uni-t 81b mulimeter /scope

[darkspr1te]

Hi All
I have this uni-t 81b scope meter that was fine but one day i turned it off then wanted to read another value so switched it back on and it's now stuck at the "Detect test-mode.." screen. I have searched for possible answers but found only one result where uni-t said it was a major component failure and would need to be replaced.
I cannot return the item as i had it bought and shipped to me in africa so no returns sadly.
It has a cyclone fpga and lattice chips as well as the main multimeter chip.
The activate calibrate option seem to work (press HOLD when detect-test mode is displayed) and it ask's for a password but any button push send you back to the detect test mode screen

Anyone have any ideas of what to check ? i want to see if there is a solution before i decide to climb in with jtag and such to discover why it's hanging.


darkspr1te

[darkspr1te]

So while the uni-t 61d in the thread next door is getting all the love I decided to dive in a little more and do some nudes, pcb nudes that is,
Also some I.C's of note (i kinda suspect i know whats failed now i studied the board layout and also i have a secondary culprit too)

Main DMM IC seems to be Fortune IC FS9721-LP3
https://www.ic-fortune.com/upload/Download/FS9721_LP3-DS-21_EN.pdf


theres not much else on this board other than interconnects.

theres the daughter board now which contains the following :-
Altera Cyclone 2
https://datasheet.octopart.com/EP2C5T144C8-Altera-datasheet-7627975.pdf


NXP LPC2138 mcu
https://www.nxp.com/part/LPC2138FBD64#/



Analog Devices AD9283 8 bit 50ms/s adc
https://www.analog.com/media/en/technical-documentation/data-sheets/ad9283.pdf


Analog device ADG711 quad switch
https://www.analog.com/media/en/technical-documentation/data-sheets/adg711_712_713.pdf




Now i suspect two possibilities here, one is the rounded top cap on one board (pic enclosed) or the other is that maybe the MCU is waiting on data from the dmm chip (detect self test part) or error in firmware on the mcu, the last one i dont think so as it allows me to activate the calibrate menu but then needs a password, but that shows that side of the firmware(mcu) is not locked up although there could be a eeprom about with currupt model specific data that i've not seen yet, a few chips that could fit the bill are too small to read easily.


 

[darkspr1te]

Some additional reading ,

https://lygte-info.dk/info/DMMDesign%20UK.html
https://sigrok.org/wiki/Multimeter_ICs
https://www.improwis.com/projects/reveng_multimeters/
https://www.eevblog.com/forum/testgear/uni-t-ut81c-scopemeter-teardownreview/


darkspr1te

[darkspr1te]

So further reading led me to the ut60e schematic which my front end board seems to compare very well with the ut60e model, So i have some testing points to go by for now. wish there was a easy way to reattach the dial, it always bugs me it's not attached.


darkspr1te

[darkspr1te]

Well scoping the outputs from the main DMM IC FS9721-LP3 shows it's output a display,
scope pic attached.
 so now will have to look at the adc devices and switching.

[shakalnokturn]

Displaying "detect test mode" for a short time when switching function is normal, I think DSO/CPU  board waits for serial data from FS9721 to know what mode that has been set to.
I don't know if lack of serial data in scope mode would freeze the display all together or only prevent  voltage/current readings, it may be worth having a look if serial output is reaching the CPU.
If your trouble showed up during the rain season, I'd certainly have a look at the function selector switch contacts.
You may also be inspired by the troubles I had with mine:
https://www.eevblog.com/forum/repair/uni-t-ut81b-scopemeter/
The fix is easy when you know it but does involve desoldering the interconnect for cleaning.

Calibration mode is only for waveform display, all the DMM adjustment is done with trimpots.

[darkspr1te]

Displaying "detect test mode" for a short time when switching function is normal, I think DSO/CPU  board waits for serial data from FS9721 to know what mode that has been set to.
I don't know if lack of serial data in scope mode would freeze the display all together or only prevent  voltage/current readings, it may be worth having a look if serial output is reaching the CPU.
If your trouble showed up during the rain season, I'd certainly have a look at the function selector switch contacts.
You may also be inspired by the troubles I had with mine:
https://www.eevblog.com/forum/repair/uni-t-ut81b-scopemeter/
The fix is easy when you know it but does involve desoldering the interconnect for cleaning.

Calibration mode is only for waveform display, all the DMM adjustment is done with trimpots.

Ha Ha, talk about timing. I had just finished drawing the schematics rough draught last night and had noticed the ONLY way the two boards could tell which position the mode switch was in was the serial output from the primary DMM chip.
And of course i wake up to your response, really gave me a chuckle.
I did check serial output from the DMM chip when it first failed and no output, oddly though it does have some output to the "DMM LCD OUTPUT" showing some function but still no serial data making me think the  FS9721 failed as i am probing at the chip output and have confirmed TX_EN is grounded and MEA1-4 are switching at the chip.
My plan today was to inject some serial data into the line to confirm last nights theory which today you confirmed.
hopefully i can purchase one of  FS9721 chips online without headaches


darkspr1te
 

[shakalnokturn]

Considering the hassle I had with mine and going by your photos that seem to confirm the standard for uni-t is to leave the PCB's covered in milky flux residue, I'd go to the trouble of making sure that the serial output is not present even when output is confirmed to be unloaded.
If you're confident enough the FS9721 is bad, I'd lift the TXD leg, slip a piece of kapton tape under it and scope output like that.

[shakalnokturn]

Did you have success in repairing this?

[darkspr1te]

No, but i did confirm my theories as i ended up returning it under warranty and checked the replacement which did have working uart output which goes directly into the second mcu.


darkspr1te