Tek p5205 HV differential probe teardown. BTW, what are the red and brown wires?

[romantao]

@Fluxed Matter, I think i was overbid by you damn!

I'll keep waiting for my right to buy one myself!

[Fluxed Matter]

@romantao, I was out bid on 3 others before winning this one!

[romantao]

Hi!

Anyone knows a reference for the probe switches ( teh ones for attenuation/ overrange and bandwidth)
They have this inscription on the side 0113 201, but from it found nothing relevant...

Two of mine are kind of funky...

regards

[128er]

Hi!

Anyone knows a reference for the probe switches ( teh ones for attenuation/ overrange and bandwidth)
They have this inscription on the side 0113 201, but from it found nothing relevant...

Two of mine are kind of funky...

regards

I had the same problem. Couldn't find any switches like them. But I repaired all switches at my four probes.

Remove the grey cap from the switch by pulling it off. Then carefully remove the top half from the switch by bending the plastic latches on both sides. Remove the white top half slowly to avoid a mess inside the switch (spring and steel wire for the latching function). Be carefull in the end, that you put all back together as it was. You can put the red piece backwards in the white top cover. And you'll not be able to close the casing of the switch.

Now the important part. You have two springy metal pieces inside. They get actuated by the red piece and shorting the contacts together. They wear out over time, because their shape gets flatter. The result is bad contact and all sorts of weird signals comming out of the probe.

I have tried to illustrate that in the last picture, where you can see the cross section of the contact pieces. The one that is worn out, is less in width. The challenge is, to bring the contact piece in it's old shape. I have done that with my fingers. I repaired twelve switches. Several switches I had to open twice. Because it's a bit tricky to bend the pieces in the right way, so that it makes good contact.

Man, I hope that is understandable. Very difficult to describe such a fiddly thing in a foreign language 

P.S.:
Good luck! 

[romantao]

Eheh! I know what you mean when you lack some technical terms in a foreign language!

I manage to fix them

I left the probe on during all day after compensating its gain and offset (internal trims).

The differential gain is accurate now.

However ... (there's always something wrong...) i'm noticing that under 500x the offset keeps running up once i close the case. That's not happening with the 50x. Even if i under-compensate while i have open case... as soon as i close it the offset keeps rising till 400mv approx.

Is this a normal behaviour and i Just need to trim it using the compensation box? This way, depending on which attenuation range i'm working, i'll have to do different offset of the 50x and 500x range.

Noticed that the lm317 and 337 get a little warm but they are putting around + - 5.1V respectively

[Shock]

I have a Tek P5205 and I'm thinking about converting it for an analog 100MHz scope with 1MR input.

Is the general consensus?
Blue +15V
Black -15V
Brown won't be used in my situation
Red is the zero compensation for low voltage measurements.

Since I don't have the power supply for it I was thinking of just of mounting a DC socket and the 10K zero compensation pot at the LV end of the probe which will need some new holes made.

Is there any potential pitfalls anyone can see in doing this, will the power need shielding or twisting or interfere with the LV signal? It's got to be better than running it up to the scope, right?

[128er]

I have a Tek P5205 and I'm thinking about converting it for an analog 100MHz scope with 1MR input.
Is the general consensus?
Blue +15V
Black -15V
Brown won't be used in my situation
Red is the zero compensation for low voltage measurements.

Yes!

Since I don't have the power supply for it I was thinking of just of mounting a DC socket and the 10K zero compensation pot at the LV end of the probe which will need some new holes made.

You mean, fitting a DC socket and the 10k pot in the probe case? Why not. But the probe needs symmetrical +/- 15V. Or do you want to mount two DC sockets? I am not able to understand it at the moment

[Shock]

I was thinking about drilling a hole for the DC supply (3 pin socket for +15V, -15V and GND). The DC supplies GND pin (not mains earthed referenced at the supply itself) then connects to the LV signal cables coax GND on the PCB (which is mains earth referenced at the scope). Then for the zero compensation put a 10K pot across the 15V rails and connect the Red to the wiper. Is that correct?

