Author Topic: Agilent 54835A scope (4 channel 1GHz / 4Gs/s) repair & uphack (Read 99286 times)

Converter · « **Reply #350 on:** November 26, 2019, 05:38:59 pm »

I don’t understand why you suffer so much time? This motherboard today is not worth $5. Do you have problems to buy another standard PC motherboard format Baby-AT with a processor socket even a little newer (for Pentium III) than you have now? See my posts:

https://www.eevblog.com/forum/testgear/agilent-54835a-scope-(4-channel-1ghz-4gss)-repair-uphack/msg1407288/#msg1407288
https://www.eevblog.com/forum/testgear/agilent-54835a-scope-(4-channel-1ghz-4gss)-repair-uphack/msg1385418/#msg1385418
https://www.eevblog.com/forum/testgear/agilent-54835a-scope-(4-channel-1ghz-4gss)-repair-uphack/msg1346487/#msg1346487

nctnico · « **Reply #351 on:** November 26, 2019, 05:53:18 pm »

That is also an option; I forgot about that. IMHO the FIC motherboard never was a high quality choice.

analogRF · « **Reply #352 on:** November 26, 2019, 06:02:46 pm »

no, just my hakko iron and lots of flux and solder wick after that
it takes a lot of time but does the job...

analogRF · « **Reply #353 on:** November 26, 2019, 06:06:11 pm »

Quote from: nctnico on November 26, 2019, 05:53:18 pm

That is also an option; I forgot about that. IMHO the FIC motherboard never was a high quality choice.

yes these are really a POS. I am surprised why Agilent chose to put these into their highest end scopes at the time
these are not industrial grade board, not even a consumer grade, just garbage...

analogRF · « **Reply #354 on:** November 26, 2019, 06:08:09 pm »

Quote from: Converter on November 26, 2019, 05:38:59 pm

I don’t understand why you suffer so much time? This motherboard today is not worth $5. Do you have problems to buy another standard PC motherboard format Baby-AT with a processor socket even a little newer (for Pentium III) than you have now? See my posts:

https://www.eevblog.com/forum/testgear/agilent-54835a-scope-(4-channel-1ghz-4gss)-repair-uphack/msg1407288/#msg1407288
https://www.eevblog.com/forum/testgear/agilent-54835a-scope-(4-channel-1ghz-4gss)-repair-uphack/msg1385418/#msg1385418
https://www.eevblog.com/forum/testgear/agilent-54835a-scope-(4-channel-1ghz-4gss)-repair-uphack/msg1346487/#msg1346487

the scope is not mine so I have to check with the owner but certainly he does not want to alter the back panel and drill holes and stuff...
so if there is a board that perfectly replaces this POS and is cheap then yeah, I think he will consider it.

Converter · « **Reply #355 on:** November 26, 2019, 06:33:10 pm »

Quote from: analogRF on November 26, 2019, 06:08:09 pm

the scope is not mine so I have to check with the owner but certainly he does not want to alter the back panel and drill holes and stuff...
so if there is a board that perfectly replaces this POS and is cheap then yeah, I think he will consider it.

Perhaps you did not attach importance, but the keyword in my post was "Baby-AT".

Converter · « **Reply #356 on:** November 26, 2019, 07:25:40 pm »

analogRF · « **Reply #357 on:** November 27, 2019, 06:25:50 pm »

alright, let me see, I cleaned up all the boards and connectors everywhere thoroughly again, reset the CMOS and then set it up according to page 5-48 of the Agilent service manual like before, changed the RAM, etc, replaced a few more of the 10uF caps...and again no luck

I could not move the PCI cards around because they hit the cpu heatsink.

I am sure the BIOS does not even attempt to access the boot record on the booting device whether it is floppy or HDD or CD (I connected a cr rom to the ide channel to check). Something else is stopping it before booting the OS. Based on my past experience when I used to build PCs about 17-18 years ago, this is usually due to a HW conflict of some sort which leads me to believe the BIOS settings are not right even though I followed the Agilent service manual (load Setup Defaults, then....)

Can somebody who has this board (FIC VA-503A) please check the bios settings and post a couple of pictures so I can compare? specially the second page which is called BIOS features and the one that is called chipset features and also Integrated Peripherals.

