EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: FrankenPC on May 18, 2014, 10:42:15 pm
-
According to the service manual, the in-guard MCU HP partnumber 1820-2539 which is an AMD AM8048 u-processor, is dead.
I'm not entirely sure it's the MCU, but it's worth swapping out if I can find a cheap replacement. I've google'd it. Can't find any suppliers. Any ideas where to look? I'm also looking for a parts unit on Ebay. No good ones yet.
Is it pin/voltage compatible with the Intel 8048?
Also, am I a fool for trying to diagnose this particular dmm?
-
Also, am I a fool for trying to diagnose this particular dmm?
Absolutely not, many here have them and love them. You'll need to pay a fair bit of money to get new equipment with the same specs.
That said, I first followed the service manual fixing mine (a front end problem) and it completely lead me down the wrong path. Perhaps if I was a HP tech with draws full of parts I could easily swap out then the manual probably would have gotten me there in the end.
So I'd suggest the more appropriate way is to go back to normal fault finding techniques. First check all the rails, then if it's a digital problem start poking around with a scope and see if any ICs aren't responding.
-
First of all:
Check the rails on the PSU, especially the ripple on the 33V rail, since I would bet that the caps there have failed, as they have done so often in other 3456As.
Then:
You are probably still f*cked, because even if you manage to get a second Intel/AMD 8048, it will be mask programmed and thereby highly unlikely contain the same hex file as the probably failed Am8048 inside your 3456A.
So, in your case, I would really make sure, it's the AM8048 on the inguard board and not some kind of other failure, e.g failed ROM or RAM on the CPU board. BTW, what happens if you power up you 3456A? Any meaningful signs of life on the LED display. If so, the CPU board is probably OK. Another good idea to repair old stuff in general is to get a second instrument, since you can swap boards easily and thereby diagnose the failure on board level and compare signals, which greatly reduces the time required and really makes repair a lot easier.
In case it's really a bad AM8048, it would be bad news, especially if it's a systematic fault, like the bad Mostek ROMs and RAMs on the CPU board. I just searched the datasheet of the Intel 8048 and found that the 8748 and 8749 are the UV-reasable EPROM variants of the 8048 and 8049 with 1/2kB ROM and 64/128 byte RAM respectively. I also found both a 8748 and a 8749 in my trash box.
So if you are sure that the AM8048 certainly is dead, I might unsolder the AM8048 in one of my 'used for parts' 3456A and try to read it's contents and to reprogram one of the 8748/8749s and then look whether I was successful and then send you one of these chips for a little money.
-
According to the service manual, the in-guard MCU HP partnumber 1820-2539 which is an AMD AM8048 u-processor, is dead.
I'm not entirely sure it's the MCU, but it's worth swapping out if I can find a cheap replacement. I've google'd it. Can't find any suppliers. Any ideas where to look? I'm also looking for a parts unit on Ebay. No good ones yet.
Is it pin/voltage compatible with the Intel 8048?
Also, am I a fool for trying to diagnose this particular dmm?
AMD in those days had extensive cross - licensing with Intel and even used Intel masks for 8088/8086 anyway. So almost certain pin compatible. Eventually Intel severed ties with AMD when second sourcing became less important and Intel grew into a monster. Unfortunately that is a mask rom* part. I would check crystals/power first before pulling that part. Good luck.
* I see MadTux posted while I was typing
-
Definitely start with the power supply, there are many electro's on there that will give all manner of issues as they dry out.
Certainly need to eliminate that first. I have repaired quite a few at work and all had the power supply problem only.
And as Harvs said these meters are absolutely worth the effort as they still hold there own and are very reliable for there age.
(except for there power supply!!).
See the thread below (and others) for further useful info.
https://www.eevblog.com/forum/testgear/hp3456a-service-notes-for-owners-info (https://www.eevblog.com/forum/testgear/hp3456a-service-notes-for-owners-info)!/msg144820/#msg144820
Some notes from HP and a few photo's including one with the AM8048 !!!, with the older style red IC socket that can cause issues. They replaced it in later serial's (see the service note).
Good luck with your 3456.
-
Thanks everyone! Great advice. I've checked a lot of the basics. But I have not done an analysis of the PS yet. The system appears to want to work, it's got a common boot error (-300000) problem with the inboard logic. The CPU theory is based on good 5V rail with correct signals on key pins like the osc and reset. The board simply doesn't respond to basic inputs like from/back input selection.
-
With the PS, the only questionable rail is the +33. It's reading 42V. The caps look OK. They aren't glued down and look really wobbly though.
-
That error is common with dodgy supply rails, replace all the electrolytics and even if they look okay.
-
With the PS, the only questionable rail is the +33. It's reading 42V. The caps look OK. They aren't glued down and look really wobbly though.
42V on that rail is normal, it's totally unregulated and 42V is within spec in the manual (mine is the same.)
-
With the PS, the only questionable rail is the +33. It's reading 42V. The caps look OK. They aren't glued down and look really wobbly though.
42V on that rail is normal, it's totally unregulated and 42V is within spec in the manual (mine is the same.)
Yeah, I just pulled that out of the manual. I'll probably to what HZ suggested and replace all the caps anyway. My DSO was RMA'd back to Tequipment, so I can't take snapshots of sag or ripple on the rails.