HP 34401a DMM with leaking segments

[iMo]

Any idea what is the filament current of the original display?
Thanks.

[coromonadalix]

10 ohms of impedance / resistance

[iMo]

Ok, 10ohm, and the voltage is 6V (when looking into the SM p.9-7) therefore 600mA AC, imho.

[coromonadalix]

6v in 10 ohms is 600ma not 60 ma

[iMo]

That is a long back I saw something like 600mA

A pretty large current, my display is normally off (I use BT), so I think I will switch the filament off as well.. Another project..

[coromonadalix]

you can turn the display off thru some commands,  you can find the user manual easily

page 132
http://ecee.colorado.edu/~mathys/ecen1400/pdf/references/HP34401A_BenchtopMultimeter.pdf

DISPlay {OFF|ON} 
Turn the front-panel display off or on.
[Stored in volatile memory]


http://cna.mamk.fi/Public/Slabra/Agilent/Multimetro_Digital_34401A_Quick_Reference_Guide.pdf

[qu1ck]

That will not cut filament power, it is not digitally controlled.

[coromonadalix]

yep you're right  quick

@imo  do you want to save the remaining life of the vfd ??

[iMo]

Sure, the filament is permanently on. The command (via remote or from the keypad) switches off all the segments (such it deloads the 2x18V anode voltage) only.

As a first step I want to decrease the power consumption by 4Watts

[qu1ck]

If I recall correctly you can just pull 2 brown wires from the transformer that provide filament voltage from the power connector block that goes into main board. Isolate the pulled contacts and that's it.

Are you using some sort of GPIB to bluetooth converter?

[iMo]

I've been using rs232 to HC-05. Powered from the meter.

[coromonadalix]

You could use an i/o line to switch of your vfd

You have pio2 thru 11 to play with ...

https://www.rhydolabz.com/documents/HC-05_datasheet.pdf

the command at page 11 section 17  should give you what you need

https://www.rhydolabz.com/documents/HC-05%20AT%20Commamd.pdf

just add you on / off circuit on pio line   Ie : an small npn transistor with an small 5v relay ??

[iMo]

I plan to use an MCU put inside the meter for such tasks..
The HC-05 will stay outside the box.

PS: the toggling with PIO pins works in AT mode only, afaik.
I've been using it as a dumb serial device passing bytes at 9k6 8N2 to/from the meter.

[coromonadalix]

The 34401 can constantly output  " talk only "    if you set it to adress 31 and select gpib or rs232 output ??


From Keysight

In the "talk only" mode, the Agilent 34401A automatically sends readings directly to the GP-IB or RS-232 port. To enable the "talk only" mode, first set the HP-IB address to "31" using the front-panel I/O Menu. Then, select either the GP-IB or RS-232 interface (also in the I/O Menu)

[iMo]

Yes it can do via rs232 with the 31 address. It sends out strings (scientific floating point, like "+9.99995790E+00") with results.
For example you may read the rs232 strings with an external MCU and display them on any display you wish then..

PS: the filament changes its resistance when ON, imho. So the current will be smaller after the filament heats up..

[hakko]

HI! I scored on 34401A (agilent dated december 2006) it have a dead front panel vfd. It does not show any digit  . The front panel have a EE80C51BH 34401-88813 REV 2.0 and HV518PJ vfd driver. The meter respond for key press.
Investigating the problem I had decoded the the data from the main board cpu to front panel with a logic analizer the data will be attached here. I also decoded the serial bus from 80C51  ---> HV518 and it have clock and data activity also latch pulse signal and  strobe are always at 0 state. I think the mcu is fine but it have a dead vfd driver. The board have no shorts or huge heat elements (checked with thermal camera).
I removed the vfd driver and applied +18V (with current limit resistor) on plate and anode segments for a short time and the vfd segments light up so the vfd display it's not faulty 
Anyone with this board rev 66522 had this problem before? I ordered HV518 from utsource but I don't know if it's good quality 

[coromonadalix]

the hv518 can be ordered here

https://www.digikey.com/number/en/microchip-technology/150/HV518/114088

[hakko]

But the shipment is 40$ for a 5$ part 

[coromonadalix]

Damn   in my location, its nothing

But for Brazil  its expensive, just checked

An complete 34401a  pn 34401-66522    board+vfd included  is around 200$ usd  from Keysight, and they dont have tons of them left ...

And you cant swap them with an older version of the pcb, the meter FW  wont allow it.

service note : 34401A-06


My curiosity would be :   retrograde the firmware and try the older display pcb with the unobtanium drivers ??

[hakko]

Damn   in my location, its nothing

But for Brazil  its expensive, just checked

An complete 34401a  pn 34401-66522    board+vfd included  is around 200$ usd  from Keysight, and they dont have tons of them left ...

And you cant swap them with an older version of the pcb, the meter FW  wont allow it.

service note : 34401A-06


My curiosity would be :   retrograde the firmware and try the older display pcb with the unobtanium drivers ??

It might work with old firmware but i don't know if the meter will work fine since newer main boards have also hw changes from the old ones specially in analog section I don't know if asics are also changed too.
HV518 is also used by arduino guys I think china ones will work since it's quite usual ic.

[qu1ck]

hakko
Thanks for interesting data. I looked at SPI data dump you posted and the control frame headers on new panels differ from the old, no wonder my firmware doesn't work. But I kinda expected that.

For example first thing immediately obvious is that text control frames start with 0xff 0x2d instead of 0x00 0x7f and annunciator update starts with 0xff 0x00 instead of 0x7f 0x00.
Also there seem to be random 0x00 frames from cpu to panel sometimes (button state query?).

If you have time could you compile a table of control frames like found here? https://github.com/openscopeproject/HP34401a-OLED-FW/blob/master/protocol.md
That would help update firmware for these new panels.

[qu1ck]

Thanks to coromonadalix who stepped up to help with capturing SPI data on a newer model 34401a meter I was able to figure out the differences in the protocol, which turned out to be minimal.
Here is new firmware that should be compatible with both old and new meters. firmware_debug.bin will spit out received decoded data onto usb serial line. If anyone can validate it I will publish new firmware and the fix to github.

[coromonadalix]

had  tons or problems to make a blue pill work,  boot loader problems i did not understood    i'm a total nooob  for this

He sent me an firmware.bin  file who replace the boot loader, and have the debug option,   thus always needing an st-link to work.

Heres the results on a serial monitor (attached photo)  , i dont have an oled display, never received it.

my comments
-----------------------------------------------------------------------------------------------------------------------
Inverted the miso mosi line going to the blue pill    (if im not mistaken)
pb13 is the clk
pb14 is the 34401a mosi line / blue pill miso
pb15 is the 34401a miso line / blue pill mosi
ground wire

Arduino IDE board enumeration is :

BN: Maple Mini
VID: 0x1EAF
PID: 0x0004
SN: Upload any sketch to obtain it

Select com9 ; it says  Maple Mini

I start the Arduino serial monitor and voila   is see the data coming out ...  woh hoooooooo

Q: What are the 2 lines of Annunciators meanings ?
they have 0000  0001   etc ...

----------------------------------------------------------------------------------------
Going right to left in the menus

Dcv - Acv - 2w / 4w ohms - freq - period - DCI - ACI   all went fine, annunciators follow fine,   played with the range arrow more digits less digits    all ok

-----------------------------------------------------------------------------------------------------------------------

problems

Continuity          spits data very fast  and crashes
Diode test         spits data very fast and crashes  the pill red light flashes ?

The rs232 monitoring seems to crash the communications ... not enough buffers ?? or slow down theses tests ??

Have to unplug power or a long reset and restart rs323 monitoring

If i do a long reset the red led flashing is gone, restart monitoring, all is fine, if i'm outside  continuity or diode tests.


Goood   job you have done     thks for your patience and many emails exchanged.

[coromonadalix]

Edit    with continuity open   data is going very fast

rs232 monitor   Termite  hold longer  before it crashes too ??

Termite wont reopen the connection   even if the pill is resetted
Arduino monitor does   after the pill was resetted

[wictor]

It would help, if serial printing had lower priority, since it is disturbing spi decoding during debug.