I have played many times with bios settings but only those that were related to IDE or booting or PCI devices, not the others because I thought setting them up based on agilent service manual should be enough.

nctnico · « **Reply #358 on:** November 27, 2019, 06:31:24 pm »

I guess you already tried to set the BIOS settings to 'defaults'?

analogRF · « **Reply #359 on:** November 27, 2019, 06:35:49 pm »

Quote from: nctnico on November 27, 2019, 06:31:24 pm

I guess you already tried to set the BIOS settings to 'defaults'?

yes, as per instruction in agilent service manual, it starts by setting to default and then it shows some of the settings that must be specifically set
but they are actually set already by default. you can look it up in the service manual

analogRF · « **Reply #360 on:** November 27, 2019, 06:39:31 pm »

I am worried that maybe because of the bulging leaky dead caps when the instrument first powered on (after several month of inactivity)
maybe one of the devices/chips on the board could not tolerate the surge or ripple of whatever and died...i don't think it is the cpu and i am sure it was not he memory stick but other chips...just a bad thought

nctnico · « **Reply #361 on:** November 27, 2019, 06:43:09 pm »

If any of the electrolyte got spilled on the board then it may create a short(ish) or have eaten away a trace. IMHO you are probably better of by installing a new motherboard as Converter has suggested. You have already tried every 'simple' solution.

Alfons · « **Reply #362 on:** November 27, 2019, 06:43:58 pm »

If the BIOS is set to default, the device boots in any case. After the Cmos battery was empty on my board, the bios was also on default settings. I just had to set the time. The fact, that the device did not boot is not due to the bios settings, but probably because the IDE controller has a defect (if the hard disk is OK).

analogRF · « **Reply #363 on:** November 27, 2019, 07:01:15 pm »

Quote from: nctnico on November 27, 2019, 06:43:09 pm

If any of the electrolytic got spilled on the board then it may create a short(ish) or have eaten away a trace. IMHO you are probably better of by installing a new motherboard as Converter has suggested. You have already tried every 'simple' solution.

luckily there was no spillage on the board at all, the caps had leaked from the top of their head

yeah I am looking at those baby AT boards, the thing is that he does not want anything changed at the back of the scope
so all the cards must sit in the same positions as they are now. By looking at the pictures of these boards I cannot
say which one would allow this

EDIT: stil pictures of the bios settings is appreciated

charlyd · « **Reply #364 on:** December 15, 2019, 03:40:02 pm »

does the board ramdomly reboot? i mean if you pull all cards and power on the mainboard does it post or even then reboot? you can try with a seperate PSU

alpher · « **Reply #365 on:** February 13, 2020, 05:03:13 pm »

OK. So I sort of resurrected my 54845A project, managed to install new motherboard and bring the system back from the dead. Unfortunately wasn't able to install the 4.5B version of software

.
Anyway back to the 4.3A we go, up and running (was running uppgraded to 4.5 to).
There is a problem though, while before the scope passed all the self tests without any issues, also the calibration went through multiple times succefully, now it somehow developed "logic trigger" faults in all 4 channels during the cal process.
Self test fails on "logic trigger" and "pattern trigger".
So began to readup on the contents of HPX4S help file that someone suggested as good source to start board level troubleshooting.
Attached are pdf versions of the relevant sections from that help file, attached here because somehow microsoft made them really difficult to open in windows 10

.

The only immediately usefull info I foun there is as follws:
"
Error Symptom   Diagnosis
The logic trigger test is somewhat self contained but does depend upon the serial input port and the 100 Mhz startable oscillator. The startable oscillator is used for pattern duration and range. To check the 100 Mhz startable oscillator:
1. Check the enable line toggles during the test.   U84 pin 15
2. Varify that the oscillator is running during the test.   U84 pins 18,19
The logic trigger also provides holdoff for the trigger system. A 50 Mhz free running oscillator is used as the holdoff clock.
1. Verify that the 50 Mhz osc. is running.   U81B pin19
"

Has anyone been there before?
I've tried to scope signals at these pins but those MECL chips are pesky (negative supply) hard to trigger on the signal properly.
Moreover the signals are on only during 'logic trigger" test which last all but few seconds, is there a way to loop the test so it runs continuously ?
Really appreciate any poiners here.

I've attached pinouts for both chips U81 and U84 , 10H117, 10H105 as they are hard to come by for PLLC20 package.

alpher · « **Reply #366 on:** February 17, 2020, 03:21:51 am »

Decided to start from scratch, used 3.5 rec.iso and ghosted the image on another CF card, everything goes peachy till I reboot the scope.
It starts to load win98 then right away flashes DOS error message to the tune "error writing device PRN, abort , retry, fail" ?

