Author Topic: HP5004a signature analyzer repair  (Read 13707 times)

0 Members and 1 Guest are viewing this topic.

Offline RobIITopic starter

  • Regular Contributor
  • *
  • Posts: 86
  • Country: us
HP5004a signature analyzer repair
« on: January 26, 2016, 03:29:58 am »
I picked up a HP5004a sig analyzer that powers on but when I connect the pod probes to the test connections on the front panel and put the unit in test, it flickers and only displays 0000.  According to the user manual, it should display some values depending if the Start/Stop are - or +.  I have not taken the unit apart yet but wondering if someone could point me in a direction.

Many thanks in advance!!
 

Offline PaulAm

  • Frequent Contributor
  • **
  • Posts: 938
  • Country: us
Re: HP5004a signature analyzer repair
« Reply #1 on: January 26, 2016, 01:55:57 pm »
You need to have the start/stop/clock leads from the pod attached to the test points on the front as well as having the probe inserted into the testpoint.

The display depends on the settings of the buttons.  Get the manual here:

arcarc.xmission.com/Test%20Equipment/HP/HP%205004A.pdf
 
If you're doing all that correctly, check the power supply.  If I recall correctly there's a fairly large cap that can go bad.
« Last Edit: January 26, 2016, 01:57:35 pm by PaulAm »
 

Offline RobIITopic starter

  • Regular Contributor
  • *
  • Posts: 86
  • Country: us
Re: HP5004a signature analyzer repair
« Reply #2 on: January 27, 2016, 01:16:09 am »
Thanks Paul. Yup had everything connected.  The probe light would flash as well but still 0000 on the display. I am going through the manual's troubleshooting section. I will crack the case open and have a look.

Sent from my GT-P5210 using Tapatalk

 

Offline PaulAm

  • Frequent Contributor
  • **
  • Posts: 938
  • Country: us
Re: HP5004a signature analyzer repair
« Reply #3 on: January 28, 2016, 10:43:52 pm »
I was looking through the manual and it looks like you can use another signature analyzer to check the signature analyzer.

It looks like a bunch of logic chips so there's probably a bad one somewhere.  You should be able to track it down with a scope and/or logic probe while it's in test mode.

 

Offline pelule

  • Frequent Contributor
  • **
  • Posts: 513
  • Country: de
  • What is business? It’s other people’s money
Re: HP5004a signature analyzer repair
« Reply #4 on: January 28, 2016, 11:53:45 pm »
I have the HP 5005B Signature Analyzer / DMM, it has a self test function for basic signature function tests.
I looked into the O&S-Manual of the 5004A.
Ch 3-14 (page 19) describes same (including the procedure in Figure 3-2 below)
Did you walked through this procedure?
What's you results - this will give the first hints for troubleshooting.
BR
PeLuLe

PS:
You may have a different version (the Keysight one is just useable for manual changes).
For your convinience I have uploaded my HQ PDF of the Q&S manual under below link.
I will delete the file in 7 days, as I need the storage for a joined project.
https://www.dropbox.com/s/xspsu9umbz889b9/HP%205004A%20Signature%20Analyser%20-%20Operating%20%26%20Service%20Manual%201977-03%20%2805004-90001%20HQ%29.pdf?dl=0
You will learn something new every single day
 

Offline RobIITopic starter

  • Regular Contributor
  • *
  • Posts: 86
  • Country: us
Re: HP5004a signature analyzer repair
« Reply #5 on: January 29, 2016, 01:35:44 am »
Thanks again for the reply.  A friend of mine is loaning me his Sig Analyzer this weekend...oddly enough a HP5004a.  I will report my findings.

I was looking through the manual and it looks like you can use another signature analyzer to check the signature analyzer.

It looks like a bunch of logic chips so there's probably a bad one somewhere.  You should be able to track it down with a scope and/or logic probe while it's in test mode.
 

Offline RobIITopic starter

  • Regular Contributor
  • *
  • Posts: 86
  • Country: us
Re: HP5004a signature analyzer repair
« Reply #6 on: January 29, 2016, 01:41:32 am »
Thank you my good man!!  I did follow through the flow chart but have not had a moment to check any components.

