Tektronix TCP202 current probe repair - Schematic and suggestions needed

[AndersJ]

The probe output is constantly at 2 Volts.
Possibly the internal negative voltage is gone.

I have double checked with other probes and other power supplies.
I believe the problem is in the probe.

Does anyone have a schematic diagram or suggestions?

Thanks,
Anders J

[tautech]

Have you got the manual:
http://download.tek.com/manual/070954202.pdf

Can't seem to find any repair threads on them. 

[AndersJ]

Yes, I have the users manual.
No schematics unfortunately.

[lukier]

Do you have any pictures of the internals, both the probe and the plug end?

Maybe this way we could help. I don't have TCP202, but I have A6302 with AM503 and AM503B amplifiers, so maybe there are some similarities. AFAIR TCP202 doesn't have the 50 Ohm amplifier with ranges and whatnot that standalone units use and terminates into 1M straight into the scope and relies on the scope frontend on AC/DC and ranges.

[AndersJ]

Here is picture of the opened compensation box.
The SO8 next to the OP27 is damaged, and the part number is partially gone.
The part number seems to end with "58", possibly prefixed by a "4".
Could be a LM1458.

Does anyone have a TCP202 they could open and verify this?

[Giaime]

Hi!

I have one: you can find a pic attached. I confirm it's an MC1458 

[AndersJ]

Many thanks for your confirmation.
/Anders

[KrudyZ]

Looks like a few of the SOT-23 parts are fried as well.

[AndersJ]

I had not noticed those before, thanks.
I have now found 2 that need attention.

Giaime,
can you take a close up photo so I can identify the two SOT-23 parts?

[Giaime]

Hi! Here are some additional photos, hope they help:

[AndersJ]

Thanks Giaime,

I cannot tell for sure how the SOT-23 next to the MC1458 is marked.
It looks like it could be GJ1 or GJ or perhaps 6J1 or 6J,
Can you try to see what it says, perhaps using a magnifying glass.

I have not been able to locate a SMD part marked GJ1.
I also tried 6J1 but cannot find that either.

The closest one I found is 6J which is a N-ch fet 2N4391
or
a some sort of resistor network.


/Anders

[Giaime]

Hi!

1AM is easy, should be MMBT3904.

The other one looks to me 6J1, could be MMBF4391 but it's not 100% guaranteed.

[vtp]

From my notes on TCP202 disassembly and reverse engineering for repair:

6J = MMBF4391
1AM = MMBT3904
26 = MMBD1203

Both JFETs are used during degauss. One is to isolate OP27 from the power amplifier (1/2 1458, BCP68/69), the other appears to enable degaussing oscillator (1/2 1458).

[lukier]

From my notes on TCP202 disassembly and reverse engineering for repair:

Would you mind sharing some of this info? Some pictures from the back, maybe simplified schematics?

I'm asking, because I don't like my AM503/AM503B amplifiers and I think it is silly to replicate scope frontend (ranges, AC/DC) and 50 Ohm amplifier there when there is one in the scope. Don't know why Tektronix went for it even in TCP300A (although allegedly there they've stopped using custom hybrid ICs).

So I always wanted to somehow mod my A6302 probe to be more like TCP202 (although with LeCroy's interface preferably), but there is very little online on these devices and signal conditioning.
I know very briefly how the DC probe works, studied old AM503 schematics a bit, but I'm not confident enough to risk damaging my A6302

[vtp]

Would you mind sharing some of this info?

My notes are four pages written and hand drawn schematics and I do not have scans of those. Not that they would do much good as the whole thing is a bit scattered around those pages.

But the operating principle is similar to other current probes. There is the head with its HALL-element and coil in the same magnetic circuit. HALL-element signal is differential input to the OP27 with 1k60s as inputs and 20k0s to GND/output. Then it goes trough 475R to one JFET that can disconnect the OP27 during degauss.

If the sensed signal is allowed to continue it goes to input of 1/2 1458 which has PCB68/69 discrete output stage driving the coil in the probe head. This is used to compensate the field sensed with HALL-element. The other end of the head coil is filtered and brought out then directly to the probe output BNC-connector (at the bottom board). So the coil is in series with the PA and probe output. Gain of the PA looks to be set with 10k0/2k00 resistor divider and it is non-inverting. OP27 output is brought to the 1458 inverting input trough the said 475R and the JFET.

The rest of the probe electronics are degaussing oscillator and power supplies. It uses +-15V and +-5V from the tekprobe interface.

TCP202 head differs from others regarding few parameters, the coil DC-resistance is higher at about 6.5 ohms to make driving it with a smaller amplifier easier. A6302 and TCP312 coil DC-resistances seem to be around 2 ohms. These are not very accurate measurements, they were taken with a Fluke 87 from the element.

...

To the OP, I do not quite understand how someone has managed to blow up those two JFETs and the 1458. The probe head may have touched live high voltage wire but even then it should have been limited to high current going trough ground shields. Unless there has been a breakdown in the head itself when you are looking at about €1200 repair (flat exchange fee from Tek - or at least was several years ago) due to unobtainium head.

It might be a good idea to check the head.

If you look at the head element from the connector end the pin out should look like

Code: [Select]

_____
|   |
|   |
|OOO| pins top row
|OOO| pins bottom row
-----

The pins being HALL2, HALL1, +3V on the top row and COIL, -3V, COIL on the bottom row from left to right. That +-3V is voltage in A6302, TCP202 may have different hall operating voltage but I will leave it like that here.

Resistive measurement between pins:

+3V - -3V = 276 ohm
+3V - HALL1 = 183 ohm
+3V - HALL2 = 179 ohm
-3V - HALL1 = 189 ohm
-3V - HALL2 = 182 ohm
HALL1 - HALL2 = 110 ohm
COIL - COIL = 6.5 ohm

Yours may differ from these several tens of ohms. If something reads hundreds of ohms different or around kohm or open then you have a damaged head.

The same way it is easy to check quickly if an A6302 is ok as those signals are all brought to the cable connector.

[vtp]

Just digging the notes again, TCP202 head HALL-element operating voltage comes from +-5V through 220R resistors in series with both voltages.

[lukier]

My notes are four pages written and hand drawn schematics and I do not have scans of those. Not that they would do much good as the whole thing is a bit scattered around those pages.

But the operating principle is similar to other current probes. There is the head with its HALL-element and coil in the same magnetic circuit. HALL-element signal is differential input to the OP27 with 1k60s as inputs and 20k0s to GND/output. Then it goes trough 475R to one JFET that can disconnect the OP27 during degauss.

If the sensed signal is allowed to continue it goes to input of 1/2 1458 which has PCB68/69 discrete output stage driving the coil in the probe head. This is used to compensate the field sensed with HALL-element. The other end of the head coil is filtered and brought out then directly to the probe output BNC-connector (at the bottom board). So the coil is in series with the PA and probe output. Gain of the PA looks to be set with 10k0/2k00 resistor divider and it is non-inverting. OP27 output is brought to the 1458 inverting input trough the said 475R and the JFET.

The rest of the probe electronics are degaussing oscillator and power supplies. It uses +-15V and +-5V from the tekprobe interface.

TCP202 head differs from others regarding few parameters, the coil DC-resistance is higher at about 6.5 ohms to make driving it with a smaller amplifier easier. A6302 and TCP312 coil DC-resistances seem to be around 2 ohms. These are not very accurate measurements, they were taken with a Fluke 87 from the element.

Thanks! This is very appreciated.

Now I get a better idea what's inside TCP202. It seems radically simplified from AM503A schematics (the most modern amplifier schematics available). There hall signal feeds MC34081P via 475 Ohm resistors, with non-inverting input grounded with 47.5K and inverting having 47.5K in the feedback loop + some RC filtering. Output from that goes via JFET (degauss disconnect) and feeds another MC34081P that drives multi stage totem-pole amplifier, with darlingtons doing the final push into the probe transformer. On both opamps there are input signals to adjust the offset, HALL_PREAMP_DC_OFS and PWR_AMP_DC_OFS, while TCP202 has only one knob. Degauss signal comes from amplified DAC output.

In AM503A the second transformer output goes to the attenuatior and the rest of these duplicated scope frontend stuff. I'm glad to hear that in TCP202 it goes straight to the scope, without any termination inside the probe or similar stuff.

What confused me in AM503A is that there is this FEEDBACK_LOOP signal that comes from Tek's custom made attenuator module into the line bettwen Hall OpAmp and PA OpAmp, adding something there for some reason.

Now I might try some DIY, but still very careful, especially with this power amplifier output and fragile transformer

[vtp]

In AM503A the second transformer output goes to the attenuatior and the rest of these duplicated scope frontend stuff. I'm glad to hear that in TCP202 it goes straight to the scope, without any termination inside the probe or similar stuff.

