HP34401 Broken uC 87C51 U700

[lucav]

Hello to all,
I have a 34401 that signal error 625-uController communication error between U500 and U700.
Following schematics I checked U704 and U506 and it seems that communication is broken at U700, that receive but don't respond.
U700 is marked as 34401-88802 and is based on 87C51cpu.

I found that oscillator is not oscillating, and reset pin is always about 3.5V despite U705-E correcly drive it low.

Now, the problem seems to be the 87C51 that burned out, since driving manually low the reset pin make the 5v supply also fall...

My question is: do you have any other test that I could perform?

Thankyou very much for ypur help.

Luca.
 

[lucav]

Hello,

sorry to bother another time, but I didn't get a solution for the broken 34401 and is really a pity that multimeter works but I/O functions are dead.

I would ask if someone have  an idea if a flavour of commercial OTP 87C51 could replace the i/o microcontroller (U700) if there is somewere a dump of that firmware to flash the replacement.
The mark on the U700 says 34401-88802 rev01.

Thankyou very much for your appreciated help.

Cheers,

Luca
 

[coromonadalix]

U700  is the main mcu / processor, U500 is the display Mcu       schematics for the 34401a  exists see below

and check the white connector,  sometime the flat cable get stuck in it,  some wire could break,   etc ......  this connector may fail, flat cable may fail ....     and you have to push down pretty strongly  if it never was removed, the wire kinda glue on the contacts, be careful ...

1st  i would check all the supply lines,  since  it's kinda old items ....   you need to be sure all the supply lines are ok

Normally  no ic's should roast

Dumped firmwares do exist too  ....    up to v10  for the display,     but we need to know if its the very old model display with unobtanium

vfd drivers,  and the more recent one     you have  to show us  pcb numbers  ...  some photos etc ...


https://www.ecb.torontomu.ca/guides/instrument-manuals/Agilent-HP_34401A_Service_Guide.pdf   schematics and part numbers in it, this  manual as the recent  front panel pcb  ....

[lucav]

Hello coromonadalix, thankyou for reply,

I have not problems with the GUI that works, the problem is related to U700 that tie down the supply to aboit 2.4V and have no oscillator activity.

As the schematics you attached (and consistent with mine) show, U700 is the I/O controller that  manage RS232, HPIB, VM Comp and Ext Trig and the beeper and is interfaced to the other parts of the board by 2 optocouplers U704 and U506 (you can check at page 9-14)

The multimeter power up and error alarm light up.
Error history show 625: I/O processor does not respond This test checks that
communications can be established between U500 and U700 through the
optically isolated (U506 and U704) serial data link. Failure to establish
communication in either direction will generate an error.

I have checked U704 that route communication from main board to U700 and it works.
I also changed U500 but no luck, before I realized that U700 have a not negligeable amount of heat up and  bring down power supply line.

I agree that is unlike that a uc die, but as I can check... it seems really that.

Please find some pictures attached.
Any help?

Thankyou

[Miti]

Check this, reply #35.
https://www.eevblog.com/forum/testgear/agilent-34401a-revisions/25/


This guy upgraded the comm cpu. You can ask him where he got the binary. I think I saw it on Xdevs but you need to ask for a user name and password now.

Cheers,
Miti

[coromonadalix]

i have xdev's access  ....    pop me your needs


here the u700 and U502   dump  on kO4BB   attached here


eeproms / fw versions / infos     from xdev's    readme.txt   attached here

[factory]

Looking at your pictures, your older HP 34401A has the clearance problem between the GDT and the relay, this can flash over as there is only a thin layer of paint between, see here;
https://www.eevblog.com/forum/testgear/check-your-34401a-multimeter/

Just finished repairing mine today and have added a thick piece of insulation between those parts. The fault with mine was a complete failure of AC measurements and the cause was several failed tants  & two bad zener diodes associated with the tants.

David

[RolandK]

They use some zener / tantalum combinations for low voltage supplies. Instead of the cpu it may be one of the tantalums or zeners.
The +5V_0 is generated on page 158 by U751. Mesure
- U553 voltage pin 1 and 3 AC and DC.
- voltage between u553 pin 2 and all +5V_0 points near u700. the short is near the highest voltage drop. Normal would be about 10mV to 100mV
 
C753 is a 1u polarized. Power off and Just measure it's resistance. it should go quick to some kiloohm. Compare this to the other power rails, each before and behind the voltage regulator.

Explanation:
On most Multimeters in Ohm mode is + delivering the positiv test current. This goes to the + of the capacitors.
The test current loads the capacitor with charge time proportional to it's capacitance.
The final displayed resistor value should be about proportional to the capacitor value, as the size of the capacitor is proportional to the foreseen load.
Just found this way 4 shorted 16V electrolytic capacitors in a 4275a from 100µ to 10000µ - the only errors in that one.

[lucav]

Hello all,

thankyou very much for your attention on this problem, I really appreciate it.
I receipt your inputs and followed your suggestions.

Update:
Searching for a overload source for supply, I ended up desolder U701 (gpib controller TMS9914) that I felt heating up to 60°C.
Desoldering that IC had the effect to free the supply voltage to it regular value.

Testing resistance resistance between Vdd and Vss of the TMS i read 45ohm either polarity. Defenitely a bad value.



I think those are acceptable values.

• Following RolandK test i had these results:
  Voltage across C753: 4.96V
  Voltage across JM751: 0.8mV
  
  Voltage input U553: 10.15V
  Voltage Output U553: 4.85V
• Checking factory issue, I can see a minimum gap between E100-101 and K101, there is a small air gap. Anyway I pushed E100 thoward S1.

Now U700 is correctly powered, U701 is missing because desoldered.

BUT.... the thing still doesnt work!

Oscillator of U700 is dead. I tried to takeout Y701, put two capacitors of 30pf and a cristal 12MHz but nothing happened.
Looking at signals, I saw Reset (pin 10) was high, so uC is held in reset state.
Look at the schematics, the chain that bring U700 to reset, became from U704.
Main board, after trying to establish communication, if doesn't have response send a pulse of about 80usec every  400msec.
In the image joined (sorry for the italian speach) you see pin 4 of U705B. Due to different tau introduced by R704-R708 and CR703, input of U705E is always high, and output always low.
So... why reset of U700 is high?
I tried to force reset pin down to gnd and the uC started to heat up to 60°C...

I don't know what happened to those ICs... but it seems that something bad happened there. Just a day turned on multimeter and this was the situation, apparently no event has trigger the damage.

So I'll try to find a couple of 87C11, I'm not sure I need to populate the GPIB controller too, but ok.. maybe I proceed one thing at time.

Thankyou very much coromonadalix for the firmware and Miti for the insight.

I think (or I hope) to have delimited the problem and I'll follow up on the results.

Cheers, Luca.

[coromonadalix]

mmm seems a catastrophic  internal supply issue(s) ??  failed tantalum or electrolythic, zenner diode etc ...  lots to check   


mostly the digital psu section, where all the main mcu's reside ...

oh by the way   is the main transformer voltage are set right ??   i had an blown xformer in a 34401a   because of bad ac settings