Author Topic: Keysight 3000 4000 MSOX4154A P-CLOCK failure  (Read 2228 times)

0 Members and 1 Guest are viewing this topic.

Offline EE-diggerTopic starter

  • Frequent Contributor
  • **
  • Posts: 391
  • Country: us
Keysight 3000 4000 MSOX4154A P-CLOCK failure
« on: December 22, 2021, 05:13:59 pm »
My MSOX4154A warranty ran out 11/2019 so I'm not sure there's an affordable way out of this.

I had seen a “System concerns detected: P-Clock failure” once or twice over the last few years.  I only use the scope about once a week.

In the past few days it has been a 50:50 chance of getting it and even if the scope works, it seems to get corrupted as it sits (same whether cool or warmed up).

Has anyone seen the P-CLOCK failure and know how to fix it?  Keysight has a page on it in their edadocs - kkbopen pages.  They call it "relatively uncommon" (unless you have it  :( ).  When it has the failure, ADC function on channels 3 and 4 seems out to lunch with large offset and various vertical alias of the waveform.

Their suggestions range from running self test to  re-installing the firmware, even if the same version.  After that, it's off to the Keysight shop.  My problem is I'm running on an  "enhanced" 7.40 and once again, was too dumb to check monthly and save versions.

This problem, publicly acknowledged by the Keysight page, seems like a timing margin problem.  As such, I would hope they would fix it for free and may need to continue hoping for that if there is no other solution.  If it will fix the scope, I would install 7.50 but lose the goodies.  Unless their update writes new FPGA code, I have doubts about it fixing the problem.

update - found my copy of factory 7.40 and installed that.  Self test now reports SIPO(): ADC failure which had also (though rarely) shown up before.

Still sounds like timing.  What do you all think?  One other tidbit ... my lab is about 10 degrees cooler than in summer (average 63F now).  I may warm it up and see if that changes things.

« Last Edit: December 22, 2021, 07:10:10 pm by EE-digger »
 

Offline TheSteve

  • Supporter
  • ****
  • Posts: 3785
  • Country: ca
  • Living the Dream
Re: Keysight 3000 4000 MSOX4154A P-CLOCK failure
« Reply #1 on: December 22, 2021, 11:18:33 pm »
That is an unfortunate failure on an expensive machine. I don't think I've ever seen anyone who has repaired that problem. If you do find temperature changes the behavior maybe you could locate the suspect component(s).
VE7FM
 
The following users thanked this post: EE-digger

Offline EE-diggerTopic starter

  • Frequent Contributor
  • **
  • Posts: 391
  • Country: us
Re: Keysight 3000 4000 MSOX4154A P-CLOCK failure
« Reply #2 on: December 22, 2021, 11:45:20 pm »
If I do find that temperature masks it, and together with Keysight's comments in the past, I will get down on both knees and beg for end of year redemption  :palm:

Interestingly enough, over a year or so ago it would not pass user cal with my precision matched test cable.  I had to  put it out in the cold for an hour and run the user cal when it first came it.  I think it complained about temperature delta and I repeated until it ran ... and PASSED !

 

Offline EE-diggerTopic starter

  • Frequent Contributor
  • **
  • Posts: 391
  • Country: us
Re: Keysight 3000 4000 MSOX4154A P-CLOCK failure
« Reply #3 on: December 23, 2021, 02:54:19 am »
Ok, here is the result of temperature testing.

63 deg F - failing self test almost 100% now

75 deg F - still failing most of the time (1 hour soak)

77 deg F - passing self test 100 % (2 hours soak)

I can do further temperature testing, taking it near max operating temp.  If it stays fixed after coming back to room temp, that may point to a bad connection.  Otherwise, if it follows temperature it may be a hardware timing problem.

The unit has never been opened so I'll wait and see if anyone has suggestions, then talk to Keysight first.
 

Offline Bud

  • Super Contributor
  • ***
  • Posts: 7276
  • Country: ca
Re: Keysight 3000 4000 MSOX4154A P-CLOCK failure
« Reply #4 on: December 23, 2021, 05:44:45 am »
There are posts about P clock failure in other threads on this forum from owners of 1000 series scopes when they performed hardware modification that involved removal and re-seating of the BLT board (the module with the fpga, cpu and ADC). One person traced it to a broken connector.
Facebook-free life and Rigol-free shack.
 

Offline TheSteve

  • Supporter
  • ****
  • Posts: 3785
  • Country: ca
  • Living the Dream
Re: Keysight 3000 4000 MSOX4154A P-CLOCK failure
« Reply #5 on: December 23, 2021, 06:08:14 am »
It certainly isn't a really common failure, I wouldn't expect a "good will" repair but you just never know. Talk nicely to Daniel, that man can work miracles.

And to be fair, running patched firmware would also invalidate the warranty(if it had one).

Maybe you want to sell it cheap because it is broken - I'm interested :)
« Last Edit: December 23, 2021, 06:10:03 am by TheSteve »
VE7FM
 

Offline EE-diggerTopic starter

  • Frequent Contributor
  • **
  • Posts: 391
  • Country: us
Re: Keysight 3000 4000 MSOX4154A P-CLOCK failure
« Reply #6 on: December 23, 2021, 02:32:15 pm »
Last night it started failing again at a few degrees higher (79 deg F).

