Protek D620: 100,000 count forgotten multimeter.

[jchw4]

I got my hands on a very unusual 100,000 count multimeter with trend capture: Protek D620.

The information on the web is very scarce: I was able to find only a few datasheets, User Manual in Polish (only!) and a Calibration Manual.

This site has some data and photos: https://www.gotronik.pl/multimetry/protek_D620.HTML

Web archive has only a single partial capture of the product page https://web.archive.org/web/20040811163957/http://www.protektest.com/ProdInfo.asp?prodId=620 .

I could not find a software for this meter in any "Download Software" captures https://web.archive.org/web/20080129111348/http://www.protektest.com/ProdDnLd.asp?type=SW .

If somebody has software or a user manual, please share!

Somebody on this forum mentioned that there was a protocol description somewhere. Anyone?

[jchw4]

The meter was dead, consuming 100mA at 3V from my power supply even when off.

After disassembly, the damage was clear.

U11 mark was difficult to read, but it turned out to be TPS763 5.0V LDO (SOT-23-5 marked PBGI). Replaced with recommended TPS709.

R3 was nearly unreadable too, but turned out to be marked "10R0" and still measured 10 Ohm!

[jchw4]

The meter looks very nice, has 100,000 counts in all ranges, live trend capture and glowing buttons! Not really as sci-fi looking as my phone decided to present, but still nice .

Meter seems to be in spec compared to my DMMCheck+ except 100k ohms range, which is significantly off. Did not investigate this yet.

My main concern was about pretty long whiskers that grew out of the fuse holders. Most of them fell off before I was able to capture them but here is a photo.

[jchw4]

Meter PCBs.

[Gyro]

That looks like a nicely featured meter for decent resolution, display functionality and data logging.

Treat the indicated CAT ratings with utmost suspicion though, there's no way that they align with the PCB photos.

[coromonadalix]

the meter chipset must be hidden under the lcd   i see a nec mcu   but ??


Oh sorry  there is another piggyback pcb it seems   


Nice find  loll  lucky you   loll

I do remember  one other meter / brand who had 80,000 count display, but cant recall the model

EDIT  Protek 6800

[Kleinstein]

I don't think this meter is using a normal DMM chip set. There is a 24 Bit SD ADC (CS5522) and lots of CMOS switches. So chances are this is more like a µC (the large NEC part), ADC chip and range and function switching with the CMOS switches. The 100000 counts may still be quite conservative for the ADC, but the reference could well be limiting.

[coromonadalix]

yeah  damn eyes, the row connectors are between the front panel and meter section ...   

Some APPA metres go to 100k counts
https://www.eevblog.com/forum/testgear/rs-pro-dmm-the-idm505-100-000-counts-0-015-dc-v-accuracy/

[johansen]

I have one of these, but need a high voltage source to calibrate it..
it reads about 15% high on all ranges as my friend didn't get all the way through the procedure.

[NoisyBoy]

Odd choice for button backlight, when each button looks the same without any writing on it.  They must assume the user would remember the function of each button. 

[jchw4]

I placed a thermistor on top of LT1004, fired up air con and did some logging of a LM399A-based 10V reference.

(Extracting data from the screen photo is pretty painful  Could somebody share the orifigal communication software and a user manual? Anyone?)

Playing with numbers I created a pretty close approximation of the value based on temperature (see PPM graph). So it seems like my meter has about 90 PPM/C.

It certanly worth improving. Install MAX6325 instead of LT1004?  But I don't have equipment to calibrate it later, so it's probably not going to happen.


Upd: reattached the photos.

[jchw4]

Odd choice for button backlight, when each button looks the same without any writing on it.  They must assume the user would remember the function of each button. 

It's not backlight. It's glow-in-the-dark plastics