Agilent 54845A Failing all trigger self tests

[TRN]

A while back I bought a 54845A on an auction, and this scope had one bad channel, due to a faulty ADC-Hybrid.
So I bought a used one, and installed it in the scope, and the bad channel is now O.K.

However the scope fails all the Trigger self tests.
I recovered the scope software to (V4.30) which appeared to have fixed the problem, but when I then upgraded to V4.50, the problem reappeared.
So I did another recovery to V4.30, and this time it did not fix the issue. A subsequent upgrade to V4.50 did not change anything.
Can Anybody give me any clues as what the problem is.
I have attached a PDF which shows thState Trigger, Pattern Trigger, Logic Trigger and Trigger level errors

Thanks

[0xfede]

Hello TRN,

I don't know if this is your case (and I never had this scope) but I remember that there is a nice video from the Signal Path concerning the repair of a 54845A:

Basically he find that poor contact between hybrids and the custom socket was causing intermittent faults.

Hope that helps you.

Best,
0xfede

[Tony_G]

I think you have a software configuration problem - The 54845A shouldn't see those tests in the self test menu - You should only see "Knob and Key", "LED" & "Scope Self Tests" in that menu.

Try completely reinstalling the initial image - Did you get it from my OneDrive?

TonyG

[TRN]

0xfede,

Thanks for your input, and the link for the video, and i have already seen this one, and have cleaned the hybrid contacts.
This did however not solve the issue.

[TRN]

Tony_G,

Thanks for your input, and if I do not mark the service extension box, I get the standard test menu which you mentioned.
However when I do mark the box, I can test all the subsystems separately.  I have attached a screenshot of the successful standard self test after the first recovery of the v4.30 software.
I downloaded the v4.30 recovery and v4.50 upgrade software from the Keysight ftp, so I assume it should be O.K.

Nevertheless I will download your version, from your onedrive, and will do a complete reinstall.
I will let you know how it goes.

Regards

[TRN]

Tony_G,

I tried to locate your OneDrive, but I can't find it.

Can you please send me the link?

Regards

[TRN]

Tony_G,

Just forget my previous post, because I found the link. (https://1drv.ms/f/s!Amqar8_XQ9Uzj6YhN1xHdM8tQdKMOA)

Regards

[Tony_G]

You're right - I missed the service extension checkbox - I still think that the unit might be thinking it is a MSO though so reinstalling per the Agilent documentation is worth a try. - I did a video on it here:

https://youtu.be/L49qug1oGuo

https://youtu.be/x9FUXcWENoU

TonyG

[TRN]

Thanks for the video, and I looked through my stuff but I didn't have an IDE CD-Rom on the shelf anymore, so I have ordered one, and should get it next week.
I will let you know how it goes.

Regards

[nctnico]

I tried to run the same tests on my 54845A (uphacked) but it passes all trigger test. I don't think your problem is a software issue because the errors don't refer to logic inputs. I suggest opening the scope and check the acquisition board carefully for bad solder joints. In my 54845A I found a chip which had come loose from the board!
This is the thread I started:
https://www.eevblog.com/forum/testgear/agilent-54835a-scope-(4-channel-1ghz-4gss)-repair-uphack/msg1251605/#msg1251605

This guy also has quite a bit of info on the internals and I recommend to check all the power supply voltages:
http://www.dasarodesigns.com/projects/fixing-an-hp-agilent-54845a-infinium-oscilloscope/

[TRN]

Thanks for the advice, and I already checked the board for soldering flaws, but it is quite possible that I have overlooked a bad soldering joint.
Anyway, i should get the IDE drive latest by friday, and if restoring the firmware doesn't work, I will do another check.

Regards

[TRN]

Hello Tony,

It took a bit longer then expected because the HDD went defect, and I had to buy a new 40GB IDE drive. I burned a CD from your 5484xBRecovery.iso  file (inside directory\54845\548xx\Upg_Rec\5484xB\) but the iso file is corrupt.
So I used the  v4.30 recovery and v4.50 upgrade software which I previously downloaded from the Keysight FTP.
The result is the same meaning that none of the trigger events passes the self-test.
So I followed nctnico's advice and have checked the board for any soldering flaws, but haven't been able to find any so far.

I also read the article from dasarodeigns.com, (thanks for the link nctnico)   
I checked all the Voltage regulators and except forXU5 and XU19, they all match the values listed.
XU5 and XU19 measure respectively +7.09V & +7.12V instead of +8.5V, but I don’t think that this interferes with the scope functionality, because it looks that they only supply power to the two large cooling fans.
   
I also checked the attenuator connector pin-outs, and there some differences with the values listed.
    J7 & J9:
    Pins 20: (+ 1.61V with attn installed); (+1.71V w/o attn installed)
    Pins 16: +0.845V
    Pins 8: (-2.78V with attn installed); (+1.8V w/o attn installed)

    J6 & J8
    Pins 20: (+ 1.61V with attn installed); (+1.71V w/o attn installed)
    Pins 8: (-2.78V with attn installed); (+1.8V w/o attn installed)
    Pins 11: +0.845V

    Pins 16 from J7 & J9: and Pins 11: from J6 & J8 are all fed from the same regulator U45 which is an LT 1086. which outputs +    0.85V, so I am puzzled how can it be that one should measure +3.9V? on pins 16 & 11?

   any ideas are welcome

[nctnico]

In the software directory your can find various help files. Look for HP4XS.HLP. This contains some information on how to diagnose the scope at the board level. It isn't a complete service manual but it might contain enough information to get you in the right direction.

[Tony_G]

    Pins 16 from J7 & J9: and Pins 11: from J6 & J8 are all fed from the same regulator U45 which is an LT 1086. which outputs +    0.85V, so I am puzzled how can it be that one should measure +3.9V? on pins 16 & 11?

Datasheet for that LT1086 is here. Can you confirm that it is really supposed to be 0.85V? That doesn't seem right as that part is supposed to have a 1.25V different between Vout and Vadj. What are the R values for voltage divider on that part?

TonyG

[TRN]

nctnico,

Thanks for the tip, however I have a hard time finding the HP4XS.HLP file on the scope, I also tried to extract the file from the Ghost image using GhostExplorer, but I was unable to do so because GhostExplorer crashes every time I try to extract anything,

Can you send me a copy?

regards.

[TRN]

Tony,

Thanks and you were right, about the regulator.

The divider values are R1= 120 ohm; R2= 260 ohm. This means the regulator is set to output 3.96 volt, and I check Uout across the divider network, and the output is 3,96v. R2 however is not connected to ground, but to a node which is at -3.1v from ground, so that's why I am measuring 0.85v between the regulator output and ground.
In a nutshell, the regulator is O.K., but I still am unable to figure out how can it be that one should measure +3.9V? on pins 16 & 11?

[nctnico]

nctnico,

Thanks for the tip, however I have a hard time finding the HP4XS.HLP file on the scope, I also tried to extract the file from the Ghost image using GhostExplorer, but I was unable to do so because GhostExplorer crashes every time I try to extract anything,

Can you send me a copy?

See attachement:

[Tony_G]

I find it interesting that there is supposed to be a -3.1 rail and you're reading 0.8 instead of -3.9.

Might be a simple coincidence but it might be an indication of something wrong with the rails. Is pin 16 meant to be the -3.9v rail? I believe that it is but I've never checked on my scope (I had issues with the attenuators but not the power supplies).


Pity we don't have a schematic.

TonyG

[TRN]

nctnico,

thanks for the help file, now I only need to find a PC which runs Win XP or 7 

Tony,

Thanks, and according to Matthew Dasaro, pin 16 is the -3.1v rail for the analog Hybrid. I will do some more board probing to see if this correct, and if I can make some sense out of this.

Anyway, I agree with your comment about the schematics, and I really do not understand that these are not available somewhere. I mean HP/Agilent designed this scope, so they must have had the schematics.
I am quite sure that there is a retired HP design engineer somewhere who has the whole documentation package laying on his desk with a pile of dust on top of it .

Anyway I will get back once I have any news

[TRN]

I am still trying to fix the the trigger self-test errors, and I am looking for the pin-out of Motorola MC10H105 and MC10H117 Ic's in a PLLC-20 case.

The ON Semiconductor datasheets only show the pin out for 16 pin dual in line packages.

Can anybody tell me where I can find the ones for the PLLC-20 case?

[MarkL]

There's a standard conversion for pinouts between ECL packages.  See attached.