[/quote]
There are posts about P clock failure in other threads on this forum from owners of 1000 series scopes when they performed hardware modification that involved removal and re-seating of the BLT board (the module with the fpga, cpu and ADC). One person traced it to a broken connector.

The unit has never been opened and is now running factory 7.40.  There are no relevant inter-board connections but it could be the FPGA or one of the ASICs.

Regarding broken and cheap, lately Ebay prices for "parts only" seem to be insane.  For a broken $10k scope, I want to pay $1200 if I can fix it, you pay me to haul it if I can't  :-DD  :-DD
« Last Edit: December 23, 2021, 02:34:10 pm by EE-digger »
 

Offline Bud

  • Super Contributor
  • ***
  • Posts: 7276
  • Country: ca
Re: Keysight 3000 4000 MSOX4154A P-CLOCK failure
« Reply #7 on: December 23, 2021, 03:38:15 pm »
If the fault is caused by a tiny clock timing margin, i would try changing the clock oscillator frequency a tiny bit or play with the supply voltage up/down a little. If you can, try measuring them as you do your thermal testing, maybe you will spot a correlation.
Facebook-free life and Rigol-free shack.
 

Offline EE-diggerTopic starter

  • Frequent Contributor
  • **
  • Posts: 391
  • Country: us
Re: Keysight 3000 4000 MSOX4154A P-CLOCK failure
« Reply #8 on: December 23, 2021, 04:43:39 pm »
I'll try the external reference when I move it back to the lab ... it's appropriately in the shitter at the moment

I'm running a bit hotter now (85 deg F) and it passes, fails, passes ...

The model is good to 50C so I'm going to see how it behaves at around 90 - 95F range.

If it passes, then fails, it may be a connection / solder problem. 
« Last Edit: January 02, 2022, 01:14:31 pm by EE-digger »
 

Offline EE-diggerTopic starter

  • Frequent Contributor
  • **
  • Posts: 391
  • Country: us
Re: Keysight 3000 4000 MSOX4154A P-CLOCK failure
« Reply #9 on: December 24, 2021, 05:18:35 am »
Further heating, cooling and variation of external clock did not give results which would correlate to the two fault conditions.  Same distribution of pass/fail instances for all.

It seems to behave best when temperature is ramping up or down, leading me back to the bad solder joint theory.  During the gradient, the connection improves due to thermal coefficient mismatches, then degrades as temperature stabilizes at a new value.

Next step is Keysight.

Thanks for your inputs.
 

Offline EE-diggerTopic starter

  • Frequent Contributor
  • **
  • Posts: 391
  • Country: us
Re: Keysight 3000 4000 MSOX4154A P-CLOCK failure
« Reply #10 on: December 25, 2021, 05:28:46 pm »
Christmas morning findings ...

Santa left a tiny, super quiet Harbor Freight compressor under the tree.  The sound level of these things is unreal.  Could have saved 30 years of canned air purchases and resided in the electronics lab.

New observations and updated theory on scope...

With scope sitting on my bench, left vents toward a "leaky" window so somewhat cooler at that end, failed 100%, all day yesterday.

On my lap, exhaust against my leg, left and right vents blocked by my hands, passed after about 3 minutes of warming   :)

Now, Christmas morning, fails on power-up and when warmed BUT, after 5 minutes of LEFT vents covered in tape, unit passes and continues to pass for several hours so far.    :)  :)  SMPS areas are no longer bathed in cooler air that's pulled in.  This is now a steady state condition in that the SMPS areas are warmer, not ramping (which could have pointed to a connection).

I thought this was the crystal oscillator end but looking at Dave's teardown, this end is all SMPS circuits on both sides of board.  This also makes me very happy because mothers of BGA packages may not be involved   :)

So now, I'm leaning or at least open to the possibility that there may be a power supply problem.

Why, why, oh why, can't an instrument in this price range give a complete dump of system status and supply voltages ?   :bullshit: :bullshit: :bullshit:

So while I wait for comment from Keysight, I'm holding out hope for a simple solution for the New Year !

« Last Edit: December 25, 2021, 05:37:01 pm by EE-digger »
 

Offline electr_peter

  • Supporter
  • ****
  • Posts: 1443
  • Country: lt
Re: Keysight 3000 4000 MSOX4154A P-CLOCK failure
« Reply #11 on: December 26, 2021, 04:24:32 pm »
Agilent/Keysight DSOX2000A, MSOX2000A, DSOX3000A, MSOX3000A (possibly DSOX3000T and MSOX3000T) had issues with internal SMPS. Based on your experience, SMPS issue remains :(
Forum topic dso-x-3024a-power-supply-defect
« Last Edit: December 26, 2021, 10:56:03 pm by electr_peter »
 

Offline EE-diggerTopic starter

  • Frequent Contributor
  • **
  • Posts: 391
  • Country: us
Re: Keysight 3000 4000 MSOX4154A P-CLOCK failure
« Reply #12 on: December 31, 2021, 03:30:08 am »
I don't think the problem is with the main supply (at least not yet).  But thanks for the link as it's a real eye-opener for the 3000 series, along with good alternative solutions.
 

Offline JacobPilsen

  • Regular Contributor
  • *
  • Posts: 149
  • Country: cz
Re: Keysight 3000 4000 MSOX4154A P-CLOCK failure
« Reply #13 on: May 22, 2024, 04:34:32 pm »
there may be a power supply problem.
Have you made any progress???
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf