Tektronix TCP202 current probe repair - Schematic and suggestions needed

[KrudyZ]

The earlier versions of Tektronix current probes had a home brew looking switch assembly in it made from stamped metal pieces to let the amplifier know if the slider was closed.
I'm not sure if they kept that "great" design in the TCP202, but that's where I would be looking.

[Tantratron]

Here is a status on my attempt to calibrate-repair the TCP202.

First I've completely de-constructed the probe head to observe any visible issue then re-constructed the probe, so far I did not see any problem (electrical or mechanical) which could explain the non-flatness response when sensing a 1KHz square signal (see attached some pictures of the de-construction).

After that, I decided to connect a 100W incandescent lamp to the 230Vac-50Hz grid then probe with my P5210 and my TCP212, see pictures again where everything seems OK.

So if we go back tony initial test, see both pictures where I'm using the tektronix 015-0672-50 calibrator 50 turns current loop, how can we obtain better flatness and rise time speed !

I've not opened yet the termination box of the TCP202 but do you think the circuit has some tuning or calibration to solve the issue ?

[Tantratron]

P.S. regarding the tektronix 015-0672-50 calibrator 50 turns current loop, does anybody knows the schematics or circuit... does the calibrator really generates a super fast rise and fall 1KHz signal from the TDS front panel ?

P.S.2. If someone owns the same tektronix calibrator and same TCP202 connected to a TDSxxx oscilloscope, could it be possible to make the same test which I did earlier to check the resultant waveform ?

[MarkL]

It seems odd that you would get a perfect waveform when the jaws are not locked CLOSED.

In your photo of the disassembled probe, are you showing all the parts?

There should be a coil spring inside a metal sleeve which forces the upper sliding piece forward, and there should also be a tiny ball bearing.  The ball bearing is important.  It puts pressure on the wavy piece of spring metal to make sure the sliding transformer piece is seated firmly against the lower transformer.  It's also the thing that "clicks" when you move the lever to the CLOSED position.  Does your probe click?

Also make sure the wavy spring metal is in the correct orientation.

Screen shot from the TCP202 Instruction Manual attached:  spring metal thing (10), coil spring and sleeve (2) and (3), ball bearing (12).  I'm pointing these out explicitly in case someone else before you lost them.  You may not know they're supposed to be there.

[MarkL]

P.S. regarding the tektronix 015-0672-50 calibrator 50 turns current loop, does anybody knows the schematics or circuit... does the calibrator really generates a super fast rise and fall 1KHz signal from the TDS front panel ?
...

Looking at the compensator output circuit I wouldn't call it "super fast", but it should be faster than what you're showing on the TCP202.

It doesn't seem likely it's the calibrator output but you can verify it's shape with a voltage probe.

Schematic snippet from http://w140.com/tekwiki/wiki/TDS540

[Tantratron]

It seems odd that you would get a perfect waveform when the jaws are not locked CLOSED.

In your photo of the disassembled probe, are you showing all the parts?

There should be a coil spring inside a metal sleeve which forces the upper sliding piece forward, and there should also be a tiny ball bearing.  The ball bearing is important.  It puts pressure on the wavy piece of spring metal to make sure the sliding transformer piece is seated firmly against the lower transformer.  It's also the thing that "clicks" when you move the lever to the CLOSED position.  Does your probe click?

Also make sure the wavy spring metal is in the correct orientation.

Screen shot from the TCP202 Instruction Manual attached:  spring metal thing (10), coil spring and sleeve (2) and (3), ball bearing (12).  I'm pointing these out explicitly in case someone else before you lost them.  You may not know they're supposed to be there.

Yes I did forgot to take pictures of the tiny ball, the spring assembly and metal sleeve... but there were there and later when re-assembling the probe head I did put them all together as shown in the service manual. Of course the initial issue was not resolved !

P.S. regarding the tektronix 015-0672-50 calibrator 50 turns current loop, does anybody knows the schematics or circuit... does the calibrator really generates a super fast rise and fall 1KHz signal from the TDS front panel ?
...

Looking at the compensator output circuit I wouldn't call it "super fast", but it should be faster than what you're showing on the TCP202.

It doesn't seem likely it's the calibrator output but you can verify it's shape with a voltage probe.

