Measuring a transformer with an oscilloscope

[kemalkemal]

Hi all
I have a 200V-24V transformer. In order to see waveforms from the secondary side i made some oscilloscope measurements. As you can see from picture i tried 4 setups.
I have found out that 1st one was totally wrong and idotic, because connecting ground clip of oscilloscope probe to one port of secondary winding means that i am shorting that port to earth.
But couldn't understand why there is a difference between 2nd and 3rd setups. The only possible idea came to me is; at 2nd setup, one of the channels act as it is connected to the circuit ground and oscilloscope displayed difference of 2 channels automatically. How this could happen when i didnt attempt to make a subtract operation between channels? Aren't oscilloscope channels independent from each other?
I think the 2nd is the closest to the correct setup. What should be the correct setup?

 

[KlausST]

Hi,

the voltage output of a transformer is across both output winding connections.

Therefore the usual way is to use method 1)
(with an voltmeter you´d use the same method, too)

Now you are concerned about earth connection.

With a usual transformer this is no problem.
But if the transformer output is related to earth GND already (before connecting the scope) then this may cause a problem.
Then you should use the differential method of 2) and use MATH and show the difference of both channels.

Klaus

 

[kemalkemal]

Hi Klaus, thanx for the answer.
In the begining, method 1 was also the usual way to me . But as i stated the resultant waveform was crazy. I checked after your message my transformer; secondary side is totaly isolated (no connection to earth). So the mystery perseveres :smile:

 

[FvM]

The only "mysterious" aspect involved with the measurements is in fact explaining why you see an output for cases 2 to 4. But if you consider transformer inter-winding and winding-to-case-ground capacitance, the results are completely expectable.

 

[KlausST]

Hi,

24V RMS (sine) gives about 68V pp.

This means method 2) gives the most reliable values.

But I´d be interested on what transformer exactely you use.
If it is a small transformer (about 1W or less) the no load voltage could easily be 1.5 times the rated (full load) voltage.

Klaus