in the garage and is fed from a fuse spur from the kitchen sockets ring main. My installation is type TN-C-S and my earth leakage device is the normal domestic 30mA.

The problem that I am having is that when the main contactor in the control panel pulls in, my RCD trips in my consumer unit. I am quite experienced with VFDs and used to use them all the time when I was designing machines. I thought the most likely problem is the filter capacitors in the VFD are leaky enough to trip out the breakers and this is a known problem.

The machine has a number of output circuit that are energised by the main contactor

1) Cabinet cooling fan
2) Power feed circuit
3) Head raise and lowering circuit
4) VFD for main spindle motor
5) Forced are fan for main spindle motor

I have disconnected each of these circuit and run just one of each at a time and found that the VFD (4) and the forced air cooling motor (5) when brought into circuit trip the RCD in the consumer unit.

Strangely sometimes the fan trips out RCD when the main contactor pulls in and sometime the main contactor pulls out.

I have done an insulation resistance test on the fan motor and is clear to over 1 gig Ohm. But it trips the main RCD almost every time.

If I run the machine via an isolation transformer every works normally and no trips.

Its difficult to fault find because when power is applied it trips before I can measure anything. I think this is a transient issue rather than an actual ground fault. Its strange that I can run my old milling machine and welding sets and basically anything and don’t really have an issue with nuisance tripping but I’m very suspicious. I do occasionally get a nuisance trips with one plug socket in my house that has various computers and power supplies hanging off it but this is only once every few months.

I have just ordered a sensitive clamp meter to try to identify any genuine leakage into the earth. The next step is to try and run the machine up slowly using a variac and measure the leakage, but the machine has to be powered for the main contactor to be pulled in.

I actually think the isolation transformer is acting as a choke rather than just removing the refence to ground and this is supressing the transient. The machine works fine working from an isolation transformer but is not good for the VFD ground to be floating. So I want to link the secondary of the transformer to ground. Do I have any issues doing this

I have also posted a thread on EEV but i think people here have more electrical experience etc.

http://www.eevblog.com/forum/chat/va...n-my-house/25/

Thanks for any advice recieved in good faith etc.

Regards Chris

[vinrads]

Hi Chris when I built the workshop I fed it with a direct feed via a thirty amp fuse to a separate consumer unit in the workshop , so as not to trip all the house power . you said that it trips sometimes when using a socket in the house . It is possible that the trip has become to sensitive. Mick.

[ChristianFletcher]

Thanks Mick

Yes your right that the RCD may have become sensitive. I have just bought a fluke installation tester so i will be able to test this fully.

Its not easy for me to modify the actual house wiring as its a old stone house with stone floors and walls and damage to the house in installing the new cable would not be worth the effort. The garage is also about 30m away across a concrete driveway. Even a 16 Amp supply would be great so I could run my plasma cutter at full wack or have a bigger HP compressor. I dream one day of buy a farm and having three phase. A man can dream etc.

My main question is my transformer solution safe ?

Chris

[GMB]

I think a possible cause of this kind of problem is where there is a powerful inductive thing like a transformer or motor which is causing a huge but short inrush current. This combines with emc capacitors to trip RCDs by the surge generating an instantaneous rise in the voltage on the neutral, which results in a pulse of current flowing from N to E via emc capacitors and thus unbalancing the current the RCD is monitoring.

[MotorBikeLes]

I would guess there is nothing "wrong" with any of your stuff, just the way the components interact at switch on.
I have an old stone cottage fitted with a 100ma RCD. I redirected the original "cooker" supply to an armoured cable, and underground to my shed. There I have a second consumer unit preceded by a 35ma RCD.
I developed a rather novel single to three phase rotary converter which involves two stages of operation, and with various capacitors incorporated. As originally envisaged, there were sections in the circuit with permanent connection to the live and neutral conductors (after pressing "start" of course. Once switched on, it worked, but periodically it would trip both the shed and cottage RCDs. No pattern to the tripping, and nothing whatsoever that admitted errors to my 500v megger.
After much hand wringing, I added relays to ensure that there were no post switch-on permanent connections. Problem solved!
Later, it occurred to me that as I had a 415v 3 phase (in and out) inverter, I could fit this between my rotary converter and the working load (Harrison lathe). The inverter came to me bolted onto a 3 phase filter block, but for convenience I connected it directly. This was both successful and very worthwhile, so I decided to install it correctly onto the 3 phase input filter. Instant tripping of both trips. Remove and check the filter with the megger, but again, everything tested OK, but it is "full of capacitors", so I discarded it.
Later I bought another more modern and cheaper inverter, which also arrived with a filter unit. I meggered the filter unit, all OK, so I took a chance, but exactly the same result. Remove it, the inverter itself was of course perfectly OK.
No, no component faults, just circuit design, which you are often "stuck with".
Les.

