Inverter output

[dor]

Following some work on my house wiring, I bought a socket tester (Kewcheck 103) which passed all my house sockets.  So thought I would test the boat sockets and it comes up with "Earth fault".

My inverter is a Powermaster 1800 PSW and has given faultless service for many years.  I suspected the earth fault might be because the AC outlet is centre-tapped.

Checking the pins I get:

L-N    227V

L-E   145V

N-E   96V

So, firstly do the outlet voltages look right?  Secondly, would a centre-tapped output give rise to an apparent 'earth fault'?

The output from the inverter goes straight to a RCD on a distribution board which trips normally when the test button is pressed.

[alan_fincher]

I'm sure others will be along that know more than me, but if it were due to a centre-tapped output would you not expect L to E and N to E voltages to be a lot more equal than they are?

 

If you deliberately induce a leakage to earth, (I do this with a mains plug that has, from memory, a 5.6 K ohms resistor wired between L & E), does your RCD trip?

[WotEver]

12 minutes ago, alan_fincher said:

If you deliberately induce a leakage to earth, (I do this with a mains plug that has, from memory, a 5.6 K ohms resistor wired between L & E), does your RCD trip?

That value won’t work on a CT output as the current will be too low. You want >30mA, so working on 40mA @ 115V you get 115/0.04 = 2k8.  Using a preferred value gives 2k7. It’ll dissipate 4.6W (but only for a very short time, if the RCD works), so ideally a 5W 2k7 resistor which must only be used on 115V supplies. 

18 minutes ago, alan_fincher said:

I'm sure others will be along that know more than me, but if it were due to a centre-tapped output would you not expect L to E and N to E voltages to be a lot more equal than they are?

Yes, I’d expect it to. The following, which was cribbed from a Gibbo post some time ago might help:

 

Inverter bonding

 

There are three basic circuits generally used for manufacture of small inverters of this type.

 

1. Totally isolated output. There is no connection whatsoever between the AC output and the DC input and/or earth. This type is perfectly fine to have neutral and earth bonded.

 

2. Partially isolated. The AC output is isolated from the DC input but the AC output is "sort of" centre tapped round the output ground. This type cannot have the neutral and earth bonded.

 

3. Mickey mouse isolated. There isn't full isolation between the DC side and the AC side. This type cannot have the neutral and earth bonded.

 

Now to identify which type you have................

 

Firstly, using a multimeter, ensure that the output ground is bonded to the case. If it isn't, take it back to the shop for a refund. End of.

 

ONLY do this if you know what you are doing with mains electricity. You could kill yourself if you get it wrong.

 

Get a good old-fashioned 15 to 40 watt 230 volt incandescent (not LED) bulb. Connect it between output ground and one of the AC outputs. Then connect it between output ground and the other AC output. It should not light up on either of them. If it does not light up then you have 1. above. It is fine to bond neutral and earth.

 

If it does light up fully on one of them but not the other then it already has neutral/earth bonding. Check it has the correct one bonded. Live should light it up, neutral should not.

 

If it lights up on both of them you either have 2. or 3.

 

If it's 2. then you cannot bond neutral and earth but you don't need to. The centre tapping will allow an RCD to operate properly and protect you in the same way it usually does.

 

If you have 3. then it should only ever be used with double insulated equipment. They are not suitable for any other type of equipment. They are dangerous if used with single insulated equipment.

 

How to identify between 2. and 3. ?

 

Connect the same lightbulb between each of the outputs and the DC negative input. If it lights up even partially on either of them then it is 3. - Do not use it for anything that isn't double insulated (ie only has a 2 conductor mains lead).

 

Finally, for inverters of type 2, using a multimeter, ensure there is no continuity (it must be greater than 1MOhm) between the output ground wire and either of the DC input wires. If there is continuity between either of them the inverter should only be used with double insulated equipment.

 

If you do anything wrong in the above, or do not do exactly what is written above you could conceivably:-

 

1. Kill yourself.

2. Kill someone else.

3. Blow the inverter up.

4. Identify the wrong type of inverter

[alan_fincher]

1 minute ago, WotEver said:

That value won’t work on a CT output as the current will be too low. You want >30mA, so working on 40mA @ 115V you get 115/0.04 = 2k8.  Using a preferred value gives 2k7. It’ll dissipate 4.6W (but only for a very short time, if the RCD works), so ideally a 5W 2k7 resistor which must only be used on 115V supplies. 

Good point!

Why do you think there is such a large discrepancy between L-E and L-N voltages?

[WotEver]

31 minutes ago, dor said:

would a centre-tapped output give rise to an apparent 'earth fault'?

Yes

3 minutes ago, alan_fincher said:

Why do you think there is such a large discrepancy between L-E and L-N voltages?

Cos I suspect that it’s what Gibbo calls a Mickey Mouse inverter. The above tests should help prove it one way or another. 

[dor]

Thanks for your replies so far.  I had thought that Powermaster inverters, although no longer marketed in this country, were considered good pieces of kit - I bought it in 2007.  Maybe "good" then is now "Mickey Mouse"!

 

I'll have a go at making the other measurements as mentioned in Gibbo's piece.  I'm pretty certain that all the equipment on the boat is double-insulated but I will check.