Schematic snippet from http://w140.com/tekwiki/wiki/TDS540

Today I've measured the resistive impedance and the inductive impedance of the tektronix calibrator, the values are 42 ohms and 85 microH so I really wondered if maybe the problem could be the calibrator speed (kind of slow 50 turns transformer). What you show is part of the TDS540C front panel circuit which generates the signal output whereas my concern is the external calibrator 015-0672-50 where the TCP202 will sense the current connected to the TDS540C.

So later today I've built a quick and dirty test fixture, see the first picture where I connect the High Voltage output BNC of my PG506 to my Q&D box which is connected to a 50 ohm coax load. At the same time, I sense the voltage output to Ch 3 and display Ch2 the output of the TCP202.

The good news now is that the TCP202 does respond very well with sharp rise and sharp fall at different 50% square signal (100Hz, 1KHz, 10KHz, 100KHz and 1MHz), see attached pictures.

Then I've made a very high sampling rate to show the transient (see last attached picture) which is less than 10ns so this proves the probe works good BUT the problem seems to be a gain mismatch. I've tuned my PG506 to make 5V as shown on my TDS540C so if my 50 ohms load is precise I should get around 100mA but in average I get more circa 103mA or 104mA.

As you see on the pictures, I've removed the top box of the TCP202 amplifier-compensator but I do not understand how to remove the bottom box. It seems to adjust the gain, one needs to access the bottom of the TCP202 box and maybe this will solve the problem.

So either the TCP202 calibrator fixture is partially detuned or my TCP202 needs a gain adjustment to meet the PASS test of the TDS540C.

It would be interesting to have someone post the circuit of the TCP202 to see if this makes sense or if someone could run its TCP202 with the 015-0672-50 calibrator to see the waveform, if if PASS or FAIL

Thank you

[MarkL]

...
As you see on the pictures, I've removed the top box of the TCP202 amplifier-compensator but I do not understand how to remove the bottom box. It seems to adjust the gain, one needs to access the bottom of the TCP202 box and maybe this will solve the problem.
...

The bottom of the comp box is a friction fit.  You can gently wiggle the black piece of plastic that is part of the BNC assembly out of the bottom.  Then you can separate the two boards and undo the insulating tape around the bottom board as needed to get access to the pot.  The coax going into the BNC is a Peltola connector and unplugs if needed.

[Tantratron]

...
As you see on the pictures, I've removed the top box of the TCP202 amplifier-compensator but I do not understand how to remove the bottom box. It seems to adjust the gain, one needs to access the bottom of the TCP202 box and maybe this will solve the problem.
...

The bottom of the comp box is a friction fit.  You can gently wiggle the black piece of plastic that is part of the BNC assembly out of the bottom.  Then you can separate the two boards and undo the insulating tape around the bottom board as needed to get access to the pot.  The coax going into the BNC is a Peltola connector and unplugs if needed.

I'm still not sure to properly understand how to safely access the gain adjustment part. I wonder if there is not a physical issue, see attached one picture showing my partial deconstruction of the compensation box. It seems unless I'm wrong that both half plastic covers have been mounted incorrectly or opposite when comparing to page 33 of TCP202 instruction manual.

When removing the the top cover catches, I see directly the DC offset pot and the de-gauss button which should be the other way. How should I remove safely without any breaking both PCB's, how to unplug from the Tekprobe to re-assemble correctly the all unit ?

What is wierd, the previous post https://www.eevblog.com/forum/repair/tektronix-tcp202-current-probe-repair-schematic-and-suggestions-needed/msg1663601/#msg1663601 shows an installation similar to my probe but does not corresponds to what tektronix instruction manual states...

By the way, the page 33 does mention that the gain of the probe must be with +/-2% for the TDS oscilloscope compensation routine to work so I guess that was my initial issue (FAIL compensation routine) and not a flat response issue.

[MarkL]

I'm not seeing the difference between your probe, the manual, and the previous photo you reference.  It doesn't look opposite to me.  Can you mark up the photo and point it out?  Maybe I'm just being dense.

Below is my comp box disassembled.  The black plastic that forms the BNC assembly is just pushed into the bottom cover.  It's a very tight fit, but it does come out.  It is not glued or fastened by any other means.

