The Megger DCM330

[Cursitor Doom]

I bought one of these the other day:

I didn't have anything to measure high currents and do non-contact tests for voltage (that was the excuse I made to myself for buying it, anyway).

I've noticed that on the no-contact checks, the thing indicates a positive on some cables even if the switch to the (light in this case) is turned off. Reminds me of neon testers which glow from induced current from nearby cables when the cable of interest is not actually live. These false positives - if that's what they are - are a PITA and I'd hoped a meter like this would know the difference. Or is it something I'm doing wrong? The instruction sheet is really awful and horrendously light on detail.

[dennis]

How would it tell the difference between induced voltages and "real" voltages?

[Cursitor Doom]

I'd have thought it would put some very light loading on the detected signal. This is one of the few advantages of old, analogue meters over modern digital meters. The modern meters impose such a light load that 'ghost' voltages appear real.

[dennis]

But isn't it a no contact test? If you could apply a small load then you may get a different answer.

[Cursitor Doom]

Yes, it's a no-contact test, but that doesn't prevent a load from being placed on it wirelessly, perhaps via the same principle as a grid dip oscillator or some other kind of resonant sensor tuned to 50hz that becomes an 'acceptor' shunt at that frequency. Just seems very odd to me that there's no apparent facility for this.

[Johnny B Good]

The meter is correctly detecting these 'phantom' voltages. The problem here is over the matter of interpretation of these voltage readings. Even a good old fashioned moving coil meter with a modest 8 or 9 kilohms per volt sensitivity on its ac range (always just less than half of its DC ohms per volt sensitivity - 20K ohms per volt in this example) will show a voltage reading on the switched live when the switch is off with no lamp fitted in the socket (or a lamp type that draws no current below a threshold voltage such as a conventionally ballasted fluorescent light fitting - a tungsten filament lamp will short out any such induced voltage).

The readings won't be as high as those obtained with a DMM, contactless or not but will, nevertheless, likely show a few tens of volts depending on the length of 'switch drop'.

As a matter of interest, just what sort of readings are you getting of these 'phantom' voltages?