Electronics > Repair

Tektronix TCP202 current probe repair - Schematic and suggestions needed

<< < (17/17)

Amphion:
I wanted to add something to this conversation that may be overlooked. I recently purchased a later model A6302 probe off e-bay. It seemed to be mostly working, but had AC overshoot that seemed too much. I finally figured out that the problem was with the core not mating perfectly, even though it would look like it was from casual observation. (The mu-metal housing looked to be perfectly mating, but the core itself was not).

It seems apparent that these probe cores are made by potting the ferrite in the mu-metal shield using a soft epoxy, and then the mating surfaces (mu-metal and ferrite) are ground/lapped flat/planar as one unit. I have much experience with epoxies and suspected that the epoxy had shrunk with time, pulling the core a little further back into the mu-metal housing, creating a gap between the ferrite surfaces. This is especially a problem with softer epoxies that have plasticizes in them.

Taking the probe apart I confirmed with a straight edge under the microscope that the core was indeed slightly below the surface of the mu-metal housing. I then starting lapping the cores, abrading the mu-metal until I could see that I was contacting the ferrite. I started with 600 grit sandpaper, working toward 1 micron diamond lapping film (used for polishing fiber optics) in multiple steps, all done on a flat glass surface.

After all this, the AC overshoot is nearly completely gone and the probe is performing to my expectations.

Some notes for "young players" on lapping, you have to have good technique!:
-You must be aware of how/where you're applying pressure as you're dragging the probe core across the sandpaper/lapping film, or you will put a tiny but meaningful convex curve in it. The ends or sides will get lapped a little more as the lapping drag force creates a tilt force and you have to compensate for this. I actually had to do some final lapping with half the core off the edge of the film to remove this small convexity I had put in the middle of it. Under the microscope I could see that I had a tiny tiny bit of a teeter totter thing going on, and it was affecting the performance of the probe.
-The ferrite is very brittle/frangible. Microscopic grains will come off, and then when you drag back over them they will take more grains out, leaving a rough surface in the ferrite. Consider this in your technique. I ended up lapping under a stream of water to wash the debris away.
-I was surprised by the probe sensitivity to this mechanical fitup. It's been talked about before, but I don't think it could be overstated. I was seeing effects that would be measured in microns.
-This process is not for the faint of heart and I don't recommend it for everyone!

I have attached a picture showing the small core pullback I'm talking about. Unfortunately this picture is half-way thought the process. I didn't think to take one at the beginning, so you're only seeing about half the gap here.

MarkL:
Thanks, Amphion, for the great insight.  I will have to look at the core mating on my A6302 that has overshoot.

Did you find it was just the lower core?  Or did you have to do both?

Do you have any before/after waveform captures?

I've also done a fair amount of fiber polishing, but it was many years ago.

Weston:
I noticed that my P6021 current probe had a scratch on the mating surface causing a poor LF response. Given my issues with the response on the TCP202 I decided to buy some lapping films and try polishing the transformer mating surfaces on both. As Amphion described I could see that the epoxy had receded, leaving the core surface slightly recessed.

I have not lapped anything before, so my process may be a bit off, but it had enough of an impact where I am convinced this is the main, if not only problem. I was not able to get the core face mirror shiny, but got it smooth enough to check flatness using optical interference fringes, and I think I got it all flat to ~1um or so.

Reference waveform in white and waveform after in yellow. Still some aberration but a decent improvement. Not sure if I am going to give lapping the cores another shot or just let it be. Similarly, I was able to bring the LF -3dB point on the P6021 closer to the spec than in was, but still not meeting the value.

[attachimg=1]

Jitterbit:
Seems an NE5532 would be a lower-noise substitute for the 1458, if you're replacing that.

Amphion:

--- Quote from: MarkL on June 02, 2022, 10:57:35 pm ---Thanks, Amphion, for the great insight.  I will have to look at the core mating on my A6302 that has overshoot.

Did you find it was just the lower core?  Or did you have to do both?

Do you have any before/after waveform captures?

I've also done a fair amount of fiber polishing, but it was many years ago.

--- End quote ---

I apologize I didn't see this until now. I'm not super active on here and I'll have to figure out if there is a way to turn on notifications on topics??

It's been a while now, but as I remember both halves needed lapping, but one was worse?? I wish I had before/after waveform captures, that would have been good to post... I might have some that unfortunately started halfway though the process, not showing how bad it was. I'll try to remember to look on that scope next time i'm in my lab.

Navigation

[0] Message Index

[*] Previous page