EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: jlmoon on April 15, 2014, 04:33:42 pm
-
Hello,
Has anyone had experience with this strange intermittent failure? I have been watching this symptom since I bought the unit a couple of years ago, and finally getting around to addressing it. When my 3325a is first powered from cold start for say 5 minutes or so all is good, then at times the display will go random (scrambled segments / digits flashing) , output drops and display goes into processor confused mode. The random characters / segments will light (brightly) as if the processor clock - multiplexing functions are failing and the signal routing relays will click a few times and then it locks up. I can recycle power and the machine will come back to life at 1000Hz 1.0mvP-P output and it usually will continue to function correctly afterwards, a few times after re powering it might lock (randomly) a second time. I know it has to do with temperature as the device is going through warm up temp cycle, some have suggested looking at power supply caps / voltages. I have been around various areas on the logic boards with freeze spray and heat gun looking for thermally related issues with no success as of today, almost as if when I run the machine with the covers off, it knows to work just fine. So far, from what I can see the power supplies do not seem to change levels or pickup noise as this problem occurs. I have been watching the threads looking for similar problems with no luck so far, I have seen on other forums a suggestion of Rom failure, is there any merit to this? Any suggestions would be great!
Thanks in advance,
JLM
-
The processor continuously monitors the signal and sets various internal conditions. Sounds like you have some kind of problem on the processor board. Could be the ROMs, might be the processor, maybe the RAMs. Might be something intermittent in the bus logic.
Probably a good candidate for signature analysis unless you can find a spare parts unit and swap boards (note that there were changes made through the series and the boards may not be directly compatible without work - connectors changed, some minor functions got moved around, etc).
-
Hello Paul,
Yes, I have suspected the processor board as well. When I first got the unit, I gave it some attention and was looking around the clock circuit that runs the cpu noticing the excessive slew rate or rise time on the leading edge of the clock signal.. so I kludged a 74s04 hex driver in the circuit and was able to reduce the intermittent problem somewhat. Obviously this was not the complete fix but did help it some. I forgot to post that others have said the RCA style plugs are a big problem with these units as well, and could be an issue for mine as well, just don't know yet. With that said, I am am about to tear into it and break it down section by section until I do find the problem.. looking on flee-bay for a parts donor as we speak as well! I sure hope it is not those ROMs..(mask programmed roms) they are some really strangely programmed parts with the Chip Select / Enable logic and the Addressing scheme..
Thanks
JLM
-
Try resoldering the crystal oscillator for the MCU and all the other crystals, it sounds like either a dry jointed crystal or a failing one.
-
The ROM images are available online and there's a 2 ROM solution as well. Both 4 and 2 ROM versions are out there. They're identical, taking into account the bank switching (ROM 1 & 3 make up ROM 1 of the 2 ROM set, 2 and 4 make up ROM2)
The CS lines are kludged like that because the nanoprocessor has an 11 bit address space, an 8 bit data bus and the code has to do bank switching.
HP used that processor in a number of instruments as well as a GPIB processor and some peripherals. It's pretty impressive what they crammed into 16K.
-
Your may want to look at this forum thread which discusses history of reliability upgrade to the 3325a/b.
http://www.febo.com/pipermail/time-nuts/2006-July/021151.html (http://www.febo.com/pipermail/time-nuts/2006-July/021151.html)
Regards
Rastro
Edit 12/2018: I just realized this link is no longer available so I've added a pdf of the original content complements of "the way back maching".
-
Paul,
Thank you for that, Yes, I think it is 'K04BB' website that has the images and address mapping info.. if I recall. Considering when this equipment was design and manufactured it is quite an impressive piece of hardware.. indeed! :-+
-
Rastro,
That's some good information.. appreciate that!
Mine has the older style relays indeed and come to think of it, I think the p/s pass transistors are still in sockets as well.. great info!
JLM
-
Right, K04BB has the 2 ROM images along with some info on compatible EPROMs. If you have to replace them, at least you won't have to try to recover the code from the existing ROMs. I think the 4 ROM images are on bluefeathertech.com
The early 3325a used a flat ribbon connector, which is a pain because the traces will delaminate. Before going to gold connectors, they went to a 0.1 pin socket and header which at least can be removed and replaced.
The command/data bus does go to the other boards; you might have a marginal part somewhere pulling down the bus. I had a bad octal latch on the function board of the one I'm working on now, although that didn't affect the display.
-
The thread is old, but the 3325A is a lot older!
In addition to the factors already mentioned earlier in this thread, random crashes with garbled characters on the screen can be caused by something as simple as the ground contact made by the screws holding the PCBs to the "ground plane" underneath. The circuitry seems to be very sensitive to ground issues.
To see if this is happening to your unit, re-tighten the mounting screws (firm, but not overtightened) on both the A6 board (the large one with the processor on it) and - possibly even more important - the power supply board at the back of the unit. If a screw is already firm, loosen it and then re-tighten to break through any oxide layers.
This "trick" has been a miracle cure on two occasions for me, hope it works for you.
-
The thread is old, but the 3325A is a lot older!
In addition to the factors already mentioned earlier in this thread, random crashes with garbled characters on the screen can be caused by something as simple as the ground contact made by the screws holding the PCBs to the "ground plane" underneath. The circuitry seems to be very sensitive to ground issues.
To see if this is happening to your unit, re-tighten the mounting screws (firm, but not overtightened) on both the A6 board (the large one with the processor on it) and - possibly even more important - the power supply board at the back of the unit. If a screw is already firm, loosen it and then re-tighten to break through any oxide layers.
This "trick" has been a miracle cure on two occasions for me, hope it works for you.
Thanks for sharing. This is an old post; I forgot I posted here. It's always good to have new insights on old(classic) equipment... :-+
I wonder if it would be advisable to use internal-tooth-washers on screws with PCB ground planes??? I can't recall if the original design used these - I sold mine to make space a few years ago.
rastro