[Station X]

Is it possible that something else connected to the supply has earth leakage and the VFD is the last straw which breaks the camel's back?

The one VFD I repaired would trip the MCB (not RCD) when switched on. This was down to in-rush current into the capacitors. Once the MCB had been reset a second attempt at switch on was successful, as the caps were now partially charged. This was cured by fitting a 'C' curve MCB

[ChristianFletcher]

Hello everyone many thanks for the comments.

Yes it’s quite possible that the RCD is on the edge due to leakage from other circuits. I did however switch all the MCBs off in the house apart from the one running the machine. This is obviously not totally conclusive but it still tripped just the same.

I don’t really get a lot of nuisance tripping and often plug in welders and lathes and even the odd dubious looking valve radio and don’t get many problems.

So as it happens my new isolation transformer arrived to tonight and i did link the secondly neutral to the earth and the machines running happily now. I’m just waiting for my new clamp meter to arrive so I can now measure if there is any real leakage.

When the fluke installation tester arrives this will test the RCD Tripping current and time to trip. This should give me a positive idea regarding the problem.

My gut feeling is it’s a transient issue. I have read the double pole switching can also cause issues if the neutral disconnects before the live. Apparently this is one of the main reasons for RCDs Tripping when the equipment is turned off.

I did also speak to one of the big transformer manufacturers in the UK today and their technically guy was really good. Apparently it’s quite common to connect the neutral of the transformer to earth on the secondary side. It’s often done for Audio application an the example I found was for boat shore supplies. I guess it’s nothing special as earth and neutral is often the same thing at the utility transformer end and it’s the same for my TN-C-S but it really feels unnerving when you link them.

I have also ordered a 100ma RCB for the output of the transformer belt and braces etc.

Thanks I will post again if I ever get to the bottom of all this.

Best regards Chris

[Trevor]

Capacitive leakage you get enough leakage to trip the rcd but on dc the megger you will get a very good reading i had this with some series connected capacitive corrected light fittings in the past
Trev

[Ed_Dinning]

Hi Gents, Trevor has pointed out one of the drawbacks of RCD's. They sense current and do not discriminate between inphase and reactive components.
30mA at 240v is a reactance of 8000 ohms. At 50Hz this is a capacitance of about 0.5uF.
Even if the capacitor has a DC resistance of 10G ohm, it will operated the RCD.

This can cause nuisance tripping where there are many IT equipment's on the same protected circuit, each with their own class Y caps fitted.
It can be worse where there are motors and similar electromechanical parts fitted in which there is a higher capacitance to earth.
If the equipment is fed from an inverter supply, which has higher frequency harmonics this will increase the leakage current and RCD's are not frequency sensitive.

It should be possible to measure capacitance from supply to earth on the equipment side; anything near 0.5uF on a 50 Hz supply should raise suspicions, much lower values can give problems on inverter drives depending in the frequency spectrum of the drive.

Ed

[ChristianFletcher]

Thanks for the comments. I downloaded the manual for the VFD and it mentions the main RCD should be rated for 300ma but locally at 30mA. It’s concerning that this is a new machine supplied with a 13 amp plug. I must assume I’m not going to be the only person having a problem using one of these at home. I still think it’s all a bit strange note I did say the RCD trips with the VFD disconnect. The spindle fan motor also trips the RCD. The transformer has got the machine working I guess I need to decide if I’m to keep digging.

