Need help for restoring an HP34401A multimeter

[JacquesBBB]

Dear All,

I scored a HP34401A in pretty bad shape. no power button, no case (but front panel).
I will attempt to restore it  with your help.

I found  various service manuals on the web, but not always very explicit.

I have shunt the power button, and power it  on from an external switch.


The unit powers on, the display is on, but nothing on the display ( all segments on ) ( see the attached picture).

I have measured  the voltages at the connector  that connects to the display (P500)



This is what I find

1:   GND
2:  5.029 V
3: 18.95 V 
4:  5.036V
5: -12.22 V
6 : 5.036V
7: -18.20V
8: 0
9: GND
10: 0
11: -12.22 V
12: 2.525V


Any hint ?

[JacquesBBB]

In the service manual of the HP34401A, they say to check  the main supply voltage, but where do I ned to check it ?

The only connector I found was the P500

Where do I check the  +7REF  and the +5REF ?

[chekhov]

Manual does not show test points, so according to schematics (page 162) at the input and output of appropriate voltage regulators. I believe there are a few dozens of "34401A repair" topics on the forum.

Also, maybe it's just shadow, but this tantalum capacitor near display connector already looks fried.

[merox]

Also, maybe it's just shadow, but this tantalum capacitor near display connector already looks fried.

That thing looks more than fried, it even looks kind of bulged 

[floobydust]

The Front Panel connector P500 voltages all look fine. This is the 1st generation uPD7527 MCU?
You have to be careful because the (Front Panel board) circuit common (GND) for the IC's is actually at -18VDC and everything is shifted to accommodate the high voltage VFD. It can be confusing.

On the Front Panel board, check the +13VDC rail is OK, and I would see if there is life in the Front Panel MCU if you scope the RC clock at pin 2/3. Check VPRE pin 4 is around +8V I think.

The tantalum cap C323 in the picture does look a bit discoloured. It is for the ACV circuit -15V rail/CR307-A but I don't think it would affect the display.

[JacquesBBB]

Thans for the feedback.

If the voltage on the P500 connector are fine, could it be that the display unit is faulty ?

The problem with it is that I am not very eager to desolder the display. Unfortunately, it seems that to inspect the circuit, you have not much choice.
I would really like to be sure that this is the thing to do before going further.

I attach here a few other pictures of the main  board.

[chekhov]

What I see on your last picture (maybe I'm not right) - something like little pieces of solder here and there - sign of warning to me.

A few things may fail, just a few thoughts:
 * tantalums on display board or on main board - check them.
 * RAM chips (btw, does meter beep at all ?)
 * CPU or it's reset line - check as well.

So still there are a few things before VFD desoldering. Maybe you have GPIB adapter or try to use RS-232 to communicate with unit - you really have to figure out whether brain has issues or is that really front panel itself.

[floobydust]

You can still take readings from the backside of the Front Panel board.  pic taken from https://xdevs.com 34401a repair.
The Main CPU issues a RESET pulse to the Front Panel processor if a watchdog trips. I thought the meter beeps when this happens.

[JacquesBBB]

Also, maybe it's just shadow, but this tantalum capacitor near display connector already looks fried.

That thing looks more than fried, it even looks kind of bulged 

The capacitor looks OK on the binocular. No bulge, no decoloration. 23 uF on  the multimeter ( without desoldering). No heat on the infrared camera. 

[JacquesBBB]

A few IR image of the unit under power done with my C200 InfiRay camera + ZnSe lens

[factory]

There is currently a newer one on ePay in the UK with possibly the same problem (all display segments lit), I suspect it's this one that had blank display a few months back, it had an SRAM failure that time;
https://www.vintage-radio.net/forum/showthread.php?p=1365376
I've no idea if the current fault is the same problem, or something else.

David

[JacquesBBB]

There is currently a newer one on ePay in the UK with possibly the same problem (all display segments lit), I suspect it's this one that had blank display a few months back, it had an SRAM failure that time;
https://www.vintage-radio.net/forum/showthread.php?p=1365376
I've no idea if the current fault is the same problem, or something else.

David

Thanks for mentioning this. Indeed, this  is similar. Although I have never hear any beep when powering the device on.

[JacquesBBB]

The Front Panel connector P500 voltages all look fine. This is the 1st generation uPD7527 MCU?

Thanks for your feedback. This is very helpful.

It is indeed the first generation display : 34401-66502

On the Front Panel board, check the +13VDC rail is OK, and I would see if there is life in the Front Panel MCU if you scope the RC clock at pin 2/3. Check VPRE pin 4 is around +8V I think.