If you're working on the two boards, you can also unplug the Peltola connector to untangle the cables a bit.  The white connector can also be unplugged, but it too is a tight fit.  The brown piece of plastic in the photo is just insulating tape around the bottom board that can be undone partially as I've done, or remove it if you want.  Just make sure you put it back.

But I think you should be able to get to the gain pot without removing the bottom anyway by pushing the top board out of the way.  You can probably shove the insulating tape to one side or poke a small hole in it to get to the pot.

[Tantratron]

Hi Mark, many thanks for the guidance where now I've fully de-constructed the compensation box, cut part of the insulating tape to access the gain potentiometer per guidance.

As you can see on the 2 pictures, my PG506 with my homemade current loop 50 ohms fixture along with the TCP202 trimming pot got it to a better gain calibration. Not perfect as a real test gear but then I've run the calibrator... the test is PASS plus we can finally see the calibrator was not supposed to generate an ultra-wide flat square (it is a 50 turns transformer).

Again many thanks where I'm this time on the road to discover, use in my plasma test lab the TCP202 current probe versus my previous P6021 and P6022.

Albert

[MarkL]

Great!  Glad you got it adjusted and it wasn't a more complicated problem.  It's been a long road for you to find a working TCP202.

Thanks for the heads-up on the 50-turn calibrator.  I've always used a load resistor, but now we know what to expect from the calibrator.  It doesn't seem like a very good accessory if the goal is calibration.

[MarkL]

Glad I found this thread.

I also have acquired a TCP202 in good condition. The probe "operates" but it doesn't "work".

I did test it with DC 1A and the display seems accurate. Also tested with DC 50mA and it seems accurate, but I need to degauss to go back to the original 0 value - after the DC 50mA test the trace is about 10mA higher then 0 every time. This may be normal, not sure.

The degauss works and so does the coarse zero (on the back of the probe head) and the fine 0 on the side of the plastic box at the BNC.

The sine wave in AC looks ok, but it is smaller than it should be. At 10KHz is about 15% smaller, at 20KHz is the worst, 60% smaller and it recovers at 40KHz to 15% again, staying around there to the MHz range. The square wave response is very ugly.

I measured the transformer in the head and I got the following values (compared to the previously posted measurements):
+3V - -3V = 276 ohm - 176 ohm
+3V - HALL1 = 183 ohm - 93 ohm
+3V - HALL2 = 179 ohm - 92 ohm
-3V - HALL1 = 189 ohm - 141 ohm
-3V - HALL2 = 182 ohm - 141 ohm
HALL1 - HALL2 = 110 ohm - 72 ohm
COIL - COIL = 6.5 ohm - 6.5 ohm

These are with the small PCB in the head disconnected. Are these still in range?

Also, the electronics in the plastic box looks good, no visibly fried parts, and no sign of any repair.

Attached some pictures. The load is a 100 ohm metal film resistor.
Any help would be greatly appreciated. Thanks.

It's been a while since that post, but I had a chance to take my probe head apart and I made some measurements.  I got the following values with the 6-pin connector disconnected and the sensor cold (after warm-up the Hall readings are different by about 10 ohms).  There is a part number on the side of the sensor head "120-2032-XX".  My measurements in [...] :

+3V - -3V = 276 ohm - 176 ohm [347]
+3V - HALL1 = 183 ohm - 93 ohm [211]
+3V - HALL2 = 179 ohm - 92 ohm [227]
-3V - HALL1 = 189 ohm - 141 ohm [246]
-3V - HALL2 = 182 ohm - 141 ohm [259]
HALL1 - HALL2 = 110 ohm - 72 ohm [154]
COIL - COIL = 6.5 ohm - 6.5 ohm [6.5]

Your Hall sensor is different by a fairly large margin.

I can also say that the TCP202 Hall sensor uses +/-2V instead of +/-3V, measured at the 6-pin connector.  This table, originally posted by user vtp, was for an A6302.  And also different from the A6302, there are two 100R resistors in parallel on the -2V input on the TCP202 sensor head.

On the degauss question... I also find that after measuring large DC currents it can leave the core slightly magnetized and show a small offset.  It's normal.

[vtp]

I can also say that the TCP202 Hall sensor uses +/-2V instead of +/-3V, measured at the 6-pin connector.  This table, originally posted by user vtp, was for an A6302.  And also different from the A6302, there are two 100R resistors in parallel on the -2V input on the TCP202 sensor head.

