08-12-2011, 09:49 PM
O.K.: temperature-rise measurements. The usual scientific approach: verify the concept with known-good test equipment and set-up prior to entering uncharted waters: in this case, try the Fluke kit first with a 'known-good' transformer.
Photo. 1: the general set-up with the thermocouple in use. (Sorry about the background clutter). The transformer shown is on a no-load / temperature rise test. This particular transformer is the one I used for the design & development work: it has a secondary winding of 235v – 0v – 235v. The current rating is not known (yet).
The result using this test transformer looked sensible.
[attachment=3536]
Photo. 2: A close-up view of the Fluke thermocouple accessory.
[attachment=3537]
So now for an unknown transformer: Photo 3.
[attachment=3538]
And that is when the fun began. On a no-load test, the temperature rise was so large that I didn't need the thermocouple to tell me something was seriously wrong. After 30 minutes in this test condition, the core was far too hot to even touch! :omg: I initially suspected that I may have configured the split-primary winding incorrectly and was over-running it, but the HV winding voltages read 330v - 0v - 330v . . . as stated on the transformer and 660v on the transformer test unit. (The LV windings were on no-load ~ and measured correct, too). If I had seriously mis-configured the primaries, then those voltages would have been different . . . and incorrect.
There appears to be one primary winding of 110v and another primary winding with taps labelled from 120v to 245v. I am sure that I correctly interpreted that and connected them in phase-additive to match a 230-v supply: linking the 110v point on primary 1 to the 120v. point on primary 2 and supplying input 230v across the 0v. of primary 1 and the 245v point on primary 2. That's much easier to follow in a drawing: the photos. below help clarify:
[attachment=3539] [attachment=3540] [attachment=3541]
Hence, question: Do I have a failed transformer or have I made a transformer configuration error?
Al.
Photo. 1: the general set-up with the thermocouple in use. (Sorry about the background clutter). The transformer shown is on a no-load / temperature rise test. This particular transformer is the one I used for the design & development work: it has a secondary winding of 235v – 0v – 235v. The current rating is not known (yet).
The result using this test transformer looked sensible.
[attachment=3536]
Photo. 2: A close-up view of the Fluke thermocouple accessory.
[attachment=3537]
So now for an unknown transformer: Photo 3.
[attachment=3538]
And that is when the fun began. On a no-load test, the temperature rise was so large that I didn't need the thermocouple to tell me something was seriously wrong. After 30 minutes in this test condition, the core was far too hot to even touch! :omg: I initially suspected that I may have configured the split-primary winding incorrectly and was over-running it, but the HV winding voltages read 330v - 0v - 330v . . . as stated on the transformer and 660v on the transformer test unit. (The LV windings were on no-load ~ and measured correct, too). If I had seriously mis-configured the primaries, then those voltages would have been different . . . and incorrect.
There appears to be one primary winding of 110v and another primary winding with taps labelled from 120v to 245v. I am sure that I correctly interpreted that and connected them in phase-additive to match a 230-v supply: linking the 110v point on primary 1 to the 120v. point on primary 2 and supplying input 230v across the 0v. of primary 1 and the 245v point on primary 2. That's much easier to follow in a drawing: the photos. below help clarify:
[attachment=3539] [attachment=3540] [attachment=3541]
Hence, question: Do I have a failed transformer or have I made a transformer configuration error?
Al.
