EEVblog #978 - Keysight 1000X Hacking

[wxqhigh]

The value is in the PDF version of the schematic in reply #835. It was selected based on minimum ringing  in a circuit simulator with  step voltage applied to pin 3.

c31 and c13 are not noted

[Bud]

I've refreshed the PDF now, not sure how that happened.

[JDubU]

Bud,

Did you add a switch to turn on the waveform generator function?

[Bud]

I have not finished that part. Only controlled the gen via GPIB.

[EinErik]

Hi,

Checked the ID resistor, no problem. I found some others have the problems too.
https://www.eevblog.com/forum/testgear/keysight-1200a-x-series-(edux1052a)-p-clock-and-i-clock-malfunction-error/

The scope showed no waveforms, and indicate P/I-clock error.

Memeber pnv57 said:

This happened to me, after I ripped the two terminals that are next to each  other on the BLT board.
I soldered them back, and got those errors.
Went back through, and soldered them all better, errors gone, now I have a trace.
The issue is that the trace would "jump" and the sine wave would be static (triggered) on the screen, but there would be a little step in it.
I went back though, and I added solder while bending each pin down to the board with a pick, and allowing the solder to harden before removing the pick.
This made all of the connections much better and fixed the problem for me.

I suggest that you remove and reinsert the BLT board, and if that doesn't work, use a pick and ensure each pin of the 4 connectors is firmly attached.

Honestly, I don't really understand what he's saying? By just re-inserting the BLT module can fix the error?
I have tried re-inserting many times but showed no improvement.

So, it looks like it's not the ID config resistor problem, but the soldering junction contact problem?

Any suggestion for my situation and comments? Please help, thank you!

[wxqhigh]

Hi,

Checked the ID resistor, no problem. I found some others have the problems too.
https://www.eevblog.com/forum/testgear/keysight-1200a-x-series-(edux1052a)-p-clock-and-i-clock-malfunction-error/

The scope showed no waveforms, and indicate P/I-clock error.

Memeber pnv57 said:

This happened to me, after I ripped the two terminals that are next to each  other on the BLT board.
I soldered them back, and got those errors.
Went back through, and soldered them all better, errors gone, now I have a trace.
The issue is that the trace would "jump" and the sine wave would be static (triggered) on the screen, but there would be a little step in it.
I went back though, and I added solder while bending each pin down to the board with a pick, and allowing the solder to harden before removing the pick.
This made all of the connections much better and fixed the problem for me.

I suggest that you remove and reinsert the BLT board, and if that doesn't work, use a pick and ensure each pin of the 4 connectors is firmly attached.

Honestly, I don't really understand what he's saying? By just re-inserting the BLT module can fix the error?
I have tried re-inserting many times but showed no improvement.

So, it looks like it's not the ID config resistor problem, but the soldering junction contact problem?

Any suggestion for my situation and comments? Please help, thank you!

It seems that he resoldered all connectors on BLT board
I guess because it's a multi-layer board and it's a little bit diffcult to take down, it may be injured when you take down it.
I wish it doesn't have any injury between internal lines

[candrian]

Has anybody managed to hack the function generator of DSOX1202G? Could it be converted to a waveform generator ?

[wxqhigh]

Has anybody managed to hack the function generator of DSOX1202G? Could it be converted to a waveform generator ?

I'm afraid not,because it needs to redesign the whole software and hardware,not just simply hack hum？
It’s better to buy an AWG

[candrian]

It looks to me the waveforms it already creates are not standard waveforms. It looks like the SW is reading a file describing the waveform's pattern and then producing it on the output. I believe it's also an easier implementation for the manufacturer to have been implemented like this.

[Bud]

There has to be UI items for each waveform in the menu. You can't hack something that does not exist.

[candrian]

Yes I understand what you have in your mind but what you are describing is in higher kevel. The menu items are the front end but in the back end it's not impossible to call an app which loads a waveform pattern from a descriptor file.

 It doesn't mean that because there is no option in the menu that you can't run something in the back end using another interface once you have gain access to the system.

[wxqhigh]

As for the wavegen button,they didn't redesign the button pannel,just cut them off to cut the cost.
So...you just need to find something to prolong it or just use something to stick it.

[wxqhigh]

I finished my EDUX1052A mod(both frontend and wavegen) but have some problems.

When i use product id 33(DSOX1202G) or 35(EDUX1052G),I can't pass the user cal(hardware self test passed),but 1202A(32) and 1052A(34) are passed.

The wavegen works not well,the amplitude of the signals getting too high causes signal clipping,and in some amplitude it's not coherent,I don’t know whether it's reason is uncalibrated.

The uart in new models is locked,it needs the username and password to login,so i can’t get any information.

SOLVED:There is another part of components on the back of board showing in #864

[JDubU]

Have you verified that the Self Cal feedback circuit in the Wavegen is working properly?
It is comparator U10D in Bud's schematic:

https://www.eevblog.com/forum/blog/eevblog-978-keysight-1000x-hacking/?action=dlattach;attach=1288201

[wxqhigh]

Thanks,There is another part of components on the back of board,I didn't notice them!
After added them,self test and user cal passed,wavegen works well.

Conclusion:EDUX1052A can be moded to DSOX1202G 70M version,enjoy it!

PRODUCT ID:
DSOX1204A: 30
DSOX1204G: 31
DSOX1202A: 32
DSOX1202G: 33
EDUX1052A: 34
EDUX1052G: 35

[raphaelCoelho]

Hello! Updating the firmware of edux1002G, without changing anything in the hardware, is there any improvement? Memory depht? Waveforms?

[Bud]

You will get more memory but if any of the added modes require enhanced trigger, the scope may likely fail user calibration because you did not modd-ed the trigger circuit.

[raphaelCoelho]

Understand. For the hack to be effective, would i'm have to modify the firmware and hardware then? And the wf/s, any improvement?
I see that the edux1052 has twice the memory and wf/s, so I'm not sure whether to buy the edux1002g (1 year warranty) or edux1052g (3 year warranty) the distributor has the 2 models here, but the 1052g is $230 (usd) more expensive

[Bud]

I do not know how to measure wf/s , and there is no indication of this mettic in the setup screen.

[raphaelCoelho]

I wrote wrong sorry. It's wfm/s.

[Bud]

I know what you mean. This is a specification on the device datasheet, you can't measure it.

[raphaelCoelho]

Yes. My question was if the wfm/s specification changes when updating the edux1002g firmware?

[Bud]

That is what I am trying to say - you can't measure it and therefore it is unknown if it changed or not.

[raphaelCoelho]

Now I get it. Sorry for this, my english is bad.
Thanks for the answers.
Do you think I should buy the edux1002g or the edux1052g? EDUX1052G is more expensive, but in addition to having better performance, it has an additional 2 years warranty.

[Bud]

It should be your decision.