Yes, the table I made was for A6302 head but the measurements I provided in the earlier post in this thread are from my TCP202. Looking at the schematics of the box, the head voltages are supplied directly from the probe interface +-5V to the head through a pair of 220R resistors.

With 276 ohms between +-3V pins and 220R resistors from both ends to the +-5V supply I get 3.85 volts for the HALL element operating voltage. Close enough to your +-2V (=4V) measurement.

Perhaps there are then different revisions of the probe - I do not have any 100R resistors in my probe head circuit board. There however are 2 paralleled 39R2 resistors and a 1u8 inductor in series with the coil circuit.

Just to give a reference of a broken probe head, here are measurements from one A6302:

+3V - -3V = 1834 ohm
+3V - HALL1 = 1701 ohm
+3V - HALL2 = 1987 ohm
-3V - HALL1 = 347 ohm
-3V - HALL2 = 626 ohm
HALL1 - HALL2 = 470 ohm
COIL - COIL = 5 ohm

[MarkL]

Thanks for the clarification.

I took apart my other TCP202 and measured it.  Below are the measurements plus the others posted.

My TCP202 #1 is newer, original purchase date unknown, and TCP202 #2 was purchased new 20 years ago.  TCP202 #1 is marked "04 21 42 61" inside the potting compound, and TCP202 #2 is marked "99 21 18 34".  I could believe the "99" is 1999, and maybe "04" is 2004.  Just a guess.

Both have the 100R resistors on a tiny PCB the female connector is mounted on.  They're easy to overlook; photo below.  The resistors are consistent with the diagram in the old "Tek Probe Circuits".

Both also have the 39R2 resistors and 1u8 inductor.  There's also a toroid wound as a CM choke that the two coil leads pass through before going to the resistors and inductor you mention.

  Readings in ohms
  "+/-3V" is actually +/-2V on TCP202

                  vtp     esc     MarkL   MarkL   vtp
   From    To    tcp202  tcp202   202#1   202#2  ex6302
  ------ ------  ------  ------  ------  ------  ------
   +3V    -3V      276     176     347     287    1834
   +3V    HALL1    183      93     211     202    1701
   +3V    HALL2    179      92     227     200    1987
   -3V    HALL1    189     141     246     234     347
   -3V    HALL2    182     141     259     232     626
   HALL1  HALL2    110      72     154     174     470
   COIL   COIL     6.5     6.5     6.5     6.8       5


How did your A6302 die?

[vtp]

Both have the 100R resistors on a tiny PCB the female connector is mounted on.  They're easy to overlook; photo below.

How did your A6302 die?

You are right. Thanks for the reminder on those resistors. I completely forgot about them. They also seem to be in A6302 heads as I have drawn the resistors on my notebook.

I can't really recall but the broken A6302 I may have bought as non-working since I had one new element to replace the broken one with.

I also have an old P6302 which died spontaneously.

Measurements of the hall-element signals give a clue about what failed. In the A6302 example the probe head has a problem with +3V pin connection.

In the broken P6302 the fault is even more clear, HALL1 pin has disconnected itself completely from the element as all HALL1 related readings give infinite resistance.

Here is a nice basic introduction to Hall-sensor:

https://www.electronics-tutorials.ws/electromagnetism/hall-effect.html and a picture from that borrowed below:



As in the picture +-3V, or really the current supply pins, provide current to the element and Hall-voltage from the diverted charges appears at pins HALL1 and HALL2. Depending on what the resistance measurements show it is fairly easy to determine if the probe head is ok. Those signals are all available on the A6302 connector so the element can be checked without disassembling the probe. With TCP202 some disassembly is needed, at least to get to the wire to board connector in the compensation box. Of course there is the probe head PCB in between but that being faulty would be very good luck.

[esc]

As for my measurements, after taking the probe head apart, I realized that my probe head has a newer transformer, designed for the TCP305 PN  120-2032-XX.
The transformer has 04-20-44-40 on it. Maybe that's why the discrepancy is there. 

[MarkL]

Your 120-2032-XX is the same part number as my TCP202 #1.  Within Tektronix part numbers, the NNN-NNNN are the main part number, and the -XX is usually the revision.  According to the TekWiki page, the last number could also be a quality or a speed grade.