[EvanShultz]

Cheers! First post!

Hi Shock, yes that's the correct connections based on the schematic posted earlier. It should work fine.

I just finished tracing the circuitry in the big probe box. I need to simulate a few things to make sure it all works, and I want to confirm the circuitry in the small box that goes into the scope, but so far it all looks correct and makes sense. Stay tuned...

[abulafia]

Hi guys,
Just for the interested parties here is R/E schematics from a P5205 (non A).
Also to answer the buttons question, they are almost identical to Digikey item https://www.digikey.com/product-detail/en/c-k/PVA1EEH11.2NV2/401-1132-ND/417713 (same manufacturer)

Best regards,
Bogdan

[walter76]

Very useful post, it helped a lot in fixing a P5205 that was scrapped.

The probe was marked as no-good and its showed a huge offest (around -2V) that could not be recovered by acting on the adjusting trimmer.
It took some time to discover that the switches of the DG444 (U150) were no longer isolated each other: acting on R160 & R165 was partially successful in recovering the offset, until the scope suddenly switched to another scale!

I removed U150 (obsolete) and I used a small 12V relais I found in my junk box: one switch to short R145, the other one to select between R160 & R165 according to the x50/x500 scale.
Just move R155 to be in series with the wipe of R165, then most of the lands for U150's pins can be conveniently used to connect the relais.

I hope this could help somebody else having the same issue.

Walter

[Dbldutch]

Just for kicks, while trying to understand how the various differential probes work, I used LTspice to model the input attenuation using the schematic provided by abulafia in an earlier post.


Interesting to see that the higher frequency response is probably used or can be used to extend the bandwidth.

I'm wondering why Tek used two identical 18R21/10-90pF in parallel???

Enjoy!

[Hydron]

At anything higher than a few MHz the model would need to include input cable parasitics (L&C) and source impedance to get an accurate model of the frequency response. As the leads can move around (thus changing the parasitics) the response will be all over the place well before 100MHz, let alone 1GHz.

[Dbldutch]

Thank you Hydron for the comments, and I agree, but that was not my intention.

I was interested only in the rather unusual front-end section and wanted to get a feel for the three things that stood out for me in this design. Tek engineers generally do that with good reason, and I wanted to try to understand why.

First the relatively low parallel HF capacitors of 2.2pF, (only 1.1pF in total) which makes the front-end much more sensitive to parasitic or stray capacitance. In all the other designs I have found a minimum of 4pF in total, which is about a factor 4 more and so is 40x the typical 0.1pF for parasitic or stray's instead of only about 10x more.

Second, the attenuation factor due to the 4M vs 6K81, which drives the rather odd LF compensation circuit with the strange, for me at least, dual parallel trimming section with R/C networks. In simulating the circuit, I found that the value of the LF 60.5pF I arrived at is extremely sensitive (I also used MC analysis), and the whole front-end is not flat. All the other attenuation designs I looked at can be easily calculated, and when simulated are flat.

Third, the interesting circuit to adjust the CMR (2x 100K and a 25K trimmer). In all the other designs I found, the engineers use a 100-500 Ohm trimmer between ground combined with the two attenuation resistors (here the 6K81), which not only adjusts the CMR, but also the attenuation factor. I like the Tek method better, so I'll use that for my own design.

I'm looking forward to get more comments about of this interesting circuit

Enjoy!

Paul

[Tantratron]

This thread is quite old but maybe since then some of you do have a document explaining how to service manual or calibrate the P5205 which would be similar to the P5210 which I'm discussing in this thread https://www.eevblog.com/forum/testgear/repair-tektronix-p5210-high-voltage-differential-probe/msg3220840/#msg3220840

Thank you in advance, Albert

[Olexandr]

I want to repeat the scheme of this device for my own use, but the scheme lacks the denominations of almost all capacitors, based on this I want to ask everyone who has such a device to help me, you need to measure and indicate which capacitors are there.

