If the scope has the signal ground commoned to frame ground, 'with great caution' would be a good starting point for the novice.
If the shunt is at mains potential, you really need to use an isolating amplifier to connect to the shunt in order to protect you and your instrument. Even better would be a current probe: the best ones (and most expensive) are from Tektronix: the AM503 amplifier and one of the 6301/6302/6304 probes, chosen to suit the current in question. LEM and Chavin Arnoux also make reasonable probes for a reasonable price, and these are perfectly good unless you are looking for very high accuracy or are interested in very high frequency (MHz).
There are other cute tricks which you can play like disconnecting the scope chassis from ground, placing the instrument on a rubber mat, and allowing the whole instrument to float at line potential. The instrument needs to be set up prior to energising the load, and treated with care because the casing is 'live'. It is not really acceptable practice in this age of high bandwidth current probes, isolation amplifiers, and fully insulated scopes.
If the scope has two channels, you can connect two probes to the shunt terminals and use the differential input feature. Don't put ground clips on the probes.
And if you are contemplating using two probes at once, depending on the scope, their ground clips are common. That is shorted together. So remember wherever the two grounds are connected it is the same as if you used a jumper to connect those two points of your circuitry together.
I do not recommend the differential approach. Not even for DC - the CMRR needs to very high if you shall be able to read a 200 mV difference on top on a x00 V potential. And the channels have to match closely.
If you are doing SCR or IGBT work, then skip that shunt. Get a DC current clamp. That is safer and usually accurate enough.
I want to measure the voltage across and current through a generator field winding in steady state, and compare the waveforms to the ones i simulated in MATLAB. There are already panel voltage and shunt current meters present in the turbine room, so i use the terminals at the back op these meters.
The best solution to connect to floating circuits is to use a differential probe, like the SI-9002. However these things are quite expensive $200-$300 at least. So i want to use a QnD solution, or a kludge if you like. I thought about using an 1:1 isolation transformer between the mains and the scope, e.g. keep the scope floating, and connect is to the circuit with a 10x probe. I'll wear safety gloves & boots, as i don't like getting fried.
OK. You can do that (did that myself lots and lots of years ago - I didn't know better). I wouldn't use a 10:1 probe, though. It eats 20 dB of the signal. And you need as much signal as you can get to overcome the inevitable common mode noise. Yes, it gets in even if you isolate the scope.
QnD solutions are great until you have to replace the instrument, or someone gets hurt or killed.
Generator fields are unfriendly places for test equipment. On the IEC 1010 categorisation, it would be a Cat III or Cat IV application, depending on the size of the unit. The fewer direct connections you make to the system the better.
We've all done stupid things in our time - I guess this might be yours. If your company won't provide the tools to do the job safely, don't do it at all. Your life is worth more than their bottom line.
Among the candidates for my 'most stupid' moments - and there have been a few over the years - was using 1.6kA semiconductor fuses as shunts to measure current in rectifer stage of an AVR serving a big directly excited generator. The AVR output was rated 1200V / 4700A continuous; available fault level is in excess of 70kA. I had a mains powered 'scope floating on an isolation transformer with a single-ended probe across the fuses to measure current. The problem we had was that there are six paralleled thyristors per bridge limb, and determining which of the six were blown was kinda tricky.
It was effective, but utterly foolhardy: if I'd shorted something out the resulting explosion would have left a glowing smoking hole in the floor. I was younger and more reckless in those days; three colleagues killed in an electrical explosion is a constant reminder of just how much energy is let out when a fault occurs in a high power system and I am much more cautious as a result. Sometimes I think we get complacent in my department where we often talk of 'only' a few MW. In context, perhaps it is only a few MW out of nearly 2GW of station output, but sometimes we need to remember what 'only' a few MW looks like when it is out of control.
Hi, i succeeded in 'safely' connecting the scope to the rotor winding through the panel voltmeter connectors and took photo's of the screen. As you can see the voltage across the winding is not constant, but pulsating which verifies the simulation results. Very satisfying indeed.
PS: Do analogue voltmeters give rms values? The needle kept pointing at a constant 30 volts, while amplitude of the pulses are 200V max.
You didn't connect to the shunt then? I wonder what the current looks like...
Yes, your voltage seems to be quite unstable. Does the AVR have to be that fast (superfast, I would say)? Is that the problem that initially made you measure the thing?
About RMS/AVG: An ordinary analogue DC voltmeter shows average value. It is only if you happen to have a soft-iron or electro-dynamic voltmeter that you will read RMS.
30 V DC seems a bit on the low side for this waveform. I would have guessed more. Like about twice that value.
Didn't bother with the current, i expect it to be constant ie flat dc.
The generator is not running, and the exciter is in 'drying' mode, supplying unregulated power to the field to keep the generator warm when it is off-line. Instead of being self-excited (the exciter voltage shunted off the generator output voltage) the voltage is provides by a auxillary source. It is essentially the no-load current, without the boost-current part.
The dc smoothing caps are way to small to smooth the voltage. This brings me to the duality of capacitors vs inductors; voltage vs current smooting. Does it matter that the power delivered to the field winding is not constant, but the current is?
PS: The harmonics in this circuit are driving me nuts! No nice phasor analysis here
