hp 3561a DSA repair - CRT "double" deviation in x-axis

[unknownsoldier90]

Hello together,

I got a hp 3561a for free from the company I work.

It has been sitting in the basement for I guess 10 years now, unfortunately it is not really useable in this condition.

The main problem is that all the writing and lines are projected to the front of the CRT two times in x-axis, which looks like in the attachment.



I did download the repair manual 1&2, when I put the jumper on A60 board to half bright or full bright the displayed pattern looks good.
Do not look at the different brightness levels, in reality brightness is even, cameras fault



Honestly I don't know how to go on, where could the problem be? Digital or analog driver board?

Somebody had a similar fault in the past?

Thanks in advance,

Constantin

[strawberry]

could be char gen is messed up (Hzz and other weird )
have you checked for bad capacitors first, capacitors in hot area and small ones wear out faster

[unknownsoldier90]

I don’t see any bad capacitors (leaky etc.), but I didn’t measure ESR or capacity.
There is one resistor R203 on a power supply board which has a crack, but resistance is fine at 22 ohms.

I also checked various power lines of the PSU, all the ones with common ground were good. But the ones with isolated ground were all out of spec - removing A10 and A15 board didn’t help…could this be the problem? I only have a really cheap multimeter, maybe measuring fault?
/Edit: All Power lines are fine. I measured the wrong pins of the linear regulators.

I managed to connect the calibration signal of my cheap Hantek oscilloscope, it shows the correct frequencies I guess.

[strawberry]

+5V MAIN is tied to SMPS feedback and should be spot on
R203 burned = C300 ?
all voltages are regulated so expected to be spot on. if measured on same isolated GND
check ripple before regulators, oscilloscope could be handy
PSU 4xLED indicators?

[unknownsoldier90]

Thanks for your reply.

As I wrote in the Edit all outputs of the linear regulators on the side are in spec.

R203 goes to C300, yes. On this board are two testpoints on the top, -12 BIAS and +12 BIAS, both voltages are good in respect to GND.

All LEDs on the PSU are off.

The solder joints of R203 were dry, I resoldered them, but no success.

I am now trying to measure the waveforms of the A60 board like shown in service manual. Since I do not have an extender I have to solder a wire to the pin of the IC, plug the card in again and look at the scope..makes fun 😬

Just measured the +5 V in respect to GND on A60 board. Is this too noisy?

[strawberry]

Wonder why R203 burned. doesnt seem like carbon composite resistor. faulty C300...
noise could be picked up by test leads/wires/loops.
noise is a bit large but could be good enough for logic circuit signal levels

https://www.eevblog.com/forum/testgear/hp-3561a-repair/

half bright / full bright. I guess it tests video RAM

have you tried run self test , errors could help

maybe XYZ output for external monitor/oscilloscope but doubt it would work here

[unknownsoldier90]

I would change R203 immediately, but since it has the specific value it should be okay.

I have put a capacitor between GND and 5 v rail to smooth the ripple, it did go down, but didn't get better on the screen, so I guess it is okay.

Did make the half bright and full bright test, the test pattern appears without any deflection, looks perfect.

I tried I think all self tests like written in the service manual,
no errors (at least I didn't manage to read one, not too easy with this kind of display

Still wondering if A60 (digital video board) or A90 (analog video board) is faulty.
If I even knew which part is wrong I could buy one, eBay seller from Italy has them.

[Tomorokoshi]

I've had plastic encapsulated resistors like R203 fail like that. Consider that the internal connection might now be intermittent due to temperature.

It's a little hard to tell from the picture, but get a close view of the pixel offset pattern. It seems to be 4 pixels. This would indicate an addressing problem into the video RAM.

[unknownsoldier90]

Thanks for your reply.

Here are two pictures, hopefully better

[strawberry]

repairing real one is more fun than just replacing
could replace when unobtainable custom IC or PCB is bad

warm IC?
part of circuit that operates normal is where is TEST function.

could be one off address lines permanently pulled high/low

byte address bus, bit address bus

[unknownsoldier90]

You are right, repairing is fun, but I don’t know how to start.
I have some experience in laptop repair, but laptops are so much different from this 😬
Also I just bought the oscilloscope to repair the DSA, so no big knowledge here.

Could somebody guide me, please?

[james_s]

This definitely looks like a digital fault to me, the defect is way too perfect to be analog. I would use a scope or even logic probe and start looking for stuck address bits in the digital section. I have no familiarity with this instrument, do you have a schematic?

[unknownsoldier90]

There are 2 service manuals, part 1 is for general service instructions and tests, part 2 does contain schematics.
Both manuals can be downloaded here:

http://ftb.ko4bb.com/getsimple/index.php?id=manuals&dir=HP_Agilent

Just type 3561 in the search box.

Digital Video board is called A60.

Unfortunately the only equipment I have is:

Cheap multimeter (Taglife DM6)
Soldering station
Hot air station
Hantek DSO 200 MHz/1 GS/s Ch.

[james_s]

That sounds like an adequate collection of equipment to troubleshoot this, even a basic scope will work.

[strawberry]

two way access to VRAM . CPU(Write to VRAM) or scan(VRAM - Video)
for each line CPU loads X Y address ,X Y increment address, line length
A60J100 tests A60 except RAM and CPU interface
A40W1 to troubleshoot RAM and CPU interface

[unknownsoldier90]

A60J100 has 4 pins,
1-2 is RUN position
2-3 is HALF bright
3-4 is FULL bright

Half and Full bright are okay, pattern is correct.

So I guess I have to continue with checking A40 board, right?

In the manual V2 on page 7-169 is described the following:



I did this test and the waveform is okay. But what is a signature? I do not understand anything of the things written in 7-51.

will continue with pin 9 of U903

[strawberry]

A40 should be fine if peripherals respond
A40W1 runs some test code stored in ROM but not clarity about that test
HP 5005A Signature Analyzer, decode error messages I guess. someone on youtube used it to find faulty RAM in some HP analyzer
could test BYTE ADDRESS BUS and BIT ADDRESS BUS basic activity with scope if there is any

[unknownsoldier90]

Hmm do you think it makes sense to move the topic into the test equipment section?

btw thank you very much for your contribution!

[slbender]

I have three of these 3561A units, only one works.  Those small unmarked small black parts near R203 are diodes, check them all, might be related to the cracked R203 if some are shorted, then replace R203 (2 watt)?  Also pair of power diodes part CR-100 is the main +5 volt pair of Schottky diodes, often a problem, if one is shorted or open, might cause the jumping or doubled display.  I’d remove it and check, it’s a TO-3 case, and also a known failure mode.  CR-100 only got about 10 or 20 Volts into it but needs to be like 20 or 30 Amps or more.  Since it’s on a heat sink, a TO-220 might also work if the original part is NLA.

Steven

[strawberry]

[slbender]

Someone else has put up seven (7) YouTube videos for repairing the 3561A DSA - maybe have a look?

Steven

[unknownsoldier90]

I will try to find a replacement for R203 tomorrow, 22 Ohms and 3 Watts.
But since the ripples aren’t too bad I don’t believe that it’s the cause for the display issue.
Will also measure the diodes, thanks

I already watched all YouTube videos of the 3561a, but no success.

The only video which shows a fault like this is here:
https://youtu.be/0eXNXumif1g

…ended in swapping a whole board

[strawberry]

replacing board is like giving up and changing work as dishwasher

[strawberry]

how to repair main and only multimeter with just soldering station and precision resistor and a week for figuring out how this meter works and where this fault could be
https://www.eevblog.com/forum/testgear/34401a-ohms-drift/msg3716629/
half year after purchase  when got used to this instrument got some strange suspicion

[unknownsoldier90]

I ordered a A60 board, if it works I can send you the defective board strawberry 😉😬