Thank you so much for the link.  I have downloaded it so you can delete it.  That one is different than the one I had.  Much appreciated!!

I have the HP 5005B Signature Analyzer / DMM, it has a self test function for basic signature function tests.
I looked into the O&S-Manual of the 5004A.
Ch 3-14 (page 19) describes same (including the procedure in Figure 3-2 below)
Did you walked through this procedure?
What's you results - this will give the first hints for troubleshooting.
BR
PeLuLe

PS:
You may have a different version (the Keysight one is just useable for manual changes).
For your convinience I have uploaded my HQ PDF of the Q&S manual under below link.
I will delete the file in 7 days, as I need the storage for a joined project.
https://www.dropbox.com/s/xspsu9umbz889b9/HP%205004A%20Signature%20Analyser%20-%20Operating%20%26%20Service%20Manual%201977-03%20%2805004-90001%20HQ%29.pdf?dl=0
 

Offline pelule

  • Frequent Contributor
  • **
  • Posts: 513
  • Country: de
  • What is business? It’s other people’s money
Re: HP5004a signature analyzer repair
« Reply #7 on: January 29, 2016, 10:23:09 pm »
File deleted.
Can be download from KO4BB.com
BR
PeLuLe
You will learn something new every single day
 

Offline RobIITopic starter

  • Regular Contributor
  • *
  • Posts: 86
  • Country: us
Re: HP5004a signature analyzer repair
« Reply #8 on: January 30, 2016, 04:49:06 pm »
Got the loaner sig analyzer and connected to my non working one as described in the user manual.

Test points 4 shows an incorrect signature (display's A446 when it should be HH46) and test points 7, 8 and 9 are incorrect as well in Normal and Service mode.  All the others show correct.  The 5v test point, I have 4.82v.

Checking the flow chart it says check Self Test circuit A1U25, A1U26 and A1U27 if signatures are not correct for TP 1 - 5.  Am I to assume it means check the Address 1 lines of U25, 26 and 27 with a logic probe to see if they are stuck or something. 

Please understand I am not an EE so I apologize for the rudimentary questions.

Thanks for the assistance!!
 

Offline pelule

  • Frequent Contributor
  • **
  • Posts: 513
  • Country: de
  • What is business? It’s other people’s money
Re: HP5004a signature analyzer repair
« Reply #9 on: January 30, 2016, 07:09:23 pm »
Quote
Please understand I am not an EE so I apologize for the rudimentary questions.
If you have no electronic experiance, you are in trouble. In that case I would try to find some help in your area.  :--

In Germany self-repair support shops are becoming more and more popular. There you will find tools, knowledge and help (I am supporter of such a group here in my area).

If you have just never worked with such "old style" electronics, I have tried to guide you through the start-up below...  :-+
Hope it helps.

Quote
Test points 4 shows an incorrect signature (display's A446 when it should be HH46) and test points 7, 8 and 9 are incorrect as well in Normal and Service mode.  All the others show correct.  The 5v test point, I have 4.82v.
The 5V (4.82V) is ok in my view 3.6% deviation is will in the by TTL tolerance.

But I am not sure, which troubleshooting diagram you have followed. Would help, if you note the page, add some pictures, schematics or similar. Please keep in mind, this is no face-to-face communication, the other side is blind  :-//
- I have looked to the path of troubleshooting-diagram on p61 (SM I have supplied for upload) - out of my understanding the start point.
- If you have wrong signature at TP1 - TP5 -> diagram recommends to look to the selftest circuit, located on the main board A1, build by the ICs U25/26/27.
This seems to fit to your description, thus I have attached the part of the board location (p73) and the schematic (p74).
Quote
Am I to assume it means check...
U25, U26 and U27 are 8 bit counter (I assume binary types).
1. I would recommend to download first the data sheets of that parts, will help to understand there function.
Look to the part list table, which parts are U25/26/27. Normally that are HP part numbers. There are cross tables on the web (just google for "HP part numbers"...)
2. Next you may want to begin with my below listed check-flow / U26(2,3) means: measure at U26 pin2 and 3 ...
Check U26(2,3) - are the levels at that logik input ok?
--> if not correct == check the board for issues on wiring or bad soldering (repair / than continue)
--> if correct --> check U26 (14) - is there clock signal?
--> no clock == check U28, may have a problem (repair / than continue)
--> clock exist --> has U26(12) clock (output)?
--> no clock == U26 has a problem  (repair / than continue)
--> clock exist --> continue same testing with U27 and so forth...

