开云体育

I thought my AA5001 was fully working, but have discovered that the GPIB board is not responding to commands. I have over a dozen working GPIB devices, including four other TM modules. Since the AA5001 does not work in any (otherwise working) slots, I have no reason to suspect my controller or chassis.

When my controller writes a command string to the device (or tries to read) it times-out, so the GPIB board appears to not be responding. I have tried several different device addresses, to rule out a bad bit on the address switch / decoder on the GPIB board.

Since the freely available service manual is missing all the service / maintenance sections, I bought the Artek manual, which I am very happy with. However, I am wondering if it might be missing the larger pull-out pages that are mentioned on page 6-11 under the (short) discussion on Signature Analysis. The pdf does include Figure 9-8 which is called “Signature Analysis Table”. However, it is not much help. It shows some parts-outlines, with some codes alongside selected pins. The codes are not hex, since they include characters like ‘P’, ‘H’, and ‘U’. It provides no detail as to how to capture or interpret them.

When signature analysis is supported, I normally will find more detail as to how to capture the data. More importantly, I would expect to find a table that provides some diagnostic commentary on where the problem may be when you don’t see the expected codes.

Does anyone have the printed manual or otherwise have any details to get me started with Signature analysis on this unit?

Thanks,
Brian.

OK, signature analysis:

Put the processor in a no-op loop that effectively reads all locations in memory and wraps.? (the processor sees no-ops on the bus regardless of what's out there.

The signature analyzer generally has a start, stop, clock, and ground leads.

Instructions on how to connect them are in the manual, as are the settings on how to capture data (clock, active levels, etc).

Once connected, you can read the pattern.? It's a hex pattern, with the H, P, U and so on replacing the hex characters.? You can, once connected, take the data lead and connect it to ground and to VCC and get an idea of what the pattern is for a node that is stuck at 1 or zero.

All you're looking for is "is the pattern as stated".? You can't really go from a pattern (132H rather than 132P) and say oh, this chip I'm not connected to is bad...

What it does tell you is that connection (node) is not performing as desired.

You then look at the inputs of the gate/element with the bad output, and look for bad inputs.? Tracing that bad input back to a bad output from something, says that the something is bad.? If all the inputs to the chip you're testing are good, and you have a bad output pattern, suspect the chip.

If you haven't used it before, you'd think it did more, but you're really looking at "functional/non-functional" for each circuit node they give you.

Hopefully, this fills in the blanks.? If you already knew this, my apologies.

Harvey

Does the AA5001 have a DIP switch for the address and terminator options?
These very often fail open and is the first thing I would check on any Tek
gear with gpib issues. Even though you think you've ruled it out by trying
multiple addresses, ohm out the DIP switch.

Dave Casey

On Thu, Nov 25, 2021, 1:54 PM Brian Nordlund <brian.nordlund@...>
wrote:

I thought my AA5001 was fully working, but have discovered that the GPIB
board is not responding to commands. I have over a dozen working GPIB
devices, including four other TM modules. Since the AA5001 does not work
in any (otherwise working) slots, I have no reason to suspect my controller
or chassis.

When my controller writes a command string to the device (or tries to
read) it times-out, so the GPIB board appears to not be responding. I have
tried several different device addresses, to rule out a bad bit on the
address switch / decoder on the GPIB board.

Since the freely available service manual is missing all the service /
maintenance sections, I bought the Artek manual, which I am very happy
with. However, I am wondering if it might be missing the larger pull-out
pages that are mentioned on page 6-11 under the (short) discussion on
Signature Analysis. The pdf does include Figure 9-8 which is called
“Signature Analysis Table”. However, it is not much help. It shows some
parts-outlines, with some codes alongside selected pins. The codes are not
hex, since they include characters like ‘P’, ‘H’, and ‘U’. It provides no
detail as to how to capture or interpret them.

When signature analysis is supported, I normally will find more detail as
to how to capture the data. More importantly, I would expect to find a
table that provides some diagnostic commentary on where the problem may be
when you don’t see the expected codes.

Does anyone have the printed manual or otherwise have any details to get
me started with Signature analysis on this unit?

Thanks,
Brian.

Dave-
BINGO - You were right on target. I have always had pretty good luck by just exercising similar DIP switches, but these have to be the worst ones I have come across. All six switches were stuck open, causing it be set to a GPIB address of "0", which of course is nonsense / non-functional.

It took about 20 minutes of spraying and flip-flopping, but I finally got the switches to work, and the module now works great.

Thank you for reminding me to investigate the simple things first, as I was preparing to dive into an overly complicated investigation.
Brian.

On 2021-11-27 1:15 a.m., Brian Nordlund wrote:

Dave-
BINGO - You were right on target. I have always had pretty good luck by just exercising similar DIP switches, but these have to be the worst ones I have come across. All six switches were stuck open, causing it be set to a GPIB address of "0", which of course is nonsense / non-functional.
It took about 20 minutes of spraying and flip-flopping, but I finally got the switches to work, and the module now works great.

Just curious what you sprayed into them. I have an HP 1340A XY with a stuck DIP switch - a signal attenuator switch - seems like bad design to have raw signals passing through these tiny cheap switches?

--Toby

Thank you for reminding me to investigate the simple things first, as I was preparing to dive into an overly complicated investigation.
Brian.

Toby-
I used DeoxIT-gold. I didn’t know whether the contacts were gold, but I have had good luck with that cleaner being gentle on plastics, so that was my choice and it seemed to work well after a lot of exercising of the switches.

Thanks
Brian.

On 2021-11-27 11:00 a.m., Brian Nordlund wrote:

Toby-
I used DeoxIT-gold. I didn’t know whether the contacts were gold, but I have had good luck with that cleaner being gentle on plastics, so that was my choice and it seemed to work well after a lot of exercising of the switches.

Thanks! I have some Deoxit D series but it's the brush type not spray.

--Toby

Thanks
Brian.