Products > Test Equipment
Help needed - Siglent TPA10 / Tektronix TekProbe offset not working
<< < (4/6) > >>
alm:

--- Quote from: tautech on February 12, 2022, 06:29:19 am ---
--- Quote from: alm on February 10, 2022, 12:35:51 am ---
--- Quote from: FlexibleMammoth on February 10, 2022, 12:23:30 am ---The p6247 is working with tape over pin5 as if the scope was set to 0v offset, so my probe is good.

--- End quote ---
That's good to know! So the problem is that the Siglent adapter is incompatible with this probe, and probably with any probe that relies on the TekprobeII interface to set the offset.

--- End quote ---
From investigations it does seem TPA10 is a TekProbe2 device. however there is no mention of this in documentation or on websites. 
Should TPA10 also be TekProbe backwards compatible ?

Can anyone test TPA10 with an older Tek probe that only uses the TekProbe interface ?

--- End quote ---
If you compare the TekProbe pinout from the Tek 1103 manual with the TekProbe II pinout (both attached to this post), they are identical except that TekProbe II adds clock and data for digital communication with the probe (where TekProbe and older probes used a resistor on the readout pin to signal attenuation). So a TekProbe II host should be backwards compatible with any TekProbe probe, save maybe the readout feature. To the point where Tektronix barely distinguishes the two. For example the TCA-1MEG TekConnect to BNC adapter is described as implementing the TEKPROBE interface, yet the pinout clearly shows the clock and data pins.

Does the Siglent adapter even support the digital interface? Or is it just implementing the TekProbe interface? Either way, a TekProbe interface should support any probe supported by the 1103, which I believe is pretty much all of them.

The ADA400A and P6247 are both compatible with TekProbe (I) scopes, like the TDS400/500/600 series and 1103, and are listed as currently available products compatible with the latest BNC scopes and BNC adapters. I'm not quite sure if they implement TekScope or TekScope II. They were introduced in 1998, when Tek says the current TekProbe (II) interface with clock and data was introduced. Yet the Tek 1998 and 2000 catalogs don't mention "TekProbe II" anywhere I can see (not even for the newly introduced scopes), and all of the then newly introduced probes list compatibility with the 1103. So I until I see evidence to the contrary, I assume that if anything the newer probes just add an optional digital interface for the scope to be able to get more information about the probe beyond what the readout resistor provided.
tautech:
Thanks alm.  :-+

It's not like Siglent has the relationship with Tek like they do with LeCroy so propriety TekProbe configuration and probe identification SW code they really need work out for themselves and it seems to need more work.  :-//

Any further clues to get the TPA10 fully working would be much appreciated.
alm:
I believe the original TekProbe was considered an open standard of sorts. The 1103 manual and the page I linked gives most of the specifications except the readout pin. The link I posted earlier gives empirical information about the values of the readout resistor, but I'm not aware of an official document specifying them. However, half an hour with a recentish Tek scope and a decade resistor should give you this information if you don't want to trust some random person posting on Usenet. There may be some subtle points, like noise and output impedance. I remember Steve on the TekScopes group mentioning that the poor CMRR of some FET probes became a problem when they introduce the TCP202, which would introduce ripple on the power rails. Still, implementing the 1103 specs should get them 95% there.

The TekProbe II standard is explicitly described by Tektronix as proprietary, so I can imagine reverse engineering that 100% may be more involved, although I'm guessing it's just an I2C EEPROM. There has been some effort reverse-engineering this EEPROM. So I could accept it for example not detecting the proper input impedance for an active probe. But the purely analog interface of the 1103 should be easy to implement, and this already specified the offset signal from -1 V to +1 V. So setting that to 2.5 V seems wrong no matter what.

I have no idea how to improve the TPA10 without Siglent's help other than taping off the offset contact or maybe generating your own offset signal from a pot.
FlexibleMammoth:

--- Quote from: alm on February 12, 2022, 02:57:41 pm ---Does the Siglent adapter even support the digital interface? Or is it just implementing the TekProbe interface? Either way, a TekProbe interface should support any probe supported by the 1103, which I believe is pretty much all of them.

--- End quote ---

I reported to Siglent some time ago that the automatic scale factor and termination (1MOhm/50Ohm) detection were not working. They told me that "The current [firmware] version is fully functional, but unfortunately automatic detection is not supported". It is not really clear if they refer to the scale factor, the input termination, or both.

Either way, there seems a lot missing here - let's assume my unit is defective and the offset on pin 5 normally works:

1) There is no analog offset control for passive probes or TekProbe1 probes.
That one is interesting - because this voltage is generated inside the scope and the signal path of the adapter is most likely a straight through. So that means when the TPA10 is connected, Siglent scopes disable the internal offset and only provide offset on pin 5. That sounds like a hack to avoid double offset problems with certain probes, but kills the offset completely for all probes that don't make use of pin 5.

2) There is no detection of the scale factor.
That one surprised me the most, as the metal ring around the BNC on the scope already does that.

3) There is no TekProbe 2 digital interface, so things like output impedance, probe names, probe termination, calibration factors etc. dont work.
Understandable, especially given their quite reasonable price for the adapter. But that would be amazing nontheless.
tautech:
This issue is firmly in front of top ppls at Siglent in the hope they can get TPA10 working as it should however I have no idea what Tek active probes they have so to go sniffing the TekProbe I2C coms therefore any sniffings you can provide and post here for their analysis could be helpful to improve TPA10.
TIA
Navigation
Message Index
Next page
Previous page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod