HP54520 Logic Trigger Problem

[mwbarth36]

I have a HP54520 oscilloscope that passes all of the Self Calibration until it gets to the last test (3) Logic Trigger, this test it fails with the error code of 00430H.  All analog triggering works fine.  I could probably easily locate the problem if I had even a partial schematic for this scope.  First question...........is there any hope of finding a schematic diagram for the 54520/40?  Somewhere, buried, there must be a schematic for this scope.  2nd question..............does anyone know what the error code can tell me?  Somewhere, there must be a definition of what this error code signifies.  Even a detailed block diagram of the scope or of the main board would be helpful.  Without any of the above information it is nearly impossible to even tell what components are part of the 'Logic Trigger' circuit.  Please, can anyone help.

[Runco990]

I have rebuilt a couple of these scopes.  How are you EXACTLY calibrating it?

You need to load defaults first, and it is picky about cables.

[MarkL]

There is a service guide, which unfortunately does not have schematics, but it does contain block diagrams and a theory of operation overview:

  http://literature.cdn.keysight.com/litweb/pdf/54542-97015.pdf

Looking at a photo of a 54520A motherboard I found online, the logic trigger chip would be the 1FJ9-0002 ASIC (U702).  The 1SE6-0056 ASIC is the timebase, and I would presume the ASIC with the white ceramic and cooling fins (nearest the ADCs) is the analog trigger.

I can't help with the error code, but maybe one of the "for HP only" self-test selections, namely one of the loop tests, may repeatedly test the logic trigger and provide more detail on the failure.  But that's only a guess.

You could also poke around for repair information for similar models and also the 16533A/16534A cards (a scope for their logic analyzers).  There's a lot of re-use of this analog acquisition design and these ASICs in other products that may provide you with troubleshooting clues.

[mwbarth36]

Yes, I have loaded the Defaults first. I am using the Self Calibration procedure from the Service Manual.  The DC and AC calibration signals are present at the rear of the scope, and the on screen calibration procedure tells you when and where to connect the signals.  It is only the 'Logic Trigger' out of all of the other calibration that does not pass.  The block diagram in the service manual shows a block of circuitry for the Logic Trigger.  Inputs are the analog trigger signals from the front end, clocks, and other signals used to do the "Logic Trigger" functions.  But there is no way to know what hardware on the board is used for this function. 

Thanks for your reply!

[mwbarth36]

Thanks so much for responding.  I will do some looking around that ASIC and see what I can find.  The Block labeled Logic Trigger had as inputs, the four analog trigger signals, a clock and two other signals from the time base system.  I know the analog triggers are working just fine, so that leaves the others.  Still, if the problem is in the ASIC, not a lot I can do to correct it short of replacing the ASIC.  I wonder how different the 520A and the 520 board are?  I have a 520.

Thanks again!

[MarkL]

I can't find any documentation which describes the difference between the 54520 and 54520A.  Everything I can find only mentions the 54520A.  Can you post a high-res photo of your motherboard, or at least of the acquisition section?

Were you able to discover anything from any of the loop tests?

In lieu of knowing what the error code means, one thing you could try is to manually configure and test features that the logic trigger is responsible for.  According to the service guide:

  Holdoff
  Delay
  Pattern Duration and Range

I would add to their list:

  State Triggering
  Qualify options

It's not practical (or even knowable) to set up and test every combination that the self-test might be performing, but maybe you'll find something that's obviously not working.  Make sure you include the external trigger input along with both channels in any testing.

EDIT: In reading the Service Guide a little more, I see there's an "ltrig" self-test (presumably for Logic Trigger) under the UTILITY menu.  Have you run this?  The manual says you need to perform a "Key down power up" before using it.  Maybe the self-test will provide some additional detail on the issue.

[mwbarth36]

Thanks for the info.  I just got back home from a Float on the Buffalo River in Arkansas, and will try to get a photo of the board.  I did not recall seeing the ltrig test listed, will look for it.  And I did not try anything yet using the 'Key down, power up, reset.  Will get to that in the next few days.  Thanks again. 

[mwbarth36]

Just ran the 'ltrig' test as you suggested, get what, it Passes every time.  Still when I try to do the Calibrate for (3), ltrig, the final calibrate step, it Fails with a code of either 00400H, or 00430H code at about halfway through the 'countdown' on the screen.  Sure wish I knew what those error codes indicated, it would take forever to do all setups for logic trigger to find which one is not working properly.  And why can it pass the self test?  Not a very comprehensive test?  If anyone has any additional information please HELP! 

[MarkL]

The service manual says it should be asking you to move cables around while doing the time and trigger calibration.  What is the configuration of cables when it fails?

It could be that something is marginal.  Have you checked the power supply voltages?

Also, at least on the photo I have of a 54520A, there appear to be several linear regulators.  You should check their output.  If their output voltages are not labeled, you can calculate what the output should be from the circuit.  I've had several instances of bad resistors making regulators output a wrong voltage, and in one case on a scope card it passed self-test but would not calibrate.

Also, did you find anything useful in the loop tests.  And a photo please.