The full app note is here (good reading if you're interested in ECL design):

  http://www.onsemi.com/pub/Collateral/TND309-D.PDF

[TRN]

Mark,

Many Thanks

[rickells]

hi,

My Agilent 54845a has  Trigger self-test "Failed"  on all three of the Trigger tests;  for all 4 channels.
The scope does power On & can measure the Sq. wave signal from the Aux Out, & other signals Ok, (with knobs twiddled manually).
  It passes all of the other self-tests, .... just Not any of the 3 Trigger tests.

I have cleaned the ADC's, examined the Acq board for solder/IC placement issues, but see none.

Over the years there have been a few individuals who had the same issue (JayW, JWalling, TRN) & perhaps others, but details seem not to be found...

   TRN seems to have started to examine voltages & signals ...
       somewhat difficult with no schematics in the service manual.

1)  -- How was this Trigger problem fixed ??

thank you,
rick

Previous references to Agilent / HP 54845a & 54831b or 54832b Trigger problem:

https://www.eevblog.com/forum/repair/sometimes-knowing-what-something-means-is-half-the-battle-agilent-54831b-error/msg1008382/#msg1008382
https://www.eevblog.com/forum/testgear/agilent-54845a-failing-all-trigger-self-tests/
https://www.eevblog.com/forum/testgear/agilent-54835a-scope-(4-channel-1ghz-4gss)-repair-uphack/
https://www.eevblog.com/forum/repair/agilenthp-54845a/msg1267378/#msg1267378
https://www.eevblog.com/forum/testgear/agilent-54831d-modernising/


https://community.keysight.com/thread/6575
https://community.keysight.com/search.jspa?q=54845a
https://community.keysight.com/message/15068

http://www.dasarodesigns.com/projects/fixing-an-hp-agilent-54845a-infinium-oscilloscope/

https://wonghoi.humgar.com/Hobby.htm

[Alfons]

If you're still interested: I've upgraded a 54815 to a 54825. After that, the device wanted a recalibration. All channels showed Trigger Fail. After a long search, I found this thread: https://www.eevblog.com/forum/testgear/_fdc-failure_-on-scope/ and the contribution of free-electron:

i may have that laying around. i definitely have the ls drives. the machine has the adapter already in it. simply unplug the floppy and floppy cable. follow the floppy cable. it goes to one of the PCI boards. and from the pci board there is another to the motherboard. ( not the IDE ! , the floppy ) unplug both those cables. ( not the IDE on the pci card ! )
look again on the pci card. you'll see a the ide being translated to the fine pith ide of a laptop drive. unplug the large IDE cable and install an IDE cable with a split connector.

hmm. i realize it is hard to describe in words if you 've never done this. leme take a few pictures. i'm working on a 825 right now with a failed attenuator for a friend of mine. just replaced it running cal now. i'll snap a few pictures so you see how it is installed.

running self calibration is a bit tricky on these machines. the service software has some quircks.
you need to run probe cal first ( vertical menu ). disconnect everything. set attenuators manully to 1:1 on ther vertical menu. then run probe cal and follow the prompts. you need a short bnc ( less than half meter). the scope uses an internal 32 bit ( yes 32 bit ... ) dac to cal the input.

once that is done you can do the real cal. make sure to let the machine heat up for half an hour prior to doing that. you may get trigger failures otherwise.

the LS drives are matsushita or mitsubishi:

http://www.ebay.com/itm/HP-3-5in-359007-001-LS120-SuperDisk-Drive-MF357H-453MG-/370811752744?pt=US_Floppy_Zip_Jaz_Drives&hash=item56561a5128

thats the exact one with the right connector. there is others with different connectors but those don't work as they dont fit the adapter

The device ran for an hour before the test. And I also ran the ProbeCal for all channels before the calibration. That worked then. All channels could be calibrated without error.

But also before I had let the device warm up. I think that's actually a software bug in many cases when all channels report Trigger Fail.

[ARF]

The same scope, the same problem.

As soon as I received it, I did a self test and the self test passed well.
After that I tried to do a calibration - but the calibration failed with an error on channel 3.
The next day I decided to repeat the calibration, but I received errors on all channels.

I did the self-test again and it already failed with 4 trigger errors.