inverters and rcd's

[rbartlett]

I had a conversation with my old business partner -and I mentioned we had just installed a Daikin inverter that had tripped the consumer unit. He said "I bet it has a rcd"

http://www.memonline.com/rcd1.html

On gaining entry it did indeed triped the rcd. He mentioned that inverters communicate down the earth so there is an imbalance hence the trip.

I re-wired it to the other side of the consumer board (a split board) and every thing ran fine and dandy.

Now this seems to confirm the advice but I welcome any additional imput

Cheers

Richard

[rbartlett]

Yes the only part of the advice that I thought 'hmmm' was the communicating down the earth bit...


cheers

richard

[frank]

Quite an interesting thread Richard. Not something that I've come across but I can see how it could happen.

The link you posted to MEM didn't work, could you see if you can find another as I would like to see what the experts say

[chemi-cool]

hi frank,
The link is ok.

Note: an RCD must not be used as the sole means of protection against touching live parts.

Can you explain this Richard?

Chemi

[rbartlett]

chemi

yes, this basically means that if you want to protect people then you should take other sensible necessary precautions and not rely on the rcd to get you out of trouble.

therefore don't install an isolator fuseboard etc without the front cover.

don't not bother to earth components thinking that the rcd will trip before anyone gets killed. etc etc

cheers

richard

[Snowman123]

Inverters don't like RCD, curcuit breaker or fuses and they must be motor rated. Vacon drives, ABB and Eurotherm all state this in the user manual.

[frank]

The link just worked

[Birds]

Inverters will trip RCDs but not all. A 30ma rcd will be found on a lot of domestic properties and would normally be in a split board configeration protecting socket outlets for use outdoors The problem would be more apparent when an inverter was installed in an installation which was fed by an overhead supply or earthing by earth rod.The solution I have found is to change the Rcd for a PULSATING D.C FAULT CURRENT SENSITIVE TYPE. MK do them cat No 566S .This is the best solution if you are installing a small inverter and are picking up the electrical supply from the existing ring circuit. Hope this helps.

[eggs]

speaking of fuses, mcb's rcd's etc...

http://www.thefusecompany.com/

a very usefull goldmine. Never struggle to find the correct breaker again.

cheers

eggs

[Dave Goodings]

Quote The VFD's we built didn't "communicate" via earth. But by the enormous level of ambient inductively produced magnetic fields there is always going to be residual current production. If the VFD is separated from the RCD by a transformer then the inductive filtering provided by the transformer helps eliminate the nuisance trips

Marc does fitting an (Isolating Transformer) not nullify the effect of using an rcd to protect the vfd!!!

Regards
Dave

[Peter_1]

A VFD creates a supply frequency but each frequency wave is itself made up of hundreds of little waves (Pulse Width Modulation) and it's the enormously high inductive characteristics associated with the high rate of change of current of the PWM that generates residual current in metal conductors located even a meter away from the VFD.

Even a normal sinusoide we know from the main supply is made of (consists) of (square) block waves and are in fact harmonics (an algebraic algorithm)
You can have 'spikes' on a sinus of double the peak voltage which can blow up large capacitors (short circuit the plates in cosphi optimisers) and you only can 'see' these spikes with a sophisticated spectrum analyser.
A VFD is generating line differences which trips the safety switch.
Normally, a line filter on the output prevents this.

All this is a Fourier analysis
Find more on http://www.dwsimpson.com/fourieranalysis.html

We use for our data aquisition the PC logger of Picotech (UK)
They also have a PC based very fast scope with the software to convert the measured values to a spectrum analysis
http://www.picotech.com/picoscope-so....html#spectrum

[Peter_1]

Again problems with my English Marc?

http://www.process-cooling.com/CDA/A...TML&u=35322512