I was not very clear about this. There is some filtering and termination and even one trimmer in the probe box between coil output and BNC connector. So the coil comes to the probe box in a coax cable, goes through some filtering and four parallel termination resistors one of which has a trimmer in series probably to adjust the output amplitude/gain.

Let me try to draw:

Code: [Select]

                         ---R---
                        |       |
------L----------L---------L------------------------------ output to BNC/scope
   |     |    |     |                  |    |   |    |
   R    R     R    R                   R    R   R    R+trimmer
   |     |    |     |                  |    |   |    |
   ------      ------                 -----------------gnd
      |          |
     C           C
      |          |
      -   gnd   -


I do not have component values for those except that the last L in chain is 10uH and has two 68R1 resistors in parallel to it (only one drawn above). Left is input from the coil.

[lukier]

Maybe it is similar to the termination used in Tek's AC probes, like shown here:
http://www.davmar.org/TE/TekConcepts/TekProbeCircuits.pdf
(pdf page 69 and 70)

[AndersJ]

Hi all,

I have replaced the visually obviously damaged parts.
After that the probe behaviour improved, but it was not fully functional.
The "undamaged" OP27 was uncomfortably warm so I replaced it too.

After that, the probe seems to work.

To the OP, I do not quite understand how someone has managed to blow up those two JFETs and the 1458.

During the repair i noticed that the grey collar that mechanically locks the probe in place, was rotated 180 degrees.
This means the probe will connect "backwards", possibly reversing power inputs, thus blowing the probe.
This is normally not mechanically doable but with some ingenuity even the impossible might have happened.
I don't know if this is the right explanation, but efter repairing it, I do not want to try that.


Thanks for all help!

/Anders J

[Pic]

Thanks' to these posts I was able to fix my broken TCP202   I did a short when I removed the probe from an unisolated 400V wire. 

I want to share some hints for everyone who did the same mistake.
I exchanged the two SOIC-8 opamps. But it still did not work. Degauss, offset was working fine but the offset adjustment was quite sensitive. I figured out there was no voltage supplied to the HALL. The two 220R and two 1k resistors in series to the supply, next to the connector have gone high impedanze. I replaced them and now - everything is working again.

Many thanks' to you folks!

[Tantratron]

Hello,

So far I've been using for most of my projects AC current probes P6021 and P6022 but lately I have new specific projects requiring DC-AC current probes.

What is the global return of experience using these TCP202, are they quality, calibrated stable with little failure if used correctly ?

Thank you, Albert

[MarkL]

...
What is the global return of experience using these TCP202, are they quality, calibrated stable with little failure if used correctly ?

I think they are great probes.  I have two of them, one is about 20 years old now purchased new, and the other one used about 2 years ago.  I've had no performance issues with either of them.

The only failure I can recall is that the older one developed a crack in the bottom plastic surrounding the core about a year ago.  It hadn't quite reached the point of breaking off completely, and some plastic welder fixed it with no further cracking since then.  I've seen several photos of used probes with this bottom piece missing, so it makes me believe it's a common weak point in the design.

You might already know this, but being a powered probe, it does need a TekProbe II connection on your scope (or equivalent adapter), or an external power supply such as the Tek 1103.

[Tantratron]

I've finally purchased one TCP202 probes from eBay in Austria, it came with a calibrate done by the vendor last week but for some reason, it does not work here in France.

When I connect the probe to my TDS540C, the corresponding channel will shift automatic to 50ohms input impedance, it will display A (Amperes) instead of V (Volts).

However when connected to a ballast source running at 250KHz, the signal is always zero plus it seems impossible to DC balance so no idea if the degauss works. The DC bias output stays always circa 20 mA flat.

I've then tried connect the probe to my spare tek 1103 power supply adapter then my tek 2465, same story with no output whereas it should display circa 1.5A peak sinewave generator at 250 KHz.

Do you know if the probe termination box circuits has some plug connector who could have moved or got loose during the shipment between Austria and France ?

Any suggestions or guidance, thank you

[MarkL]

I'm assuming you mean you can't get it to zero with the knob on the compensation box.  Did you try the coarse balance adjustment on the underside of the probe body?

See page 3 here:

  http://w140.com/tek_tcp202_instruction.pdf

Also sounds like you're not getting any signal out of it.  Can you see any deflection with just a test DC current of 1A?