Author Topic: Tek p5205 HV differential probe teardown. BTW, what are the red and brown wires?  (Read 46755 times)

0 Members and 1 Guest are viewing this topic.

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17427
  • Country: us
  • DavidH
What I find really odd about this discussion is that such a device used to exist!

Oscilloscope input "normalizers" are used to adjust the input capacitance of the vertical inputs of an oscilloscope as well as the compensation of the input attenuators but these normalizers also function as compensated precision divide by 2 attenuators.  I have a couple of them in my collection of calibration instruments.
I figured there had to be; this didn't seem like such an unusual situation.   But I couldn't find any to say, "Here, buy this!"  Thanks for the pointer.  Is there a compensation adjustment on a "normalizer"?

Yes, it is not shown in the photo because it is on the bottom but there is an access hole for the trimmer capacitor.

In practice they were compensated against one channel and then used to adjust the compensation of the other channels and all of the input attenuator sections.  They could also be used to compensate multiple oscilloscopes against each other so probes could be moved between them easily.  A x10 probe may be used to do the same thing but it will have 5 times more attenuation of the test signal which may or may not be a problem.

Quote
Quote
There's no information I can find that says exactly *which* specs are invalidated if you go with 50 ohms (I suspect accuracy with max signal input), so in reality the above may be overkill for your purposes.  I would say try the 50 ohm termination with various signals that you typically measure to see if it works well enough.

The accuracy of the 50 ohm termination will degrade the calibration.  The high voltage differential probes I am familiar with specify two different attenuation factors depending on if their output is terminated into 50 ohms or not.
Right.  I'm expecting the P5205 won't be able to swing to the peaks of its specified output with 50 ohms and the waveform will get squashed.

The waveform will certainly be halved but if they designed it to drive a 50 ohm transmission line to its 100 MHz bandwidth, I doubt it will have problems besides that.
 

Offline MarkL

  • Supporter
  • ****
  • Posts: 2302
  • Country: us
Right.  I'm expecting the P5205 won't be able to swing to the peaks of its specified output with 50 ohms and the waveform will get squashed.
Did a quick test and this is correct.  Instead of the output swinging 6.2V Pk-Pk (measured; 5.2V is the spec), I only get 2.8V Pk-Pk with 50 ohm termination.

A side effect of 50 ohm termination, which I didn't expect, is that the overrange LED does not light.  It's obviously looking at its own output to determine overrange and the level never gets high enough.  So, 50 ohm termination with this probe is not a good thing.

Edit: Since the source impedance is 50 ohms, 2.8V is actually not too far off expectations.
« Last Edit: October 08, 2014, 08:35:48 pm by MarkL »
 

Offline Mechatrommer

  • Super Contributor
  • ***
  • Posts: 11714
  • Country: my
  • reassessing directives...
make active 1/2X divider or 5X amplifier with an opamp then as suggested by the OP earlier.
Nature: Evolution and the Illusion of Randomness (Stephen L. Talbott): Its now indisputable that... organisms “expertise” contextualizes its genome, and its nonsense to say that these powers are under the control of the genome being contextualized - Barbara McClintock
 

Offline MarkL

  • Supporter
  • ****
  • Posts: 2302
  • Country: us
Not sure I saw that suggestion, but why would you want to start adding active components?  Plus added complications of DC to >100MHz?

The question I was trying to answer was if a simple, widely available, 50 ohm terminator could be used.  A: It can, but there's a tiny amount of non-linearity on max signals, and you lose the overrange light.  Or, to stay in compliance with the probe output specs, build something passive with 1 or 2 resistors and a trimmer cap.
 

Offline onlookerTopic starter

  • Frequent Contributor
  • **
  • Posts: 395
Right, I did not mean to add my own active components. The talk about x100/x1000 (/2) or x10/x100 (5x) was to say that, after all, mentally dividing by 2 or multiplying by 5 may not be that bad, though from the beginning, I did contemplate about what might be possible by some simple unofficial adjustments of the current PCB to reach x10/x100.

If at all possible, x10/x100 (5x) is more preferable than x100/x1000 (/2) since it reduces the attenuation that the device currently has and smaller signals will likely be less noisy (yes, small signal after P2505 appeared to be a lot more noisy).

As you guys had mentioned,  x10/x100 can be done by rearranging the voltage divider at the HV side, then, at the expense of de-rating the HV performance (but, how much de-rating, 5 times worse? I do not have a quick answer). Or, is there any room in the x50/x500 calibration pots on the PCB?