At this moment I cannot even type, keyboard seems to locked up.
Has anybody encounter such a thing?
Could be that the 3.5rec.iso from Tony's archive is corrupted somehow?
I'm honestly stumped.

alpher · « **Reply #367 on:** February 18, 2020, 12:07:59 am »

Short update, the "cannot write to PRN" error occurred for me on 2 motherboards, an Intel ZX chipset and Via VP3 . Somehow managed to install 3.5 recovery on Intel i810 mobo without hiccup. Unfortunately my conclusion is that something must be borked on my acqusition board. Managed to probe the signals on both the ECL chips (U81, U84) and they're there during the test.
Still all 4 channels fail on logic and pattern trigger, doesn't matter the mobo, be it i810 and software ver. 3,5 or ZX chipset and soft. 4.3 or 4.5, the result is the same.

So if someone has an extra A13 acqusition board, preferably cheap one, without FISO's or attenuators , as mine are working fine, I'' be willing to buy or trade for something else ( have a couple of sets of A5, A6 boards, maybe other stuff).
Thanks again.

nctnico · « **Reply #368 on:** February 18, 2020, 12:10:16 am »

Did you take the acquisition board out and inspect it? On mine one of the chips came loose. Be sure to work in an ESD safe environment otherwise you might do more damage than good.

alpher · « **Reply #369 on:** February 18, 2020, 12:14:50 am »

I did take it out a couple of times, didn't see anything obvious even under a magnifying glass. Without any more board level info I don't think that there is a chance of repair.

charlyd · « **Reply #370 on:** February 18, 2020, 09:59:54 pm »

Here are the screenshot where analogRF askes for in post #363, sorry just came across this post. Like NCTnico said don t forget to first default the bios settings.

screenshots are from a repaired 54845A (FIC VA-503A)

JDubU · « **Reply #371 on:** February 20, 2020, 02:33:37 pm »

I think that "error writing device PRN, abort , retry, fail" is a Windows boot problem, probably associated with the motherboard BIOS settings and not with the scope electronics.
It can occur when a DOS program is trying to print to the default printer and there is no device to print to. In DOS days, printing was done to a parallel port so make sure that the parallel port is enabled in the BIOS.

alpher · « **Reply #372 on:** February 27, 2020, 05:59:09 pm »

I've ordered a used acquisition board on eBay, we'll see when it comes.

alpher · « **Reply #373 on:** March 31, 2020, 11:10:53 pm »

Now I've had time to sober up, I've finally managed to put my 54845A project back together. Replacing the acquisition board fixed the logic trigger issues that i had.
The board I bought on ebay came with a full set of A/D converters(FISOs) but no attenuators. No problems for me as attenuator assemblies on my original A13 board were fine, passing all the selftests. Luck had it that 3 out of 4 FISOs calibrated without an issue, only channel 3 wouldn't. It will pass all the self tests but would fail the self calibration routine. As I couldn't find a way to try to calibrate only one channel and the whole calibration routine is quite long, I'm not sure if this was a defective A/D converter or the socket. I did try to clean the contacts on the chip and the socket it didn't help, so I just replaced chip and socket from my original board.
After that it all went smoothly, scope passed all the selftests and calibrated just fine.
Upgraded the OS first to windows98SE then windows ME, mostly to get a decent USB support as I also installed a supermicro slimline USB panel instead floppy drive.
A ha, forgot to mention that the scope runs now of 8GB CF card mounted on the rear bracket and the mobo is an intel ZX chipset with P3 933MHz, and 512MB ram, quite a decent machine for win ME.
I'm running the scope application with /noram /internal parameters as it seems to be booting much faster this way, gone is also config.sys with the ramdrive.
All in all doing away with ramdrive cuts the boottime by~30%, switching to windows ME cuts that still in half.
Anyway now that its all done I'm left with quite a bit of extra parts, if any one is interested I have :

one A13 acqusition board with 3 good FISO's and one most likely bad (there is a slim chance it could be a socket), no attenuators. (it has that logic trigger issue)

2 x A5 boards that I burned during testing and consequently repaired once I got my hand on a supply CT t65550 chips. both running fine .

2 x A6 interface boards also running fine, got them together with the A5 boards, on one of them I had to replace 60 pin connector , couldn't find the original so I installed just the .05" 60 pin header and it works fine.

Plus misc other small items here and there, some MECL chips and couple of CT vga chips as well.

alpher · « **Reply #374 on:** April 01, 2020, 01:56:39 am »

Couple more pics:


EEVblog Main Site	EEVblog on Youtube	EEVblog on Twitter	EEVblog on Facebook	EEVblog on Odysee

Author Topic: Agilent 54835A scope (4 channel 1GHz / 4Gs/s) repair & uphack (Read 99286 times)

Share me