Rogowski current probes? 2

[Renovator1]

And on a totally related note to my other post... anyone home-brewed their own Rogowski current probe? I got really excited about the CWT series from PEM - you can uncouple the loop and slip it in between the legs of a TO-220 part to measure, eg., drain current! - but wasn't too excited about the price: $1445?!?! Since that is clearly not in the budget this year (or next, etc., and so on) I was thinking of rolling my own with some thin (2mm?) teflon tubing for the coil form... Or if someone knows of another supplier of a similar style Rogowski probe that *isn't* smoking crack, I'd love to hear of it (yes, I've done a lot of Google searching... coming up a bit short).

 

[ScottyUK]

Yes, I've rolled my own a long time ago. The coil is fairly easy, if time-consuming. Getting a flat wide frequency response takes a little more effort in the intergrating amplifier but is not especially difficult.

Have you contacted Rocoil in the UK? This is a small company making custom and semi-custom Rogowski coils and associated electronics. The guy to speak to is the owner, David Ward. He is very knowledgable on the subject and has done some unusual applications for ultra-high accuracy and for measuring some outrageously large currents. Highly recommended!

http://www.rocoil.co.uk

BTW, I have no association other than being a satisfied customer.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[Skogsgurra]

I have used the PEM coils. Wide bandwidth but be careful with the ringing when current pulse risetimes are down to a few hundred nanoseconds. Otherwise very good for the price.

The problem is that they get destroyed when you measure rotating shafts on motors and other machinery. I lost two that way.

I also made myself a few when the conductors (heavy tubing in this case) were too wide for standard coils. Picture attached.

I have used these with passive integrators (RC filter with around 1 ms time constant) and active integrators. Lately, I have used the integrating function in USB scopes. That gives a great flexibility but also a problem with the DC offset, which drives the integrator into saturation. It can be overcome in many ways, but you have to take care of it if you measure over several seconds.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[Renovator1]

ScottyUK - any practical details on the coil construction should you remember them from such a long time ago would be most welcome! E.g. - how many turns and what gauge wire? What did you use as a former? And, most importantly, how did you get the two ends to meet as closely as possible in an also easy-to-disconnect way??? That's what I'm really scratching my head over. The CWT series from PEM looks like it uses a tubing Tee connector but I can't imagine how that would work AND the coil be accurate as the coil ends will, by necessity, be at least two diameters apart from each other... this is supposedly Very Bad.

Gunnar - Interesting observations. I'm going to need to measure current rise times in the 50-300nS range so it would seem the PEM integrator is not up to the task for me anyway.

The most common solution to integrator saturation is to short out the integrator capacitor with a FET either periodically, on demand by the operator or, ideally, at every zero-crossing. Frankly, trying to pull off the latter without sampling the voltage of the circuit under test seems a bit difficult, though.

Thanks for the picture of your home-brewed coil!

FYI - Athena Energy, LLC, is willing to sell the probe only - everything but the integrator, that is - for $150! That's definitely well within the budget. Also, for larger industrial applications I found out that Fluke sells Rogowski current probes. Check out the i3000S, for example. It's totally useless for my application - measuring the individual collector currents of a bank of IGBTs - but it looks perfect for buss bar monitoring, VFDs, etc... and the price is definitely right: $400US!

 

[Skogsgurra]

I don't know about the i3000S, but the similar Rogowski probe from LEM, Switzerland has a very limited bandwidth. Something like 20 kHz. Far from what you need. I think that the Chauvin-Arnoux probe, which (I also) think, is identical to the Fluke device has a similar bandwidth.

I have also used an FCC HF clamp-on transformer. Goes to hundreds of MHz. But has a definitive influence on the impedance. Which a Rogowski doesn't have. Or very slightly so.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[ScottyUK]

I used them for a similar application to yours but on a very large rectifier with six thyristors per limb on the bridge to monitor current sharing. The coils passed around the hockey puck thyristors.

As a former try some of the medical PTFE tubing which is occasionally used to insulate compenent leads. The joint doesn't matter too much and overlap is preferable to a butt joint with a gap. For coil as small as you are trying to make the connections to the outside world will need to be simimilarly small. The ideal would be shielded twisted pair but I don't think you'll find it in very small sizes. Enamelled wire twisted will probably be ok. Wire gauge is fairly arbitrary but you need a reasonable number of turns in the length available and you need to be able to work with it. Anything below 40AWG is going to be too fragile - 40 breaks if you breathe heavily - but I would guess you are looking at 40 AWG on a 1.0mm former.

Are the TO-220 leads splayed at all or staggered? Might be easier to drill a pair of holes straddling a PCB track to the IGBT and thread the coil through them rather than trying to get around the device itself. You then have the chance to anchor the wiring etc securely to the PCB.

How many of these are you looking for?


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[Renovator1]

I got some more details on the probe from Athena Energy. He uses 36awg wire and 2mm PTFE barb fittings and the bandwidth with his integrator is ~30MHz (should resolve a ~33nS rise time, then). He indicated he reduce the number of turns to allow measuring up to 1000A; my understanding is that the number of turns would only need to be reduced to preserve the same bandwidth at the higher current as the coil itself cannot be saturated.

At any rate, I definitely can't justify making my own probe when his price is $150US; making my own integrator for his probe, though, is definitely compelling as the combo costs $950. Yeah, I'll definitely spend half a day to make an integrator and calibrate it with my IGBT pulse tester to save $800. I mean, How Hard Can It Be??? (Famous last words...)

 

[Skogsgurra]

Great info!

Thanks.



Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[Renovator1]

Summary (in case you can't be bothered reading this whole post): If you design SMPS, or anything else where it would be Really Useful to know what the current waveform looks like without disturbing the circuit, then you *need* to buy or build a Rogowski Probe. I bought one from Athena Energy and have been *extremely pleased* with it! See:

http://www.athena-energyllc.com/athena_energy_rogowski_information.htm

Update: Dimos Katsis of Athena Energy was extremely helpful when I talked to him on the phone and even managed to convince me that rolling my own integrator would probably cost more time than I expected, and I'd still never be sure it was perfectly calibrated. So, I bought the complete probe package - coil + integrator, rather than just the coil as I had intended. When it arrived I was not terribly impressed with the way it looks - a lot like a ham radio tinkerer's weekend project - but man did I get a big ol' grin the first time I used it. Being able to thread the probe coil around the source (emitter) of a To-247 part in-circuit will make power circuit design (and the troubleshooting thereof) infinitely easier. Really, folks, if you are at all involved in switching high currents around at high speeds a Rogowski probe is damn near indispensable. Yeah, I know it sounds like I've definitely been drinking the Kool-Aid here but for good reason.

 

[ScottyUK]

Pleased you got a good solution to your problem. I love 'em, ever since I first encountered them in the early 90's. The only downside is the lack of any DC measuring capability, but even that can be advantageous in certain applications. I am slightly bemused why they remain relatively unknown and are treated as mysterious exotica by many people. Perhaps it is a part of The Great Conspiracy.


----------------------------------

If we learn from our mistakes I'm getting a great education!