Using a current transformer as an oscilloscope probe

[cvanc]

Hi all-

Have any of you ever used a current transformer as a scope probe?  Like in the attached spec sheet.  Did it work OK?

[tautech]

Yep, one from an old PC switchmode PSU so to hopefully have some range into the 10's of KHz.

Current waveform reproduction and accuracy good enough for low frequencies when checked against a Tek P6022.
We had to trim it some with shunting a resistor across the outputs so to get meaningful measurements from it.

Not sure I'd wanna use one that was rated to only 400 Hz.

[T3sl4co1l]

How do you use a current transformer "as" a probe?  You can certainly connect it up with a burden resistor and thread it on a wire; but it's not exactly clip and go like a split-core probe is.

Tim

[tautech]

How do you use a current transformer "as" a probe?  You can certainly connect it up with a burden resistor and thread it on a wire; but it's not exactly clip and go like a split-core probe is.

Tim

Yep, this ^
TBH it's a right PITA and you get real sick of breaking connections to insert the thing.
Wasn't long before I bought a Tek P6021.....well waited until there was one on eBay that I could afford.

[cvanc]

Thanks everyone.  Here's a bit more detail.

I do a lot of audio amp testing.  Some amps have a single ended output and some are balanced.  So if I'm, say, looking to see where they clip using my scope across the load resistors I need to take care about that unbalanced scope input.

I do most of my bench testing (all the simple 'triage' stuff anyway) with a 400Hz oscillator.  I have big 8Ω non-inductive loads.  I want to use transformers to sniff the current in these loads so my scopes' unbalanced input is isolated from the DUT.

I would hook the transformer secondaries directly to the scope inputs; no shunt or other load would be added.  The 8Ω loads and associated transformer would be mounted onto a small piece of wood and permanently paired to each other.

This is solely to see the shape of the waveform, I have nice bench meters across the loads to tell me what the voltage is.  I do not need any amplitude accuracy from the scope.  Just need to see that 400Hz sine wave, and see when it flattens out.

These transformers are rated for 400Hz, so... why wouldn't this work?  I think it should be a decent, cheap solution to one (and only one) measurement issue.  Right?

Thanks again.

[tautech]

For that use I'd give them a go for sure cvanc.

You can always confirm their signal fidelity is more or less correct by using math and 2 channels for scope isolation from the amp.

[T3sl4co1l]

Yeah, that's fine for measuring output current.  And as long as your load resistor is resistive at the relevant frequencies (I would guess it's good up to ~low MHz), current or voltage will be an equivalent measurement.

You can solve the endedness problem by using two probes, with the scope set to difference (usually MATH menu).  You still need the grounds connected; it should be okay to connect these to chassis or signal or power ground.

Tim

[cvanc]

Just to close this out I have been happily using this setup for months now and it meets my needs well.

All they ever look at is 400Hz sine waves as this is on a dedicated audio amp test bench. They are hardwired to 8 ohm load resistors.

This setup is nonlinear at low levels; I haven't bothered to investigate why.  It works fine from a couple watts up and gives new life to one of my old 2 channel scopes, which can't handle stereo differential inputs.  Thanks for the input, folks 

[trobbins]

Non-linearity would likely be reduced if you use a burdon resistor to load the secondary.  Even a 50 ohm load from a scope may improve your low end linearity.  But obviously your scope has to manage a lower FS signal level, and you aren't really interested in low-end linearity.

As an aside, LEM have a variety of small active CTs with very wide frequency range, and the smaller ones such as LTS6NP are quite cheap as they are used for smps - they are linear across a wide range, and can be used with soundcard inputs as well to look at much wider SNR than a scope can provide - obviously more hassle to set up as they require powering.

[nctnico]

Non-linearity would likely be reduced if you use a burdon resistor to load the secondary.  Even a 50 ohm load from a scope may improve your low end linearity.  But obviously your scope has to manage a lower FS signal level, and you aren't really interested in low-end linearity.

Correct. With the right loading resistor a current transformer can have a very flat frequency response. A CT for use in switching power supplies can be flat between 10kHz and >5MHz. I have used these in the past as a cheap way to measure the AC ripple in switching power supplies.

[joeqsmith]

https://interferencetechnology.com/the-hf-current-probe-theory-and-application/

[capt bullshot]

I would hook the transformer secondaries directly to the scope inputs; no shunt or other load would be added.  The 8Ω loads and associated transformer would be mounted onto a small piece of wood and permanently paired to each other.

This is solely to see the shape of the waveform, I have nice bench meters across the loads to tell me what the voltage is.  I do not need any amplitude accuracy from the scope.  Just need to see that 400Hz sine wave, and see when it flattens out.

You can't leave the secondary open (the scope input is equivalent to open here), that'd just result in a spiky waveform that has nothing in common with the primary waveform and might damage the scope due to high voltage spikes.
Calculate a load resistor according to the specified load (the VA rating), connect it across the secondary and bring the voltage across the resistor to the scope input. That'll work nicely. For a ballpark value, 1R load for the AL-500 should work.

[David Hess]

I would hook the transformer secondaries directly to the scope inputs; no shunt or other load would be added.  The 8Ω loads and associated transformer would be mounted onto a small piece of wood and permanently paired to each other.

This is solely to see the shape of the waveform, I have nice bench meters across the loads to tell me what the voltage is.  I do not need any amplitude accuracy from the scope.  Just need to see that 400Hz sine wave, and see when it flattens out.

You can't leave the secondary open (the scope input is equivalent to open here), that'd just result in a spiky waveform that has nothing in common with the primary waveform and might damage the scope due to high voltage spikes.
Calculate a load resistor according to the specified load (the VA rating), connect it across the secondary and bring the voltage across the resistor to the scope input. That'll work nicely. For a ballpark value, 1R load for the AL-500 should work.

If you had not said it I would have.   The maximum volt-time product before saturation means that a higher secondary voltage due to high load resistance produces a high pass response and of course saturation makes the response very non-linear.

For higher performance applications, an amplifier with zero input resistance can be used to extend the low frequency response.  See the old Tektronix Type 134 for an example of this.  With some cleverness the low frequency response can be further extended by using a negative input resistance to remove the effects of the transformer's winding resistance.

Some current transformers could be derated to operate directly into a 50 ohm oscilloscope input.