Hope these information helps and is not obvious. Good luck...
BR
PeLuLe
You will learn something new every single day
 

Offline RobIITopic starter

  • Regular Contributor
  • *
  • Posts: 86
  • Country: us
Re: HP5004a signature analyzer repair
« Reply #10 on: January 30, 2016, 07:26:19 pm »
Thank you kind sir!  While I don't have official education in electronics, I am familiar with basics and able to understand on a somewhat beginners/intermediate level.

My apologies, I should have clarified which diagram I was looking at.  I was referring to the flowchart on pg 53 and 54 of the PDF you provided.

I will review your suggestions and reply with my findings.

Thank you for your assistance and patience!!

Rob
 

Offline dom0

  • Super Contributor
  • ***
  • Posts: 1483
  • Country: 00
Re: HP5004a signature analyzer repair
« Reply #11 on: January 31, 2016, 10:45:59 am »
Quote
In Germany self-repair support shops are becoming more and more popular. There you will find tools, knowledge and help (I am supporter of such a group here in my area).

I didn't know about this until this post, at first I thought you meant hackerspaces, but a hackerspace isn't really a "self-repair support shop".

Turns out these things have a trade name, they're called "repair cafe" and convene regularly.
,
 


Offline RobIITopic starter

  • Regular Contributor
  • *
  • Posts: 86
  • Country: us
Re: HP5004a signature analyzer repair
« Reply #13 on: February 11, 2016, 01:43:36 am »
Finally getting back to this.  Started probing U25, 26 and 27.  The only abnormal thing I am seeing is pin 11 on U25 is stuck low.  The same pins on U26 and 27 have activity.  I am going to socket and replace it and see what happens.
 

Offline pelule

  • Frequent Contributor
  • **
  • Posts: 513
  • Country: de
  • What is business? It’s other people’s money
Re: HP5004a signature analyzer repair
« Reply #14 on: February 11, 2016, 04:40:48 pm »
Code: [Select]
The only abnormal thing I am seeing is pin 11 on U25 is stuck low.I recommend some more measurements bevor soldering.
U25 is a counter: pin 2/3 are enable inputs (AND), pin 5 seems to be the reset (active high), pin 1 is the clock input, pin 11/8/9 are the counter outputs.
If pin 11 (1st counter stage output) stays low, this may also caused by "no counting" or "reset":
- no clock at pin 5
- a low state at pin 2/3, a short low pulses may have the same effect if tming fits
- a high state at pin 5 (f.e. self test switch in Position ON).

Do you have a logic analyzer or a 4ch scope available?
In that case I recommend to measure at same time pin 1/2/5/11 to check if the condition prevents counting/change of the ouput

BR
PeLuLe

Then first take some other measuremens  at same time singnals at U25(
You will learn something new every single day
 

Offline RobIITopic starter

  • Regular Contributor
  • *
  • Posts: 86
  • Country: us
Re: HP5004a signature analyzer repair
« Reply #15 on: February 12, 2016, 12:38:51 am »
I do not have a logic analyzer but I do have a four channel scope.  I will double check what you suggest.

Thank you!!! :-+

I recommend some more measurements bevor soldering.
U25 is a counter: pin 2/3 are enable inputs (AND), pin 5 seems to be the reset (active high), pin 1 is the clock input, pin 11/8/9 are the counter outputs.
If pin 11 (1st counter stage output) stays low, this may also caused by "no counting" or "reset":
- no clock at pin 5
- a low state at pin 2/3, a short low pulses may have the same effect if tming fits
- a high state at pin 5 (f.e. self test switch in Position ON).

Do you have a logic analyzer or a 4ch scope available?
In that case I recommend to measure at same time pin 1/2/5/11 to check if the condition prevents counting/change of the ouput

BR
PeLuLe

Then first take some other measuremens  at same time singnals at U25(
 

Offline RobIITopic starter

  • Regular Contributor
  • *
  • Posts: 86
  • Country: us
Re: HP5004a signature analyzer repair
« Reply #16 on: February 12, 2016, 01:32:13 am »
A quick update.  Pins 2 and 3 are tied together on U25, 26 and 27  and all were showing low.  Pins 1 and 14 on U26 and 27 were showing clear clock signal BUT pins 1 and 14 of U25 were just flip flopping high and low.  So I am going to guess there is a problem with the clock signal to U25?
 

Offline pelule

  • Frequent Contributor
  • **
  • Posts: 513
  • Country: de
  • What is business? It’s other people’s money
Re: HP5004a signature analyzer repair
« Reply #17 on: February 12, 2016, 05:43:04 pm »
A 4ch-scope is much better than logic analyzer, as you are able to see quality of the signals also.

I am not sure what does mean "pins 1 and 14 of U25 were just flip flopping high and low".
But to check of U28 (clock IC) should be easy. Output Pin 3 must be a stable frequency (~50% duty cycle).
I expect it is OK and you need more measurements to dig to the error.

For all measurements you need to look to U10/Pin 6 (reset) at te same time to be sure it stays (?) at LOW.
It is the signal NOT(U28/QD & U25/QD) building a feedback loop to generate the test signal.
So it will not stay at LOW forever...

U25, U26 and U27 are 74LS93 (4-bit asynch binary counter, HP parts 1820-1478).
Pin 2 and 3 are &-wired reset inputs, both need to be HIGH, to get an reset - here they are connected and LOW.
Pin 14 is clock input CKA of first counter stage A, Pin 1 (CKB) is clock input of the remaining counter stages (B-D).
Typically (as also here) the first stage output QA (pin 12) is connect to CKB to buld a complete 4-bit counter.
Conclusion: U25/Pin 1 must have exact the half of the clock speed of U25/Pin 14 (binary devider).
Is that the case?
If NOT: U25 may have a defect and should be replaced.
If YES:
U25/Pin 14 must be same signal as U27/Pin 11 (U27/QD).
U27/Pin 11 (U27/QD) must exact the half speed of clock of U27/Pin 8 (U27/QC).
In that way should be able to check/follow the signals back to U28 (clock IC)
BR
PeLuLe
You will learn something new every single day
 

Offline RobIITopic starter

  • Regular Contributor
  • *
  • Posts: 86
  • Country: us
Re: HP5004a signature analyzer repair
« Reply #18 on: February 12, 2016, 05:56:32 pm »
Thank you pelule.  Rather than try to explain what it is doing, I will make a small video tonight when I get home to hopefully better explain what I am seeing.  I will also go over what you have suggested.

Thank you so much for your assistance!!

Rob

A 4ch-scope is much better than logic analyzer, as you are able to see quality of the signals also.

I am not sure what does mean "pins 1 and 14 of U25 were just flip flopping high and low".
But to check of U28 (clock IC) should be easy. Output Pin 3 must be a stable frequency (~50% duty cycle).
I expect it is OK and you need more measurements to dig to the error.

For all measurements you need to look to U10/Pin 6 (reset) at te same time to be sure it stays (?) at LOW.
It is the signal NOT(U28/QD & U25/QD) building a feedback loop to generate the test signal.
So it will not stay at LOW forever...