On the other hand, as I mentioned earlier, for now, I am just trying to make it usable with minimum effort. My eBay adventure just failed. The RG174 I got was junk. The shield has probably less than 60% coverage, that resulted to significantly high signal loss. The new hunt for lower loss RG174 is in progress and the price is likely more than double. 
« Last Edit: October 10, 2014, 12:29:13 am by onlooker »
 

Offline 128er

  • Regular Contributor
  • *
  • Posts: 229
  • Country: de
Little progress with my P5205 probes.

Today I had some time and etched a pcb to easier test my (supposedly) broken P5205 probes. As I said, bought them cheap on Ebay. It seems that they come from some sort of lab. They all have "defective"-stickers with a short failure description on them. On two of them is written: "generates voltages that are not plausible", and on the other: "x50 range defective".

At the moment I can only test them with my variable isolation transformer. Actually no chance to test them to full Bandwidth. Maybe at work I can find some (reasonable) high frequency power stuff.

But...
In both cases, they work fine (with my limited test conditions). Only the x50/x500 switch looks dodgy. You touch the switch slightly and the probe output goes nuts.

I don't expect to find these switches somewhere. Have to look. Has anyone ever observed the same issue?

Have two more probes to test. But not this evening.
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17427
  • Country: us
  • DavidH
But...
In both cases, they work fine (with my limited test conditions). Only the x50/x500 switch looks dodgy. You touch the switch slightly and the probe output goes nuts.

I don't expect to find these switches somewhere. Have to look. Has anyone ever observed the same issue?

Are you referring to the three push-button switches?

From your picture it looks like the body of the switch easily comes apart so you can clean the contacts and put it back together.  What kind of switch mechanism do they use inside?

I have had rubber dome switches with a carbon conductive surface and gold plated contacts which just needed cleaning.  My guess is that they were contaminated during cleaning to remove flux but took years to fail.
 

Offline 128er

  • Regular Contributor
  • *
  • Posts: 229
  • Country: de


Are you referring to the three push-button switches?


Yes.

That's not my picture, but no matter.  :)

All Switches, at all probes working fine, except for the x50/x500 switch (probably the most actuated of all three switches). However, I opened the switches. They were perfectly clean. A  little metal spring contact shorts two carbon contacts to close the switch. I had to bend back the spring contacts. Obvious, they were worn out over the years.

Extreme fiddly work. :palm:

Thanks David for the hint!
 

Offline onlookerTopic starter

  • Frequent Contributor
  • **
  • Posts: 395
I have done a few things since the last update. These include

---- Came up with a power supply that is light and cheap;
---- Fitted connectors on P5205 for power supply and output signal;
---- Some performance tests about the power supply and the P5205.

The 1st picture shows the overall set-up.

----------------------------------
1). P5205: connectors for power supply and signal output

The high voltage input had fitted with HV wires before this round of experimenting. But, I still need some HV  accessories. They are expensive and relatively rare. Other than this, the wiring was more or less usable.

As for the low voltage output, I settled with a simple solution: just fit a female BNC to the existing short coax left on the P5205. For connecting to the external power supply, a short cable with a micro USB female jack was used.

Though the signal output and the power connection are on the low voltage side, I may later still like to have some silicone shroud to cover the exposed metal for, in case, the P5205 has to move around in HV surroundings. 

-----------------------------------
2). Power Supply: The construction

The power supply I ended up to use was a "DIY" hybrid type and quite simple: a SMPS pre-regulator followed by a linear regulation stage.

The pre-regulator SMPS was one I already had. It was unregulated and speced as 12V and 1.5A, though, the ripple can be quite high. The low cost of $1.50 made it disposable for experimenting. The other important consideration was that its half bridge topology with the typical center tap full wave rectification secondary can be easily modified to have +/- rails. For my use, I would only need up to about 0.2A per rail and, for such load, the modified SMPS supplied +/-13.2V.

For the linear regulation part, a $5 small dual rail 317/337 board rated to 1A was used. The board was adjusted to output +/-10.5V. On the board, the spots for rectifier diodes were fitted with 2 inductors taken from burnt out CCFL bulbs to form PI filters.

Now, there are two important questions: 1). Will the power be clean enough at the required load level; and 2). The P5205 was speced to use +/-15.0V power. Will a +/-10.5V supply be adequate?

----------------------------------------------
3). Power Supply: The noise measurements

AC differential measurements were taken at different test points with the P5205 as the load (100mA) and also for a resistive load (230mA). Of particular interest were the results at the power output.