[mwbarth36]

First of all, I have to admit to a really stupid mistake.  The scope I have is a 54520A, the A is there as plain as day, I just did not look at the front panel for some time and assumed it was a plain 54520.  Sorry, have been kicking myself all day.  However, that makes everything you have said even more important. So, today, I first went back, entered the defaults using 'key down' power up, did a self calibration, just to confirm all passes except 'ltrig', and that is the case everything passes except (3) logic trigger.  'ltrig' fails at around 580 in the countdown with an error code of 00430H.  I again checked that when the 'ltrig' test is run in the test mode it always Passes.  The loop test will run forever without error, one has to end it by depressing a key.

When one enters test (3) logic trigger, the instructions are: Run a cable from the AC calibration signal to Channel 1 and press 'continue', and this I do. 

There are six or seven on board regulators.  Most of them are labeled as to the output voltage except for u51, a 317T.  It's voltage output is 7.89 Volts which seems odd. It is one of three regulators in a row toward the back of the board, the other two are a negative 6 volt regulator for the probe power output which measures 6.34 Volts, and a -2 Volt regulator which measures -2.14 Volts.

I now have the scope with the main board back in the case so I can operate it, as you know it is difficult to operate with the main board removed from it's position in the case.  Also the reason I have not tried to see what determines the output voltage of the 317T, any resistors must be on the bottom of the board.  I do suspect this may be an output controlled by some variable reference because of the unusual voltage output voltage measured..........even if suppose to be 8 Volts, seems strange.  And, also why I have not provided a picture, not a major job to remove the board...again...but also not the 'funnest' thing to do.  Maybe will do that tomorrow to see if I can determine what is controlling this voltage. 

If you have any other thoughts............please share them! 

[Runco990]

Did you try a different coax?  I had those problems and tried several cables before I found a short QUALITY 50 Ohm cable.  I then kept THAT cable for future scope cal's, worked every time.

Try that before you tear your hair out, like I did.  The cable has to have the extra white insulator sleeve in the BNC connector. 

You COULD have a different problem, but I ran into this on TWO of these scopes... the right cable fixed it.

[mwbarth36]

Well, certainly worth a try..................was  your failure confined to the logic trigger calibration?  Just a question, will try a different cable in any case!

[mwbarth36]

Just re-ran the logic trigger calibration (3) two more times with different cables, one about as short as will work, both failed at the same point, with the same error code 00430H.  I then reloaded the defaults and ran complete self calibration so logic trigger cal 3 would start with a default loaded, result is the same, Failed with the same error code.  Damn, wish I knew what that code signifies!  Thanks for the suggestion, it was certainly worth a try.

[MarkL]

The 317T has a reference of +/-4% plus other inaccuracies.  If you're seeing +/-5%, I think it's fine.  Something that's off by +/-10% or more is worth investigating.

The scope cards I've worked on with this chip set have a bunch of strange linear regulator outputs: +/-6V, +/-2V, and +7.3V.  Granted, that may not be much of a useful comparison, but unfortunately I don't have a 54520A to compare voltages to help you further.

Some other ideas:

Because the test is failing on the AC cal, maybe it's using a high speed signal, and for good signal integrity it's turning on the 50 ohm termination on the Channel 1 input.  It might be worth checking with a DMM that Channel 1 measures 50 ohms when the termination is enabled.  I've certainly had my share of bad or flaky attenuator modules of this style.

You could also try swapping the attenuator with Channel 2 as another verification.  I doubt it's the ADC, but that's another component that's very easy to swap that you could also try.

Have you looked at the high frequency pulse response?  There's a section in the service guide on it.  You might not have the equipment to perform the adjustment, but you could at least look at the waveform with the fastest edge you can generate.

If you have another scope, it might be interesting to snoop on the AC Cal output at the time of failure.

Tip: If you start swapping things, label them first so you can put them back exactly where they came from when done.

[mwbarth36]

I have not abandoned this project, had Carpal Tunnel surgery several days ago and have been somewhat restricted as to what I can do.  What I know so far is that the voltage regulators all appear to be within normal range.  The 7.x voltage seems a bit odd,but was glad to see that you found one too.  Mine is a bit higher at 7.8 Volts, but who is to say which one is correct,,both seem strange.  Will add more to this tomorrow, a pain in the neck typing with only my left hand                1

[RaymondMack]

@mwbarth36

If you're interested I have a spare mainboard to a 54520A. I should point out that I've removed the attenuators to fix a 54542C that has two flakey channels. The 54520A arrived with the "blank screen + all lights illuminated" problem that I just found out is due to the battery going flat. So the mainboard should probably work, or at least give you spare parts to fiddle with.

As a side rant, HP made working on these scopes a pain in the ass! You literally have to pull the entire scope apart to replace the attenuators... I have been slowly working up the mental energy and space required to fix my 54542C. Maybe I'll do it this weekend.

[Runco990]

Actually, those scopes are very easy to strip if you follow the logic.  Takes me 15 minutes to COMPLETELY field strip one to clean and overhaul one.  Forget Tektronix.....  Give me an HP scope any day to work on.

[RaymondMack]

Actually, those scopes are very easy to strip if you follow the logic.  Takes me 15 minutes to COMPLETELY field strip one to clean and overhaul one.  Forget Tektronix.....  Give me an HP scope any day to work on.

Fair enough. They aren't hard to pull a part; I've done it three or four times now. But removal of the attenuators is more hassle than it needs to be. I just wish the screws holding the shields were on the top side of the board rather than the bottom or at least some holes were in the underside of the chassis to allow quick access to them. And now that I'm thinking about it, a battery holder would have been nice too.

Otherwise I don't have much to complain about aside from the relatively small capture memory of 32k points, slow update speed for low frequency signals (though manually upping the sampling rate fixes this), lack of X-Y mode and no averaging in repetitive mode. The lack of X-Y mode is a real head scratcher.

[mwbarth36]

Managed to monitor the ac calibration with another scope while trying to run the calibration, here is what happens:

1- AC Cal is a 496 Hz square wave prior to starting the test, when test starts, countdown on screen starts, 2 relay 'clicks', and signal on AC Cal changes to a high frequency with low amplitude, around 100 MHz.  Countdown starts at around 1000.

2- When the countdown reaches around 900, pattern on AC Cal output changes to 2, 50 ns pulses, 100 ns apart, repeated every 500 ns.

3.- When countdown reaches 582, pattern changes back to original 496 Hz square wave; also, 2 relay 'clicks'.  Cannot tell which comes first, waveform change or relay action.  If I had to guess, I would say some test fails and relays signal abortion of the test. 

Since Logic Trigger setups involve more than one input from the 4 possible inputs of this scope ( Input1, Input2, Ext, and Aux), it is difficult to imagine what one would 'calibrate' with the AC Cal output attached to only Channel 1 Input.  Especially since the Analog Trigger tests and delay tests all pass. 

One other point, in the Service Mode, the Logic Trigger Test Passes!  Any thoughts?? 

[mwbarth36]

Thank you so much for the offer!  I would bet that if I had the board and put the attenuators from this board in it would 'calibrate' just fine. Now thinking the issue is Software (Firmware) issue, see my previous Post for why I am thinking along these lines. 

Over the past several years I have worked on probably 30 scopes in the 545xx series, the 54542C is a fine scope, best of the lot.  Hope you are able to get yours working.  Yes, a bit difficult at re-assembling, but it is worth the effort.

[Runco990]

The scope DOES have X-Y.  It's in the math menu.  Not exactly obvious.  Does FFT as well.

[RaymondMack]

The scope DOES have X-Y.  It's in the math menu.  Not exactly obvious.  Does FFT as well.

Looks like you're right! I never tried to use the "vs" option in the math menu. Makes sense now that I think about it. Seeing how this is a four channel scope, one might want to use any two channels for this. For some reason I figured it would be under the horizontal menu. Thanks for pointing this out!

[RaymondMack]

Thank you so much for the offer!  I would bet that if I had the board and put the attenuators from this board in it would 'calibrate' just fine. Now thinking the issue is Software (Firmware) issue, see my previous Post for why I am thinking along these lines. 

Over the past several years I have worked on probably 30 scopes in the 545xx series, the 54542C is a fine scope, best of the lot.  Hope you are able to get yours working.  Yes, a bit difficult at re-assembling, but it is worth the effort.

I'd be willing to let the board go for $20 plus whatever shipping comes out to be. Probably $15 using FedEx Home, but maybe USPS Priority might be cheaper.

Send me a PM if you're interested.

[MarkL]

...
Since Logic Trigger setups involve more than one input from the 4 possible inputs of this scope ( Input1, Input2, Ext, and Aux), it is difficult to imagine what one would 'calibrate' with the AC Cal output attached to only Channel 1 Input.  Especially since the Analog Trigger tests and delay tests all pass. 

One other point, in the Service Mode, the Logic Trigger Test Passes!  Any thoughts??

There are a bunch of different trigger modes that use configurable timers (e.g., delay by time, pattern present for < time, > time, or time range, holdoff, glitch trigger, etc.).  The scope may be trying to calibrate one or more of these timers.  The timer(s) would be common to all the channels, so it really doesn't matter which channel it uses for the reference.

The logic trigger self test may not exercise the timers in the same way, especially if the scope is not asking you to connect the AC cal output while doing the self test.  (Is it?)

Note that if you replace the main board with the one RaymondMack has so kindly offered but without attenuators, you will need to do (or should do) the manual calibration steps starting with the high-frequency pulse response (pg. 4-6).  The manual is pretty adamant about doing all the steps, and in the specified order.

[mwbarth36]

Good thought about the timers, yes, that is about all that could be calibrated.  The connection is made prior to starting the test.  I have swapped the attn's in a number of these scopes, and have always had them calibrate with know good attn's, but I have done the complete manual HF calibration when necessary, and will certainly consider that if needed.  Thanks so much for your thoughts, always appreciated.  The new board is on the way and will post the results when I have tried it.