Transformer Interlayer Isolation 1

[dunya]

Anyone has an idea on how to test layer insulation within a winding on foil wound delta-wye transformer (115VAC, 400Hz) . The test will serve as an isolation and verification for shorts (latent that can not be discovered during functional testing) across layers test were interlayer winding is isolated. I am thinking of applying a high voltage (maybe high frequency pulsed) across the winding and monitor the current. Any ideas?

 

[VEBill]

High frequency will test the interwinding capacitance (probably not what you want). Insulation can be tested with DC.

Limit the current otherwise the test may be destructive.

Google: Hi-Pot Testing (and variations).

 

[ScottyUK]

A technique often used to test generator windings is the Recurrent Surge Oscillogram, or RSO, which injects a fast pulse into the winding under test and uses analysis of the returned pulse based on Time Domain Reflectometry techniques to determine if a fault exists, and approximately where it is in the winding. I haven't heard of it being applied to a transformer, but I see no reason why it can't be.


----------------------------------

Your body may be a temple. Mine is an amusement park...

 

[Warpspeed]

I am not sure that a high frequency or pulse test would give you a conclusive answer. You are looking for perhaps one shorted, or "partially damped" turn out of many turns in series.

Jacking up the test frequency will also raise eddy current losses in the core significantly, so the drop in Q at a higher test frequency due to one (partially) shorted leaky turn may not then be quite so obvious.

Possibly the best you can do is to shock excite the winding with a fast pulse, and watch the damped resonant decay. Count the number of cycles to half amplitude on a storage oscilloscope. If you have a number of identical transformers to test, any significant change in one transformer may be trying to tell you something.

 

[ScottyUK]

The RSO certainly picks 'em out. Interpreting the trace is the key - it takes more experience than I have to be able to locate the fault within the winding.


----------------------------------

Your body may be a temple. Mine is an amusement park...

 

[VEBill]

Ah, within a single winding [sorry].

If you want to detect a hard short, then you could simply measure the end-to-end resistance using milli-ohm 4-wire techniques to see if the reading is lower than it should be (due to several turns being shorted out).

To detect defective insulation where it isn't quite shorted, then perhaps the transformer could be exercised for very brief periods (low duty cycle) at extreme voltages and 'listen' (*) for arcing (* noise burst at frequencies other than 400Hz). If it doesn't arc at voltages far above the operating spec, then it is probably good enough.

 

[Warpspeed]

A foil winding will usually be as wide as the transformer, so the edge of the foil may perhaps be visible or accessible somehow ??

A shorted turn is going to run fairly hot, perhaps it could even be seen with infrared sensitive equipment, or probed with a thermocouple down the edge of the winding.

Could the volts per turn be probed ? A shorted or leaky single turn would have conspicuously less voltage across it than all the other turns.

 

[Skogsgurra]

A surge test is the standard method for normal transformers and motors. I think that it could work on this (small?) transformer as well.

See Baker Instruments

http://www.bakerinst.com/BakerWeb/Products/Products.htm

Gunnar Englund

http://www.gke.org

 

[VEBill]

A veritable feast of good ideas...