But given that these both came from TCP202 probes with the same ratings, I think it's less likely that the same major part number would have such different resistance readings, and by implication that the problem is elsewhere other than the Hall sensor.

So at the moment I think the Hall sensor is the prime suspect.  However, I can't come up with a good explanation of the strange frequency response.  Maybe the lower impedance has affected parts of the active circuitry that it's connected to.

Instead of just resistance, I will see if I can come up with an out-of-circuit dynamic test or two so you can compare to your Hall sensor.


vtp: Thanks for the nice pictorial on the Hall effect.

[MarkL]

Attached is some frequency characterization for the Hall sensor.  And although you don't seem to be having a problem with the higher frequencies, I also did the coil up to 1MHz since I already had the sweep set up.

The Hall sensor was supplied by +/-5V with each leg in series with a 220R resistor, as previously described by user vtp.  I also confirmed I have the 220R resistors on both my probes.  The output voltage was measured differentially across the two outputs.  The outputs have an offset above ground of about +236mV, and between them about 4.7mV.  No load was placed on the output except the probing.  A test of a 1k load did not affect the output significantly, so it seems the Hall output is a fairly low impedance.

Likewise, the coil output was measured differentially across its two outputs, but because it's floating it won't matter if you do it single ended.  The coil output was loaded with a 50R resistor.  I don't know if that's the actual impedance seen by the coil, but at least it will serve as a basis for comparison.

The signal generator was set to 16Vpp with offset 0V.  One leg of the output was looped through the probe transformer halves and terminated in a 50R load.

I don't have a Bode plot on my scope, and I didn't feel like writing one.  So, what you see horizontally is a full sweep at the noted start and stop frequencies, and then a closer look at the start, stop, and 3dB point.  The scope has an RMS ratio function between two traces, so it was easy to find the 3dB point by spinning the horizontal delay until the ratio was down 3dB from the maximum.  The sweeps include 20kHz, which is where you were having the most severe drop-off.

The DC response of the Hall sensor was 44.4mV/A, measured at 1A.

Hopefully this info will help you evaluate your Hall sensor.

One thing you should check, if you haven't already, is that your TekProbe interface has all the +/-12V and +/-5V supplies working. You could also verify the Hall +/-2V supply at the head.  But given the lower impedance of your sensor, it's probably going to be low.

Also make sure the Hall and coil are isolated from the sensor's metal shell.


EDIT#2:  Ever have one of those days?  I discovered my connection to the 50R load was bad throughout the AC measurements.  I've removed all the previous screen shots and re-posted updated full sweeps for the Hall and coil, both 100Hz to 100kHz.  I didn't mention it before, but it should be obvious the frequency on the X-axis is logarithmic.

I'll just tell you the numbers derived instead of posting all those detailed screen shots again:

         Frequency at reduced signal
           level from maximum (Hz)

           -3dB    -6dB   -10dB
          ------  ------  ------
  Hall      2.3k    4.0k    8.1k 
  Coil      2.3k    1.3k     775 

  Hall @ 100kHz = -21dB
  Coil @ 100 Hz = -24dB

The sweep envelopes now look right and the numbers make a lot more sense.

[esc]

Thanks MarkL for the info.

Since your transformer in TCP202#1 and mine are both the same part 120-2032-XX, the resistance measurements should be close, and they are not. So I agree, it is likely that the Hall sensor in my transformer is the problem. The interesting part is that my probe measures DC accurately. Maybe the small resistance between the +3V and -3V and the two 220R series resistors causes a too small supply voltage for the Hall sensor, and even though it "works", but causes the anomaly at 20KHz?

[Tantratron]

Maybe the small resistance between the +3V and -3V and the two 220R series resistors causes a too small supply voltage for the Hall sensor, and even though it "works", but causes the anomaly at 20KHz?

If you have an anomaly circa 20KHz, this could be related to the electronics amplifier and mixer, namely the 2 responses (current transformer and hall sensor) which are fused or hybrid. If you look the service manual of the A6302 and AM503A, it mentions the cross over region. It seems clear visually that the TCP202 probe head is same technology or design than the A6302, maybe check more circa pages 3-6 of the manual. In the case of the TCP202, the compensation box contains the amplifier and probably the crossover electronic management.

System Behavior in the Crossover Region: When the frequency of the current being measured is in the vicinity of the crossover region (approximately 20 kHz), part of the measurement appears at the Hall device and the other part appears across the coil. The waveforms in Figure 3 8 illustrate how these two components combine to make the total measurement. Waveform (a) shows the current being measured. The fre quency of the current is in the crossover region of the probe. Waveform (b) shows the low frequency bucking current produced from the Hall device output. Waveform (c) shows the high frequency component that is induced directly into the coil from the current source. Waveforms (b) and (c) are summed together in the probe coil and amplified by the AM 503A amplifier. The amplifier output is connected to a measurement instrument, such as an oscilloscope, where a waveform similar to waveform (d) appears. Waveform (d) conforms to the original signal shown in waveform (a).

[Mrnt]

I picked up a used TCP202 and I am not sure if this is me being thick, but it only seems to measure positive DC current ie it seems to accurately measure dc when flowing in the direction of the arrow on the probe but if I reverse the current direction, the probe only measures approx -700mA no matter what current is measured. Is this normal?
I have not tried AC yet and nothing appears scorched inside the head.

[MarkL]

The probe should have a symmetric output for AC and DC signals.  I don't think it's likely to be something in the probe head.  Maybe one side of the Hall driver is not working.

You could start by checking the power rails in the comp box.  A convenient place is on the little white ribbon that connects the TekProbe connector to the main circuit board.  The ribbon pads on the circuit board are numbered starting from the end where the probe cable enters:

  1: GND
  2: -15V
  3: -5V
  4: +15V
  5: +5V

Taking a look at a high frequency AC signal (like >1MHz sine) could also be enlightening, since the Hall driver should be out of the picture at that point.

EDIT: Fixed description for pin 1.

[Mrnt]

Thanks Mark for your prey response! It look like the voltage while not exactly a you listed are within the +\- 2% pc for the Tekprobe connections:

At the white ribbon
0
-14.9
-5.16
14.82
4.96

Trying at 1.5Mhz AC with approx 10v pp into 50 ohm the current waveform was fairly smooth, correct amplitude and symmetric about 0 mA. Unfortunately I do not have a higher current source.

I notice that there is a different  opamp on this board from the other pictures I have seen - an OP275.

Any thought as what to look at next?

[MarkL]

Those voltages are fine.  It was worth checking to make sure one of the rails wasn't completely missing.

My older probe has OP27 and MC1458, but the newer one has OP27 and OP275 like yours.

Have you done the coarse balance procedure on the underside of the probe?  What kind of a range do you get for the coarse adjustment?  Does it also bottom out at -700mA?  For comparison, my newer TCP202 can be adjusted from -7.9A to +6.1A (or in volts: -0.79V to +0.61 into 50R).  My older probe has a much lower range of -1.6A to +1.1A.

If it does bottom out, I think a general approach would be to start working backwards from the Hall power amplifier which includes 1/2 of OP275 and the BCP68/69 SOT-223 driver pair (see vtp's earlier post).  One approach would be to vary the coarse adjustment back and forth to find who's responsible for the amplifier falling out of operating range.  Possible probe points would be the emitter and base of the BCP68/69 pair, and examining the OP275 output and inverting and non-inverting inputs.

A few other simple voltage checks might also be in order.  On the white connector on the main board, they are numbered 1 to 7 (same order as the ribbon connector):

  1 - 0V (Black to bottom board)
  2 - Varies with DC signal/offset (White to bottom board)
  3 - Hall power -2V
  4 - Hall power +2V
  5 - Hall 1  0.206V
  6 - Hall 2  0.259V
  7 - Center tap on course adj. pot, varies +/-2V

  5 and 6 will vary depending on the specific Hall device, but should be close to each other.

I don't have any specific suspicions in mind related to this connector; just gathering data.

Without a schematic and without resorting to shotgun parts replacement, you're probably going to need to do some reverse engineering.  I haven't done any of that for the TCP202, but I can help with voltage comparisons from a working probe.

[Tantratron]

I picked up a used TCP202 and I am not sure if this is me being thick, but it only seems to measure positive DC current ie it seems to accurately measure dc when flowing in the direction of the arrow on the probe but if I reverse the current direction, the probe only measures approx -700mA no matter what current is measured. Is this normal?
I have not tried AC yet and nothing appears scorched inside the head.

Have you press the degauss button as a start ?