[richipedia]

Just in case somebody needs it: here are some nice pictures of the P5205 PCB's top and bottom sides.

[minifloat]

(......)
A few pictures of my PSU . . .

A year ago or so, I got a pair of P5205 HV probes fairly cheap, without noticing the TekProbe supply needed. Yikes!

I did not want to cannibalize the probes - being inspired by @128er, I made my own.

Hacktronix
- simple ±12V
- no ±5V nor offset rubbish
- some 3D print parts
- mains filter waiting in the parts drawer
- contact plate made with Dremel, drill press and carpet knife, out of a salvaged PCB with huge  ground planes
- mains xfmr salvaged out of a phone system
- housing made from a salvaged MEADE battery charger
- ...

If someone needs the shroud, I can share collada or stl files.

Thanks for all the information shared in this thread!

Attached you find some pics.

br, mf

[Neomys Sapiens]

(......)
A few pictures of my PSU . . .

A year ago or so, I got a pair of P5205 HV probes fairly cheap, without noticing the TekProbe supply needed. Yikes!

I did not want to cannibalize the probes - being inspired by @128er, I made my own.

Hacktronix
- simple ±12V
- no ±5V nor offset rubbish
- some 3D print parts
- mains filter waiting in the parts drawer
- contact plate made with Dremel, drill press and carpet knife, out of a salvaged PCB with huge  ground planes
- mains xfmr salvaged out of a phone system
- housing made from a salvaged MEADE battery charger
- ...

If someone needs the shroud, I can share collada or stl files.

Thanks for all the information shared in this thread!

Attached you find some pics.

br, mf

Please put the shroud .stl here:
https://www.eevblog.com/forum/testgear/replacement-knobs-feet-and-fittings-for-test-equipment
Would help me much with a TekProbe breakout box!

[Mechatrommer]

(......)
A few pictures of my PSU . . .

A year ago or so, I got a pair of P5205 HV probes fairly cheap, without noticing the TekProbe supply needed. Yikes!

I did not want to cannibalize the probes - being inspired by @128er, I made my own.

Hacktronix
- simple ±12V
- no ±5V nor offset rubbish
- some 3D print parts
- mains filter waiting in the parts drawer
- contact plate made with Dremel, drill press and carpet knife, out of a salvaged PCB with huge  ground planes
- mains xfmr salvaged out of a phone system
- housing made from a salvaged MEADE battery charger
- ...

If someone needs the shroud, I can share collada or stl files.

Thanks for all the information shared in this thread!

Attached you find some pics.

br, mf

if BNC cable from that PSU output to scope is long, you risk reduced signal integrity (reflection), if its me, i will make an adapter from tek BNC input directly to normal scope's BNC output with power pins socket on the side, so PSU is external device with wires connecting to the adapter (which will be connected directly to the scope) much like tek 1141a diff probe with external 1142a psu module.

[minifloat]

(...)
If someone needs the shroud, I can share collada or stl files. (...)

Please put the shroud .stl here:
https://www.eevblog.com/forum/testgear/replacement-knobs-feet-and-fittings-for-test-equipment
Would help me much with a TekProbe breakout box!

done so: https://www.eevblog.com/forum/testgear/replacement-knobs-feet-and-fittings-for-test-equipment/msg4611142/#msg4611142 

(...)
Attached you find some pics. (...)
br, mf

if BNC cable from that PSU output to scope is long, you risk reduced signal integrity (reflection), if its me, i will make an adapter from tek BNC input directly to normal scope's BNC output with power pins socket on the side, so PSU is external device with wires connecting to the adapter (which will be connected directly to the scope) much like tek 1141a diff probe with external 1142a psu module.

Yes I know - the P5205 requires the scope inputs to be set to Hi-Z.
I am using 30cm of BNC cable after my box, since I placed the Hacktronix supply right below my 'scope.

Using the original Tektronix 1103 supply with lengthy cable at the output would lead to the same problem

br, mf