[McMurdo]

Repairing VFDs is a daily business for me in the main job and it's almost impossible to power one of any size from a 30mA RCD-fed socket, single or 3 phase, at least with the earth connected. For soak testing purposes I have a bank of non-RCD protected sockets for powering VFDs and to enable me to safely connect a laptop for diagnostic purposes etc.

VFDs are normally used on industrial supplies without RCD protection and have heavy Y-class filtering to earth, for RFI reasons. Synthesising a 4kW 3-phase, high frequency pwm waveform into an induction motor from a single phase supply can lead to lots of stray noise which is not allowed these days.

Some VFDs have an internal jumper to disconnect the RFI filtering from earth, useful in rcd-fed cases.

[philthespark]

Hi, if you have an rcd tester then do the following,
1 Test the tripping current of the rcd with everything else disconnected, the best way to do this is to remove the leads from the outgoing ways of the rcd, it will probably trip at quite a bit less than the rated trip current, usually I fing they trip at anything between 22 and 26 milliamps.
2 with everything reconnected, do a ramp test, this will reveal a lot about what is going on in the electrical system, existing leakage current etc, it's surprising how certain pieces of kit have a fairly high leakage.
Finally remember that an rcd will trip on leakage from neutral to earth as well as live to earth, so dropping a circuit breaker in an attempt to isolate a circuit from the rcd won't work as most standard breakers found in a domestic environment are single pole.
If you do go down the route of feeding it via a transformer, be very careful about linking the neutral and earth on the TX (transformer), while a TNCS system does have the neutral and earth linked at the substation, current uk wiring regs prohibit the linking of neutral and earth at any point within the customers premises. If you suspect your VFD is faulty then I may have a spare one out of a lift in my shed (the vfd not the lift, lol) drop me a pm and I'll try and help you.

[ChristianFletcher]

Thanks for the advise on the VFD. There is a link that can be cut to take out the filter capacitors as I spoke with drive company (Delta) they say cutting this is common for domestic applications.

Note I did not buy this drive it came installed in a brand new machine I have purchased. The machine comes from a well known machine supplier in the UK. I tried to get some support from them but they were useless. They just said the machine works fine in our workshop which was disapointing. I have a lot of experience with VFD mainly Siemens Vector and ABB. But these are large three phase and not RCD protected in my experience so not had this problem before.

Im going to go with my isolation transformer for now with the earth neutal link. This is simply a plug in box so not part of the installation. If it goes wrong im the only person put at risk by the modification.

Im still waiting for my new test gear to arrive and do intend to try and measure if the machine has a genuine earth fault or is just normal leakage from the EFI capacitors. Given that i can now actually just power up the machine thats helping the fault finding process.

Thanks for the advise on testing the RCB. I had assumed i could just turn off all but one curcuit and plug into that for testing but you make a good point about the neutral earth fualt I had not considered.

Regards Chris

[Refugee]

From the first sentence of the previous post we have got the answer.
It is either a non RCD socket or remove the link.

[ChristianFletcher]

The clamp meter arrived today so I was able to measure the earth leakage. This was around 14ma. There’s quite a few circuits within the machine. But I measures about 14ma at the mains lead. With the clamp across the Vfd supply I get around 10ma. This actually drops when the VFD is running the spindle motor. I have a few mA unaccounted for but there’s a bit of variation depending on how the clamp is positioned and adjacent metalwork.

So I think this is going to give problems on a domestic supply without some modifications by the user.

Thanks for the comments and advice.

Regards Chris

[trobbins]

Originally Posted by ChristianFletcher

My gut feeling is it’s a transient issue. I have read the double pole switching can also cause issues if the neutral disconnects before the live. Apparently this is one of the main reasons for RCDs Tripping when the equipment is turned off.

Is the main contactor for the machine switching active and neutral? If so, then you could link out the neutral contact (if all appears safe to do that) as a fault-finding step.

[ChristianFletcher]

Yes I did consider that it would be easy to link out neutral switch on the terminal rail in the main panel. But I think I’m happy at this point to just stop digging. It’s likely two issues the normal leakage from the VFD and the double pole switching.

Thanks regards Chris