Agilent 54835A scope (4 channel 1GHz / 4Gs/s) repair & uphack

[nctnico]

Perhaps some electrolytic capacitors turned bad.

[analogRF]

Perhaps some electrolytic capacitors turned bad.

That was also my suspicion but where...it's been sitting on the corner for 2 or 3 month only. caps on the motherboard at least look ok ...PSU voltages are all good too...

[Converter]

The most perishable product is aluminum capacitors with liquid electrolyte. Check the voltages stability from the power supply using a different oscilloscope in roll mode. Make sure that the ripple is normal. Some contacts may also oxidize, especially if you store it in a humid environment. Check connectors.
Check if a forced shutdown signal is received from the oscilloscope control circuit, or the protection in the power supply unit is activated against overcurrent, temperature, or a signal from the supervisor for voltage.

[Jwalling]

what do you think could be wrong? I have been dicking around with reseating PCI and ISA cards and RAM sticks but it keeps rebooting randomly. I don't think it is heat related because sometimes it happens right after cold boot before even win98 starts.

Try running memtest86+ on it.

[analogRF]

First of all my own 54845A which was randomly rebooting seems to be working fine again...I ran memtst86 for an hour and it didn't find any error. I had cleaned up all the board edges, connectors, etc...and I checked all the main filter capacitors on the PSU and on the mobo and they were all very good even better than the brand new ones I was going to replaced them with! (they were very good both in terms of ESR and value)
but after playing with it for a couple of days and keeping it running eventually the reboot problem went away and it has been working 1-2 hours at time for a while now. Maybe there is a bad cap somewhere that was about to die when the unit was stored for a couple of month and now that it has seen voltage across it, it has come back to life (and it will probably die again) but I don't know where that could be...I suspect one of the caps in the PSU but not in  the output filters, perhaps on that control daughter board which is soldered on to the main PSU board...

However, I am now dealing with another 54845A which is not mine. Belongs to a colleague who is not into repairing stuff although he does a lot of electronics specially RF...This is a VIN 033 model made around 2000 with FIC VA-503A motherboard (newer than mine) with one stick of 64MB SRAM and K6/400 CPU. Based on his description the unit had been working until the day that it was put aside about 5-6 month ago and now it does not boot. It stops at "Verifying DMI pool data ...." and nothing after that. Recognizes the HDD and I have checked the HDD outside the unit with chkdsk /r and no errors. I also connected the original HDD of my own older unit to it just to see if it boots.
There were tons of leaky bulged elcos on the motherboard which I replaced them all (all the 1000uF and the three 220uF and only 2 of the many 10uF but not all of them. The 10uF caps are not bulged or anything) but nothing changed...I tried to boot it with only the HDD, interface board and the video board and still a no go....tried different SRAM sticks...

any idea what else I should be looking for? All voltages on PSU are accurate and I have cleaned all the connectors...changed cmos battery and played with cmos many times...tried to boot it froma DOS floppy in the LS120 zip drive but still the same issue...

I am out of ideas now....

EDIT: any help is appreciated. I know there are people on this thread that have worked on these infinium series a lot more

[analogRF]

any ideas?
nctnico, I am counting on you 

[nctnico]

any ideas?
nctnico, I am counting on you 

It beats me. I'd try to take all the cards from the motherboard and use a standard PCI videocard to see if the problem is in there. Perhaps try to connect a speaker to the motherboard and check for beeps. There are also diagnostic cards which you can plug into a slot but I never used these.

One thing you could do is take the battery from the motherboard, short the battery terminals to make sure the CMOS memory is powered down fully. There may be a jumper on the motherboard for this purpose.

[analogRF]

i dont have a video card but I just took the CF card and adapter that I had put in my own unit (and of course it is working) into this unit
to replace its hard drive. NO change  the BIOS recognizes the CF card and reports its size and everything though

Actually that little CF card IDE adapter has a data LED on it that flashes when there is access to the CF card and it flashes when bios is detecting the IDE devices. However I noticed that after that "Verifying DMI pool data..." when the win98 should start, there is no access to the CF card at all
unless it flashes for like a few micro seconds that I cannot see but I don't think so, it appears that it never tries to access it really. It seems the BIOS gets lost at that point...is it possible that the BIOS chip is gone bad somehow after sitting idly by for a few month? 

[2N3055]

Remove battery and try booting (without it). On some old PC motherboards empty battery would prevent boot. It would behave exactly like that.
Of course, there is a chance of corrupted BIOS...

[analogRF]

battery is brand new...i have reset CMOS 100 times even shorting the battery holder pins

but the unit had been working until few month ago and after that it was sitting in the corner of his workshop (I had seen it in his workshop myself when it was working and also when it was put aside) how the bios could get corrupted? and exactly at that point when it is supposed to load the OS. everything before that takes place as it should

By the way, I notice that in the device table, it shows an unknown device (which I suppose is the interface board) and its IRQ is NA? everything else has a name and proper IRQ number. Is that normal?

EDIT: my mistake, the interface board has IRQ=10 and the display board has no IRQ number and that is exactly how it is on my own working unit. Still stuck at Verifying DMI pool data....I am out of ideas pretty much except maybe programming the BIOS chip?!  but why? and with what BIOS file? the one in the chip does not seem to exist anywhere on internet...

[analogRF]

connected a speaker to the mobo, there are no beeps until right after the "Verifying DMI pool data...." and I get one short beep
and then nothing.

shouldn't there be any beep earlier like after or before RAM test or right after recognizing IDE devices?

[nctnico]

Can you check whether the disk drive is working? Suddenly I recall I had a mysterious boot problem when I disconnected it in my scope. It is a long shot but worth trying.

[analogRF]

Can you check whether the disk drive is working? Suddenly I recall I had a mysterious boot problem when I disconnected it in my scope. It is a long shot but worth trying.

I tried 3 different disk drives including the one from my own working 54845A. I also tried the HDD of the faulty unit on my 54845 and it boots
I tried booting from floppy to DOS but it seems that it just does not access the boot device at all 

have tried different RAM sticks in different slots...all PSU voltages are good...

"Verifying DMI pool data...." and one short beep 

[nctnico]

Did you Google yet?

I have found this:
https://forums.tomshardware.com/threads/stuck-at-verifying-dmi-pool-data.501230/

It seems to have something to do with the boot device order.

[analogRF]

yeah, I have been googling this thing for several days...i have seen that page,  too...have tried pretty much all their "solutions" ...

these people usually have done something to their system and then it failed to boot or they built a new system

the thing that is most irritating is that this unit had been booting fine before it was set aside for a few month and now it doesn't boot
nothing has happened to it during this time...

I am gonna clean the connectors for the 3rd time again and also move over the video card and interface board one slot (or 2 slots if possible)

I also tried another video card that I had for these scopes by the way, no luck. I do have another interface board but it is an older
model and the ribbon connector (that goes to the acq board) does not fit...but what the hell can be wrong with that board anyway?

I cannot even boot it from floppy to DOS. I will try to find a regular floppy drive (instead of the LS120) and try with that and see what happens
I must have one laying around somewhere....

EDIT: I have also moved the HDD to the secondary IDE channels just for the heck of it....of course the boot device and boot sequence in BIOS have been checked and re-checked 100 times...

[AJ3G]

Hello:

I am not sure if you are the same person, but I spoke to someone on the HP board regarding an almost identical problem. I too had the error you speak of, and it was resolved by replacing the HDD. I pulled the old mechanical drive for a solid state drive.

Your next step seems like a good one. I had a SRAM fault, which was intermittent. It seemed to work fine for about 1/2 hour and then the display would go on the fritz, followed by the SRAM error and some really weird relay contact counts. I finally took the scope completely apart, cleaned all the connectors, and card edges, and the the scope is functioning again. I have had it running now for three days without issue, so I hope the issue is corrosion on one of the connectors.

These scopes seem to be a bit finicky. I hope this helps..

Rich

[analogRF]

yes I am the same person 
but if you read my posts, I have already gone through all those things more than once...
playing around with RAM sticks and their place, using multiple known good hard drives and flash card drive, cleaning all boards and connectors 2-3 times....but what the hell, i;ll do that again 

[nctnico]

Maybe the CPU got bad. Does the heatsink fit properly?

[analogRF]

Maybe the CPU got bad. Does the heatsink fit properly?

the cpu is stuck to its fan as you may have noticed yourself since you had the same model, and yes it fits properly. it is barely warm...
why would the cpu die on its own sitting in the corner? besides i would imagine the bios would complain if the cpu was bad...
during capacitor replacements I took it out just to be safe.

[AJ3G]

That is annoying. It would seem to me if the final attempt at cleaning does not resolve the issue, the Motherboard maybe malfunctioning and need to be replaced. Not great news, but not the end of the world either.

[analogRF]

that crappy motherboard costs a lot of money these days so i doubt he will want to buy a motherboard for it...when it didnt boot, he was going to sell it as is, i convinced him that i can repair it and it must be a piece of cake and here we are   

My gut tells me it must be a very simple thing that I am overlooking...

if push comes to shove, I am going to program a new BIOS chip with one of the BIOS files on the FIC ftp server and see what happens

anybody knows what is the BIOS EPROM model? it has a sticker on it I dont wanna remove for now...

[AJ3G]

I forget where, but I thought there was some online documentation suggesting that other boards may be used, but the chassis would have to be modified (i.e. drill new holes). Yes, I agree buying the same motherboard would be expensive, but using an alternate would not, provided you were willing to make some modifications to the sub chassis.

I do not know the BIOS EPROM Model off the top of my head. Later today I can take a look if you like.

[nctnico]

Let's do a recap:
- Original problem is that it is stuck at 'verifying DMI pool data' and this didn't change. So nothing you did made the problem better or worse.
- Getting stuck at this point hints towards a BIOS / boot problem problem but it could be anything.
- All the contacts have been cleaned (I assume including the configuration jumpers) so let's rule out a contact problem.

Perhaps you need to replace more capacitors. I assume you have used low-ESR types from a reputable source to do the replacements earlier.

[analogRF]

yes, your recap is spot on. Configuration is done by microswitches which I sprayed IPO and switched them several times (even number of times!) and yes made sure they are exactly where they were before and match the 54845 service manual , so are the BIOS settings

I have replaced all 1000uF and 220uF caps (all with high quality brand new Nichicon and Panasonic) those were the bulging caps and most of them part of dc-dc converters

then there are like 20-25 tiny 10uF bypass caps all over the place which I randomly chose 2-3 of them and replaced

but taking the caps out of this multilayer board is really a nightmare. 10uFs were not that bad although not as good as the new ones i installed
maybe there are a few of them somewhere that are really bad? but they dont seem to be that huge of a deal that can cause such problem

i can try to replace more of them if you guys really think that could help

[nctnico]

Did you use hot-air during removal and fitting of the capacitors? In my experience this is the only way to get a capacitor replaced on a PC motherboard.