Are clamp ammeters able to measure current through SMPS coils?

[debininja]

Let me expand my question a bit: is it possible to measure the average current consumed by a SMPS by placing the cuff around the trace(s) or wire(s) coming out of the transformer? Right now, I just use a 0.5Ohm shunt resistor connected in series to the mosfet and the xfmr, measuring the voltage drop on my multimeter. I've never owned a clamp meter before, hence the nooby question.

I'm thinking about getting one of these:

[Ian.M]

Probably not.  The UT210E clamp is only rated for DC and true RMS between 50Hz and 60Hz.  A SMPSU will run at 60KHz or more.  It also has rapidly increasing error for non-sinusoidal waveforms.  e.g if the crest factor is 3, the measurement accuracy is degraded by +/-7%, so it basically cant cope with measurement at the SMPSU transformer.  There is even a risk of damage to it as its internal current range circuits may not be designed to cope with such a high frequency pulsed signal.

However, if the SMPSU has a PFC circuit on its input controlling the crest factor, you should get a reasonable result measuring the input current before the line filter. and the DC output current should also be measurable subject to the accuracy limitations of the meter range you are using.

[debininja]

Thanks for the reply, just quickly looked up what crest factor is.

Last question: the cuff might work on the secondary side after rectification to DC and proper LC filtration, right?
The crest factor will be greatly diminished, but there might still be some high frequency ripple, so I imagine it wouldn't work all too well, huh. 

[Ian.M]

It should work OK on DC Amps after the LC filters - the meter's DC Amps circuitry almost certainly has low pass filtering and you'd have to be fairly unlucky for the residual HF ripple to upset the sampling circuit after that.

[SeanB]

I have built small current transformers in the past, using a largish ferrite bead ( or a very small toroid) and placing the lead of the switching device through it, and winding on 10 turns of thin enamelled wire, and twisted pair of the wire to a thin coax cable, terminated with 50R into a scope. Accuracy not too great, but has a very high AC bandwidth, and calibrable with known currents to get a transfer ratio of current. does have some effect on the high frequency components, but if there is already a bead there it can be a good substitute for it.

[P90]

This is the proper tool...
http://www.ebay.com/itm/322536436620

With the proper price!    :-)

[David Hess]

If you can accept AC only operation, then Rogowski type current probes are less expensive.  PEM makes them.

It is often better to simply include a current sampling circuit during development although if this is done on the line side of an off-line switching power supply, then precautions need to be taken to use it.

[debininja]

Thanks for the reply everyone! I'm just a po' ol' college student, and you guys are linking me items that cost $1k or more...  <--I love this scared emoji
I'll settle for the shunt resistor method for now, until I save up... quite a bit...
(And I really like your idea, SeanB. I could definitely build one of those).

(P.S. I don't know why this forum doesn't have a setting to close the thread when someone's question is answered).

[tesla500]

Tektronix makes some very nice clamp on current probes. The TCP312 goes from DC to 100MHz at 30A. Not cheap though, a used one is about $2k

[tautech]

Thanks for the reply everyone! I'm just a po' ol' college student, and you guys are linking me items that cost $1k or more...  <--I love this scared emoji
I'll settle for the shunt resistor method for now, until I save up... quite a bit...
(And I really like your idea, SeanB. I could definitely build one of those).

(P.S. I don't know why this forum doesn't have a setting to close the thread when someone's question is answered).

If you close a thread it wouldn't permit other replies to come along later........you don't want that. 

Years ago I was in a similar pickle and a viable option is a Current xformer that sometimes are populated on SMPS, especially the beefier ones. The one I found on a very old PC PSU was a potted 2 pin job with the current measured passing through a cable from the PCB passing through its center hole.
A mate has a Tek P6022 (100 MHz) AC current probe and using the 2 in tandem we were able to shunt the output of the CT to get them to roughly agree at frequencies up to a few 10 KHz.
I considered it plenty good enough for normal SMPS work but set my sights on getting a P6021 (60 MHz) of which I now have 2.
Sure a DC capable current probe is what we all want but that's normally applicable for pure DC or frequencies below the spec of an AC probe where you're likely within the realms of other measurement techniques.

If you are patient you can find P6021 and terminator for a few 100$ on eBay. Took me ~2 years to get 2 units at what I was happy to pay. Other equipment can date and need replacing but good current probes are an investment for life.