U25, U26 and U27 are 74LS93 (4-bit asynch binary counter, HP parts 1820-1478).
Pin 2 and 3 are &-wired reset inputs, both need to be HIGH, to get an reset - here they are connected and LOW.
Pin 14 is clock input CKA of first counter stage A, Pin 1 (CKB) is clock input of the remaining counter stages (B-D).
Typically (as also here) the first stage output QA (pin 12) is connect to CKB to buld a complete 4-bit counter.
Conclusion: U25/Pin 1 must have exact the half of the clock speed of U25/Pin 14 (binary devider).
Is that the case?
If NOT: U25 may have a defect and should be replaced.
If YES:
U25/Pin 14 must be same signal as U27/Pin 11 (U27/QD).
U27/Pin 11 (U27/QD) must exact the half speed of clock of U27/Pin 8 (U27/QC).
In that way should be able to check/follow the signals back to U28 (clock IC)
BR
PeLuLe
 

Offline RobIITopic starter

  • Regular Contributor
  • *
  • Posts: 86
  • Country: us
Re: HP5004a signature analyzer repair
« Reply #19 on: February 13, 2016, 06:46:40 pm »
Hopefully I will have answered all of your questions below.  I want to make sure that I have clarified that I have the analyzer probes connected to the front of the unit and it is in test mode.

Here is a short video of scope readings on pin 3 of 28 and U27 pin 8 and pin 11.

https://youtu.be/2rQlKIzqcpE

But to check of U28 (clock IC) should be easy. Output Pin 3 must be a stable frequency (~50% duty cycle).
I expect it is OK and you need more measurements to dig to the error.


Output of U28 pin 3 does not look like it is 50% duty cycle to me but not sure if that looks to be with in tolerance.


For all measurements you need to look to U10/Pin 6 (reset) at te same time to be sure it stays (?) at LOW.
It is the signal NOT(U28/QD & U25/QD) building a feedback loop to generate the test signal.
So it will not stay at LOW forever...

U10 pin 6 is staying low all the time

Conclusion: U25/Pin 1 must have exact the half of the clock speed of U25/Pin 14 (binary devider).
Is that the case?

I get a similar output as in the video, it jumps high and low.
 

Offline pelule

  • Frequent Contributor
  • **
  • Posts: 513
  • Country: de
  • What is business? It’s other people’s money
Re: HP5004a signature analyzer repair
« Reply #20 on: February 13, 2016, 11:11:34 pm »
The video does not answer the questions unfortunately.
It just show one signal at the time, so I am not able to compare the signal timing and sychronization. What can be seen: you have triggered the first signal to a stable picture, for the other not. They are slow (as expected, see below) and you keep the horizontal setting time/div at the scope unchanged.
You may want to connect at least two signal (CH1 & CH2) at the same time, set time/div to the slowest signal. Than you are able to detect, if the 2nd signal is excatly the half speed of the 1st - as it must at a binary counter stage.
Each stage is a divider by 2, so the 4 bit binary counter is a diver by 16.
Signals should look like:

CH1: __HHHH____HHHH____HHHH____HHH
CH2: __HHHHHHHH________HHHHHHHH___

Clock of U28 (clock IC) is ~600 Hz (0.6 KHz), so:
U26(12) ~300 Hz - U26(11) ~38Hz - U27(12) ~19Hz - U27(11) ~2.4Hz - U25(12) ~1.2Hz - U25(11) ~0.15Hz
Means the loop runs ~7 seconds. Than it should reset the counters and restart at the next low output of U25(11).

You will learn something new every single day
 

Offline RobIITopic starter

  • Regular Contributor
  • *
  • Posts: 86
  • Country: us
Re: HP5004a signature analyzer repair
« Reply #21 on: February 14, 2016, 02:44:31 pm »
My apologies for not answering your questions.  Thank you so much for the clarification.

I did as you suggested and verified the clock on the given pins.  They all seem to be working though I was having great difficulty triggering on the slow clock pins. 

Where should I turn to now?

