3 Phase rectification problem

[matt99199]

*All measurements are made with a voltage divider circuit to cut the voltage into 1/3 of the value and observed with an oscilloscope. The voltages below have all been multiplied by 3 to give the actual voltage.

So I have a 3 phase setup with the following characteristics (this is a very old machine):
*Voltage readings taken from phase to earth ground (no neutral line is available from the transformer, is this a problem?)
Frequency: 60Hz
Phase A: 231V +27V DC offset
Phase B: 234V +24V DC offset
Phase C: 231V +15V DC offset

These three phases are being full wave bridge rectified to DC and the output has the following characteristics:
Max: 440V
Min: 370V
Delta: 70V
Frequency: 360 Hz (doesn't this seem high?)

So we have an electrolytic capacitor bank of 40uF 600V 85 degC tied to this output (and .1 uF biasing caps to earth ground on - and +) along with some kind of control circuit (it is impossible to determine what this load is but it does sink a good amount of power) when these readings are taken and this bank has failed multiple times in the past. My question is, would this frequency and delta combination cause the electrolytic capacitors to fail? Could they just be overheating from this combo? They blew out at the cathode and melted the insulation around that area last time.

The other strange part of this is that sometimes one or more of the 30A phase fuses will pop, making me believe its the load's issue but I could be wrong.

Note it is impossible to run this machine without at least 20uF of capacitance across the DC output and the load must stay attached. I do not currently have an ammeter capable of this load rating.

 

[Minder]

The ripple frequency of rectified full wave 3ph is 360Hz.
Sounds like you have a delta supply, a neutral may not be necessary for three phase full wave rectification, but if a ground is required, it would either be one intersection of two phases or the star point, if star secondary.
It sounds like you may either have voltage under rated caps or an excess load?
M.

 

[duke37]

If the capacitor is failing due to overheating, you are passing too much current. If the capacitor goes short circuit, it will cause the input fuse to blow.

You could try a much larger capacitor to share the ripple current. Do some calculations to find the dissipation in the capacitor with the ripple voltage. I have not looked up the ripple which is generated without a capacitor but 70V seems a lot, perhaps the capacitor is doing nothing.

 

[Minder]

Three phase unsmoothed ripple is only 5% compared to 1ph which is 100%, so charge current should be minimal.
M.