I have checked the pin 2 and 3 . I found  clock signal on both pins : nice square, 5.36 V pk to pk , 563 khz on pin 3,  more deformed signal  460 khz, 5.44V pk to p on pin 2.

SO for the moment, it seems that the display is not the culprit.

[floobydust]

Upon power up, the Front Panel processor will light all segments, like what you are seeing. It waits for serial data from the main CPU.
It might work fine, is not the problem. It's just getting no data from the main board CPU.

You could look at the serial data lines IGFPSCK, IGFPDI, IGFPDO and see if there is communication. Try hitting buttons. No comms and error E601 happens and I thought it beeps.

[JacquesBBB]

Excuse my lack of experience, but I must say I am a little bit confused with the use of floating logic and Earth referenced logic.

I am contemplating the block diagram from page 9-7, and I am sure that with some concentration the light will come out, but a little help will certainly be welcome.

DO we agree that the floating earth is an AC signal of 50 hz and 6.82V amplitude (13.7 Vpp) ?

I have nevertheless  been able to check that the  oscillator is OK.  With 12Mhz frequency and two outputs at 3.66 Vpp and 2.24 Vpp.

[JacquesBBB]

You could look at the serial data lines IGFPSCK, IGFPDI, IGFPDO and see if there is communication. Try hitting buttons. No comms and error E601 happens and I thought it beeps.

I dont now if I did it right, but I could not find any activity  on these lines, even pressing the various key of the keypad.

[floobydust]

The entire multimeter is isolated so it's floating from earth-ground, EXCEPT for the RS-232/GPIB interface.
So multimeter common GND will have a little weak AC hum and noise present. It's OK to connect your scope ground there (which connects the floating logic to earth-ground) as long as you aren't making measurements to items that are floating to a potential themselves.

You might have to see which MCU seems to be hung or not running. The beeper should be making noise, it's driven by U700 on the earth-grounded side by the RS-232/GPIB section. Just weird. Is there data flowing across the opto-isolators at IGTXD, OGTXD? You could try connect to the RS-232 port and see if it's alive.

I would back up a few steps and check power the 5V rails like Vcc and PONRST from U553 in the power supply, and +5V_D.

Check for a stuck pushbutton, that will also cause the display to stay all segments lit - but it causes two beeps from the main controller MCU which you are not getting.

[JacquesBBB]

You might have to see which MCU seems to be hung or not running.

Are the infrared pics useful for this ?

The beeper should be making noise, it's driven by U700 on the earth-grounded side by the RS-232/GPIB section.

I confirm that I never heard the beeper.

Just weird. Is there data flowing across the opto-isolators at IGTXD, OGTXD? You could try connect to the RS-232 port and see if it's alive.

I already connected to the RS232 port. I had nothing. But I read in the doc that the GPIB is activated by default, but not the RS232 which need to be activated

I would back up a few steps and check power the 5V rails like Vcc and PONRST from U553 in the power supply, and +5V_D.

Will do that when I access again my workshop.

Check for a stuck pushbutton, that will also cause the display to stay all segments lit - but it causes two beeps from the main controller MCU which you are not getting.

I doubt on this one as I had the same behaviour when  the keypad was removed ( display connected ).

[JacquesBBB]

I would back up a few steps and check power the 5V rails like Vcc and PONRST from U553 in the power supply, and +5V_D.

It seems that there is indeed a problem here. Still investigating.

By the way, I am surprised to see that the LM340 and LM337  are in To-220 packages and not in special SMD packages. Is that  common, to just bend
then pins of a TO-220  to make it SMD, or is it a special version ?

 

[JacquesBBB]

I was too optimistic. The +5V_D is perfectly at 5V, with no ripple,  and still no beep .

I am not sure  +5V_D was not always good,  even before I mande any change ( replaced the LM340) and the associated tantalum cap.

[Swainster]

Hi Jacques,

I'm working on a 34401a repair at the moment. I'm think it is likely that your display is fine, or at least not the only problem, as when I remove the display on my 34401a it still beeps. With a working display attached it gives one or two short beeps, and with no display it gives 2 long beeps.

I would suspect an issue with the mainboard floating logic. Perhaps a good place to start (after the obvious power supply stuff) is the XPONRST signal (power good) from pin 5 of U553. Can't really add much more as my problems are all analogue related - my floating logic section seems to be working fine. Also might be worth scoping out the signals on the optocouplers (U596, U704) to see if the ground referenced logic is talking to the floating logic.

Good luck!

edit: the LM337 on this board is a variant of the SMD D2PAK TO-263 package. These dont normally have the mounting hole though. You can easily buy the D2PAK variant without mounting hole, but I bent the legs of a TO-220 to save buying more parts.