The video does not answer the questions unfortunately.
It just show one signal at the time, so I am not able to compare the signal timing and sychronization. What can be seen: you have triggered the first signal to a stable picture, for the other not. They are slow (as expected, see below) and you keep the horizontal setting time/div at the scope unchanged.
You may want to connect at least two signal (CH1 & CH2) at the same time, set time/div to the slowest signal. Than you are able to detect, if the 2nd signal is excatly the half speed of the 1st - as it must at a binary counter stage.
Each stage is a divider by 2, so the 4 bit binary counter is a diver by 16.
Signals should look like:

CH1: __HHHH____HHHH____HHHH____HHH
CH2: __HHHHHHHH________HHHHHHHH___

Clock of U28 (clock IC) is ~600 Hz (0.6 KHz), so:
U26(12) ~300 Hz - U26(11) ~38Hz - U27(12) ~19Hz - U27(11) ~2.4Hz - U25(12) ~1.2Hz - U25(11) ~0.15Hz
Means the loop runs ~7 seconds. Than it should reset the counters and restart at the next low output of U25(11).
 

Offline pelule

  • Frequent Contributor
  • **
  • Posts: 513
  • Country: de
  • What is business? It’s other people’s money
Re: HP5004a signature analyzer repair
« Reply #22 on: February 14, 2016, 09:01:55 pm »
Lets first check, what kind of test equipment you have available beside the 4ch scope.
Do you have
- a logic analyzer (>= 8 channels, >= 10 MHz)?
- a frequency counter (>= 10 MHz)?
How is your knowledge regarding signature analyzers principle? If you not understand the principle of signature analysis, you will stumble around like a blind in the fog.
I have attached an very good article from HP Journal, when HP introduced the 5004A. This may help to get a better understanding.
I have attached the link to a very good article from HP Journal, when HP introduced the 5004A. This may help to get a better understanding (file too big to attache):
http://www.hpl.hp.com/hpjournal/pdfs/IssuePDFs/1977-05.pdf%2Fwww.hpl.hp.com%2Fhpjournal%2Fpdfs%2FIssuePDFs%2F1977-05.pdf&usg=AFQjCNGFSQWK8RkSBEG2EfBv7AtRSyW5aA&bvm=bv.114195076,d.bGs
You will learn something new every single day
 

Offline PaulAm

  • Frequent Contributor
  • **
  • Posts: 938
  • Country: us
Re: HP5004a signature analyzer repair
« Reply #23 on: February 14, 2016, 09:59:33 pm »
 

Offline RobIITopic starter

  • Regular Contributor
  • *
  • Posts: 86
  • Country: us
Re: HP5004a signature analyzer repair
« Reply #24 on: February 14, 2016, 11:06:40 pm »
Sorry, the only other test equipment I have is a logic probe.  I may be able to get my hands on a frequency counter.  I do understand the principle behind signature analyzers.  I am currently working on troubleshooting an Asteroids Arcade PCB, built a NOP adapter for the 6502A CPU.  Just waiting for some parts to finish that project. 

Not sure if it would help at all but it was test point four (U29 pin 3) on the troubleshooting chart that gave me the first incorrect signature. 1, 2, 3 and 5 were correct.

Thanks for the links and of course, your patience.

Rob

Lets first check, what kind of test equipment you have available beside the 4ch scope.
Do you have
- a logic analyzer (>= 8 channels, >= 10 MHz)?
- a frequency counter (>= 10 MHz)?
How is your knowledge regarding signature analyzers principle? If you not understand the principle of signature analysis, you will stumble around like a blind in the fog.
I have attached an very good article from HP Journal, when HP introduced the 5004A. This may help to get a better understanding.
I have attached the link to a very good article from HP Journal, when HP introduced the 5004A. This may help to get a better understanding (file too big to attache):
http://www.hpl.hp.com/hpjournal/pdfs/IssuePDFs/1977-05.pdf%2Fwww.hpl.hp.com%2Fhpjournal%2Fpdfs%2FIssuePDFs%2F1977-05.pdf&usg=AFQjCNGFSQWK8RkSBEG2EfBv7AtRSyW5aA&bvm=bv.114195076,d.bGs
 

Offline pelule

  • Frequent Contributor
  • **
  • Posts: 513
  • Country: de
  • What is business? It’s other people’s money
Re: HP5004a signature analyzer repair
« Reply #25 on: February 15, 2016, 07:04:18 pm »
This tools are not a MUST, just would make the search much much easier.
Let me look into the schematics and think of wich next steps are best suitablke.
BR
PeLuLe

You will learn something new every single day
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf