what causes these glitches in a transformer output?

[ledtester]

I'm wondering what is causing glitches in the output waveform when I run a sine wave from my function generator through a transformer.

In the attached pics note the major glitch at the crests of the output and also the smaller ones to the left and right.

This is with a 2 KHz sine wave input. As I increase the frequency the major glitch at the top goes away, but the one of the left is still perceptible up until about 70 KHz (under x10 magnification). The location of the glitch also seems to shift with frequency moving to the left as the frequency increases.

The transformer looks like one you'd find in a PC power supply, ie. something like this:



The function generation is a FY3224. I've connected it to one winding and I'm probing another winding with the scope.

[Seekonk]

Your signal generator is incapable of driving that load.

[jitter]

I would have expected much more severe distortion if the generator was overloaded. More like clipping or similar.

[Jeroen3]

Looks like the H-bridge has a short for a few microseconds.

[danadak]

Is this a DDS based signal generator ? Looks like the flip of the
sign and LSBs of a DAC producing the glitch.

Regards, Dana.

[T3sl4co1l]

Hah, looks like a shitty class-C output stage. Get a better function generator?

Is this a DDS based signal generator ? Looks like the flip of the
sign and LSBs of a DAC producing the glitch.

Regards, Dana.

That seems second most likely. If it's present for all loads...

Get a better function generator!

Tim

[ledtester]

This is a DDS function gen, but I don't see the glitch with all transformers.

Some other observations:

- no glitch with resistive loads - I've gone down to a 3 ohm load. The output impedance of the FY3224S is 50 ohms.
- when I observed the glitch I was driving the transformer with a 1 Vpp signal and observing an output level of 0.135 Vpp. I've performed the test with other transformers and have observed basically the same output level but without any glitches.
- the FY3224S does have issues with it's switching power supply (see this thread from about reply 137 onwards):
https://www.eevblog.com/forum/testgear/feeltech-fy3224s-24mhz-2-channel-dds-aw-function-signal-generator/137/

Can anyone suggests any other tests to perform to help understand the issue? This is an AWG so I can download a custom waveform.

[T3sl4co1l]

If it's the output stage, you should see the same glitch in the middle of any even waveform.  For example, in the middle of the flat section of a square wave.

Tim

[kfitch42]

Just to clarify, when using the "bad" transformer with the glitch in the output signal, do you see a glitch on the signal at the input of the transformer?

[ledtester]

Just to clarify, when using the "bad" transformer with the glitch in the output signal, do you see a glitch on the signal at the input of the transformer?

Yes, the glitches are visible on in the input side - but, of course, only when it's actually connected to the transformer.

See the pic below when I drive the transformer with a triangle wave. Top trace is the second output of the function gen displaying what the first output is set to; bottom trace is the output of the transformer which is being driven by the first function gen output.

So I think the "glitches" are an amplification of the MSBs (?) of the DAC changing, and you only see it in the sine wave near the crests since that is where the signal is relatively flat (i.e. second derivative is near 0). This transformer just happens to be good at amplifying the DAC switching noise.

Do "real" DDS function gens have this problem?