With a HP3457A, the 2 rails showed readings of 0.13mVAC and 0.18mVAC at a 230mA load respectively (short leads reading=0.1mVAC).

On the scope, the +10.5V output did not show any noise above the noise floor; while the -10.5V rail output showed the worst at the (2mV/div, 20MHz limit, 1x probe) setting and it is shown in Picture #2.



The common mode (and radiated) noise was tricky to measure and my noisy scope just added the difficulty. Nevertheless, the comparative picture (#3) can still show the level of such noise. In this measurement, the ground lead of the probe was unused; while the measurement was taken with the probe tip touching the PS ground. The comparison was between whether the PS was plugged into the mains.



-------------------------------------------
4). P5205: The performance tests

These were aimed to provide quick (non-exhaustive) answers to 2 questions: a). Is a +/-10.5V PS a good enough replacement for a +/-15V PS; and b). How usable is a P5205 for "small" signals.

For the first question, a concern was that the reduced supply voltage may lead to waveform clamping for full range signals or it may lead to reduced band width of P5205. Up to now, I only did the clamping test.

For the 50X setting, P5205 is speced for a voltage range of +/-130V. To test the clamping, the mains (with a +/-170V range) was used as the signal source and no visible and measured difference in the waveforms was observed comparing using the +/-10.5V supply with a +/-15V supply.

On the other hand, I think being over range for the 50X setting when using mains did show a slight different waveform than that for the 500X setting (Picture#4) and the difference was independent of the supply voltage (10.5V or 15.0V). It was also more interesting to see that the mains waveform was not very "sine".



As for "small" signal tests, a ~10MHz square wave with Vpp ~= 0.7V was used. Picture #5 shows a comparison among the "original"  (without P5205 involved), and the P5205 input/output. With using P5205, the signal was 50 Ohms terminated at the P5205 input; While for the "original", the 50 Ohms terminator was on the scope input.

Two things are worth to note here: a). To get stable waveform from the P5205 output, my scope needed to set at average=16 samples; and b). There are excessive ringing when using P5205. Can these ringing be reduced? Then, maybe, my scope and cabling are to blame here?

« Last Edit: November 17, 2014, 01:29:37 pm by onlooker »
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17427
  • Country: us
  • DavidH
Two things are worth to note here: a). To get stable waveform from the P5205 output, my scope needed to set at average=16 samples; and b). There are excessive rings when using P5205. Can these rings be reduced? Then, maybe, my scope and cabling are to blame here?

If you needed to use averaging then the noise level seems high.

The frequency of the ringing may yield a clue about the cause but I cannot see it accurately on your image.  Is it 130 MHz?

Was the differential probe terminated into 50 ohms on the oscilloscope?
 

Offline 128er

  • Regular Contributor
  • *
  • Posts: 229
  • Country: de
2). The P5205 was speced to use +/-15.0V power. Will a +/-10.5V supply be adequate?

Can't remember which parts of the circuit use the +/- 15V. But IIRC the +/- 15V input is directly followed by two adjustable linear regulator, one for +5V and the other for -5V. And they power most of the probe circuit. Have to look in one of my probes.

Was the differential probe terminated into 50 ohms on the oscilloscope?

Don't know if the OWON is capable of 50 Ohm termination. But the probes are specified for 1 MOhm input impedance.



But it is nice to see that the thread is not dead. I have worked a little bit on my Tekprobe PSU at the last weekend.

I like Onlookers idea. It is more like a stand alone solution, instead of my brick like Tektronix 1103 clone  ;D

A few pictures of my PSU . . .

« Last Edit: November 17, 2014, 08:10:23 pm by 128er »
 

Offline onlookerTopic starter

  • Frequent Contributor
  • **
  • Posts: 395
Quote
The frequency of the ringing may yield a clue about the cause but I cannot see it accurately on your image.  Is it 130 MHz?

Yes, that should be about right.

Quote
Was the differential probe terminated into 50 ohms on the oscilloscope?

MarkL had already quoted the manual earlier: The scope end must have an impedance > 50kR. Or as 128er mentioned, 1MR should be fine.

128er, Your 1103 clone looked quite authentic  and nice.

Quote
Can't remember which parts of the circuit use the +/- 15V. But IIRC the +/- 15V input is directly followed by two adjustable linear regulator, one for +5V and the other for -5V. And they power most of the probe circuit. Have to look in one of my probes.

The output is speced to be +/-2.6V.  So, in theory, it is possible that a +/-5V supply is all needed, though a few chips on the PCB can work up to +/-15V. In any case, it should be an interesting item to followup on. If true, the requirement for the PS is further reduced.
« Last Edit: November 18, 2014, 01:38:58 am by onlooker »
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17427
  • Country: us
  • DavidH

Quote
Was the differential probe terminated into 50 ohms on the oscilloscope?

MarkL had already quoted the manual earlier: The scope end must have an impedance > 50kR. Or as 128er mentioned, 1MR should be fine.

I am thinking that the original cable had a compensation box on the end and was made from wire similar to x10 probe cable with a resistive center conductor to dampen reflections.  If that is the case, then modifying the new cable to work into a 50 ohm termination may be necessary to get full 100 MHz performance.

The 130 MHz ringing may be from the reflections along the length of the new cable and not anything to do with the probe end.
 

Offline onlookerTopic starter

  • Frequent Contributor
  • **
  • Posts: 395
I would quote MarkL again about the "compensation box" of P5205 that connects to the scope:
Quote
there's no compensation in it.  The coax goes right into the scope.

Also, my P5205 had about 5" coax left. The core was made of multi strand silver(?) plated copper wire like RG400.

Then, the manual says: "Output Type: Single-ended  Source Impedance of 50 ohms drives 1 M ohm oscilloscope input". That is, yes, the scope end will reflect, but the scope would never see it since the P5205 end has a matching impedance and will not reflect back the reflection. This is what was meant by "single-ended".

The square wave picture I posted earlier also showed that the output of p5205 followed the input "closely", including the ringing. That is, the ringing should be from the input of P5205. My input cables are about 20cm in length that are shorter than the original cables which are 30cm. I should expect less ringing? 

The problem is that the manual did not spec the ringing. Maybe they should spec it  in some way since the manual did have pictures showing extra ringing as an expected result of using P5205 and adviced to twist the input cable pair.
« Last Edit: November 18, 2014, 01:28:58 pm by onlooker »
 

Offline MarkL

  • Supporter
  • ****
  • Posts: 2302
  • Country: us
I would quote MarkL again about the "compensation box" of P5205 that connects to the scope:
Quote
there's no compensation in it.  The coax goes right into the scope.

Also, my P5205 had about 5" coax left. The core was made of multi strand silver(?) plated copper wire like RG400.
All the clues you point out correctly conclude it's 50ohm coax and not special resistive coax.

Plus, I'll also add that the 5205 can be used with the 1103 power supply, which you extend to the scope using 50ohm coax.

And to put the question to a final rest, I took the cover off mine and measured 0.1ohm from the BNC center pin to the cable termination solder joint in the 5205 main box.


In one measurement, you're showing ringing on both the input and output.  If you disconnect the 5205 from the dual banana adapter, do you still see ringing on the input coax?
 

Offline free_electron

  • Super Contributor
  • ***
  • Posts: 8550
  • Country: us
    • SiliconValleyGarage
2). The P5205 was speced to use +/-15.0V power. Will a +/-10.5V supply be adequate?

Can't remember which parts of the circuit use the +/- 15V. But IIRC the +/- 15V input is directly followed by two adjustable linear regulator, one for +5V and the other for -5V. And they power most of the probe circuit. Have to look in one of my probes.

Was the differential probe terminated into 50 ohms on the oscilloscope?

Don't know if the OWON is capable of 50 Ohm termination. But the probes are specified for 1 MOhm input impedance.



But it is nice to see that the thread is not dead. I have worked a little bit on my Tekprobe PSU at the last weekend.

I like Onlookers idea. It is more like a stand alone solution, instead of my brick like Tektronix 1103 clone  ;D

A few pictures of my PSU . . .
that's a nice metal box. who makes that ? i like the punched noches to slide the board in.

schematic of the supply ?

i am sitting on a bunch of 6247 diff probes and their active counterparts but no 1103. and 300$ is too expensive for that stupid thing.

By the way, these probes are not made by Tektronix. Tek installs their own pigtail on it. The probe itself is made by Pintek. it is sold under many brand names. Agilent , fluke Tek , BK precision : they all have this probe in one form or another , all made by Pintek.
Same goes for all those current probes out there. only tek has their own. almost any other brands are rebadged hioki's.
Agilent scope probes are made by PMK in germany ( only the old infinium probes were an in-house job )
« Last Edit: November 18, 2014, 03:47:07 pm by free_electron »
Professional Electron Wrangler.
Any comments, or points of view expressed, are my own and not endorsed , induced or compensated by my employer(s).
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17427
  • Country: us
  • DavidH
I forgot about the description of the original coaxial cable and compensation box.

What kind of cable was used on the input side?
 

Offline 128er

  • Regular Contributor
  • *
  • Posts: 229
  • Country: de
that's a nice metal box. who makes that ? i like the punched noches to slide the board in.

Bought that case here: http://www.eibtron.com/epages/eibtron.sf/en_GB/?ObjectPath=/Shops/eibtron/Products/754-5979

Also available on RS: http://de.rs-online.com/web/p/gefahrstoffschranke/7545979/

I don't know who the original manufacturer is. Top, bottom and side covers are steel. Front an rear panels are aluminium.

schematic of the supply ?

Is attached.

JP1-6 are header connectors for LED's. Two indicates that the +/- 15V are present. And two for each channel to show the attenuation factor for the channel.

Didn't heard of PMK electronics before. For those who are interested: www.pmk.de/en/
« Last Edit: November 18, 2014, 07:14:16 pm by 128er »
 

Offline MarkL

  • Supporter
  • ****
  • Posts: 2302
  • Country: us
That's a really nice job on the power supply.

All you need to do now is add the offset controls and, with a reasonable price, you could clean up selling them as Tek 1103 replacements.
 

Offline onlookerTopic starter

  • Frequent Contributor
  • **
  • Posts: 395
 

Offline onlookerTopic starter

  • Frequent Contributor
  • **
  • Posts: 395
I did more tests to determine the power supply requirements. These were done by measuring the signal amplitude reading on the output of 5205, as a function of the power supply voltage.

1). Power supply dependency of bandwidth:
A sine signal of 100MHz and Vpp of a couple volts was connected to the input of the 5205.

For supply changing from 15.0V to 9.0V, the signal output stayed flat  at about 2.6V. When supply further reduced to 8.0V,  the signal output was dropped to  2.45V, or  a 0.5dB attenuation.

2). Power supply dependency of signal clamping:
For this test, the mains voltage with its voltage range of +/-170V was applied to the input and the 5205 was set to the 50X setting which has a speced voltage range of +/-130V.

The waveform amplitude and shape stayed the same for the supply varying from 15V to 10V. When the supply further dropped to 9.0V, the output voltage clamped about 7% and other artifacts started to show up at the peaks of the waveform . Considering that 170V was a ~30% over range signal, I would guess an in-range signal should not show clamping for a supply of about 9V.

Based on these tests, I think a supply voltage between 10V (or 9V) to 15V should work fine. 
 

Offline free_electron

  • Super Contributor
  • ***
  • Posts: 8550
  • Country: us
    • SiliconValleyGarage
Professional Electron Wrangler.
Any comments, or points of view expressed, are my own and not endorsed , induced or compensated by my employer(s).
 

Offline 128er

  • Regular Contributor
  • *
  • Posts: 229
  • Country: de
Hi all!

Just wanted to show you my finished PSU for the P5205 differential probes that I have. It works fine.  :-+

The labeling is made with self adhesive letters. Sticks quite well. The LED's seems a bit dim, but that is due to the camera flash. The brightnes is good. Not too bright, so that it is irritating on the bench.

The BNC connectors are okay. I had prefered some sort of front panel BNC's with crimp connection. Or one of those front panel BNC's with a plug on the rear side (couldn't find a picture at the moment). But none of my usual suppliers had them (for a reasonable price).

But all in all I'm quite happy with the thing.

The last thing left to do, is to properly connect alle enclosure parts with the PE. And maybe change the screws on the front panel, to some nice looking round head with hex socket.

Thanks to all who had replied to this thread and gave some good infos.
« Last Edit: December 18, 2014, 12:57:40 am by 128er »
 

Offline Fluxed Matter

  • Regular Contributor
  • *
  • Posts: 114
  • Country: us
Just picked up the P5205 for cheap off fleabay and it arrived yesterday and have a couple of questions.

The first thing is how do you remove the enclosure and get at the pcb? Been trying to figure it out without destroying the case. Do you just pry it open?

Love the power supply 128er! Do you have the image for the probe board available? I would like to etch one up for the PS I will build.

I picked the probe up and it has a few calibations stickers on it from '08 but someone wrote bad in permanent marker on both sides. It came complete with cables and connectors on both ends. Hoping to get this working because I been wanting a probe for power supply work.
Have a Great Day!
Fluxed Matter
 

Offline onlookerTopic starter

  • Frequent Contributor
  • **
  • Posts: 395
The case was fastened with four screws. You can access the screws by lifting the top plastic sticky sheet cover. You need only to lift it at each corner just enough to see the screw well. A sharp knife can be used to do the initial lifting.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf