You are seriously over your head here and should not be playing
around.
A multimeter is not the correct tool for this job!
You need
3-4 channel scope with triggering capability
4pcs. X10 probes
4pcs 4kv probes
3pcs 3000A Dc/high frequency current probes.
480V to 24 V control transformer
4pcs. 1500v wideband isolation probes
In any case, you are not even measuring the voltage correctly.
The incoming AC is measured:
buss to buss <A,B> ie Black - Red <B,C> ie Red - Blue <C,A> ie Blue
- Black (DELTA or WYE)
or buss to neutral. <A,N> ie Black - White <B,N> ie Red - White
<C,N> ie Blue - White (WYE only)
or buss to ground <A,G> ie Black - Green <B,G> ie Red - Green <C,G>
ie Blue - Green (WYE only)
The main 3 phase thyristor bridge you describe should not be making
DC, it should be making high frequency AC, hence the description
inverter! I would assume, the multimeter measurements would be
consistent from pair to pair but totally inaccurate- you will have no
idea what the true voltage or waveform really is.
Now there are 3 input AC busses, each connected to a pair of
thyristors, one positive, one negative. On the output of the thyristor
set, are there just 2 output busses or are there 6 and do they connect
directly to the load? Is the load connected to ground or neutral in
anyway, perhaps in the center of the coils?
This whole thing sounds really dangerous at best. The AC supply side
should be stable +/- 15% with the thing on or off and in no case
should a 440vac phase start reading 600v or 1000v to ground at the AC
Distribution (Circuit Breaker or Fuse) Box. If it does, you've got big
problems and they might be utility related. An installation of this
size will usually have a company supplied step-down transformer from
say 13.5KV to 440 made up of 3 interconnected transformers one for
each phase. If the junction that connects the three secondaries
becomes weak, that is, fails under load, the system goes up for grabs.
That is in a WYE connected system. In a Delta connected system it is
yet....Well here you read about it:
http://www.elec-toolbox.com/usefulinfo/xfmr-3ph.htm
Since the secondary of the Delta is not ground referenced,
measurements that are ground referenced or neutral referenced have no
meaning and can vary wildly. You must measure Buss to Buss, Phase to
Phase. I am guessing you have a Delta Connected distribution
transformer for this inverter.
With the inverter off at the input side:
Trigger the scope with a 440v to 24v control transformer. Connect the
440v side from phase A to B and ground one leg of the 24v secondary.
Connect the X10 probe to the other 24v leg and the probes ground where
you grounded the other side of the secondary. Connect the probe to the
trigger input. Set the scope for external triggering and adjust for a
stable trigger at a 60hz rate. ( You can cheat here and use the LINE
TRIGGER setting instead)
Using the 4kv X100 probes, Connect both probe grounds to the ground
used above. Set the Ch1 and CH2 scope inputs to add. Press the invert
button on CH2.
1) Connect one to A and one to B.
Set both CH1 and CH2 to identical attenuator settings in the CAL
position so the trace fills the screen vertically (about 1-2 V per
division). Adjust the sweep so you can see 2 complete sine waves.
2) Move the probe on phase A to C and measure the B - C waveform
3) Then move the probe on phase B to A and measure the A - C Waveform
All three should be about the same
Connect your current probe to CH3 and set the attenuator for about
1500A full screen using the probe factor, DC coupled .
Start the inverter
4) Clamp it around each of the 6 thyristors or it's output lead, one
at a time noting the CH3 waveform
5) Clamp it around each of the 3 AC input busses and note the CH3
waveform
Repeat 1) 2) and 3) with the inverter still running.
Repeat 1) 2) and 3) with the inverter still running but this time
connect to the output side of the thyristors.
Be careful, setup each measurement, use protective gloves, a
switchgear mat and have an assistant standby with an 8' 2X4 to pry you
off if you get grabbed. You are working at massive power levels. The
fact that Busses are overheating indicates a high current fault that
could fry your face at any moment. Death by electrocution is very
painful. Death by molten copper plasma is excruciating.
At this point the faulty section should be obvious. You could take a
chance and replace the offending thyristor.
The next test now that you have narrowed it down to a single device
requires you to measure the trigger voltage and this can be tricky
since it is .6v (actually it could be 20) referenced to one side of
the thyristor in a really noisy environment. Measuring it wrong can
trigger the thyristor at the wrong time and then you're in a world of
shit. Use 4KV probes on the three terminals of the thyristor. The
gate terminal will be pulsed a few volts above one of the other two
terminals.
Using the 4kv X100 probes, Connect probe grounds to the ground used
above. Set the Ch1 and CH2 scope inputs to add. Press the invert
button on CH2.
a) Connect one probe to the GATE and the other to each of the large
terminals of the thyristor.
Set both CH1 and CH2 to identical attenuator settings in the CAL
position so the trace fills the screen vertically for each
arrangement. The correct orientation will require a lot of gain around
1-10mv/Div . Start at 10v/div and reduce 1 click ch1 then 1 click ch2
etc.
Repeat this test on a known working thyristor for reference.
http://www.littelfuse.com/data/en/Product_Catalogs/PowerThyristorApplicationNotes.pdf
Where are you located? Thought of hiring someone to do this for you?
Tom
