OWON XDM1041 the unknown multimeter...

[RogerG]

The last weeks the batteries are showing full. No drain visible now.

[abeyer]

I have no idea what you're trying to tell us.
What does this thread have to do with Zeeweii?

aliex has been doing that to me frequently of late... if you're looking at something new/unusual and the AI doesn't have much info on the product it sometimes starts spewing nonsense about some other unrelated product in a similar category instead.

[Aldo22]

aliex has been doing that to me frequently of late... if you're looking at something new/unusual and the AI doesn't have much info on the product it sometimes starts spewing nonsense about some other unrelated product in a similar category instead.

Got it, AI slop. 

My point regarding the XDM1051/1251 was that I don't think the accuracy matches the 120,000 counts.

[Kleinstein]

The resolution / number of digits is a function of the chip set used, especially the ADC.
With advanced chips more resolution to this level can come cheap.
The accuracy is to a large part set by the imput divider and voltage reference and maybe the calibration.
Accurate resistors are expensive and it is hard to predict how much they drift over time. Especially Chinese companies tend to be optimitic with the specs. So I would not give too much to those numbers - it is nothing they can measure, like the noise or weight. Only a long time will tell how much the meters typically drift and what specs would have been right. However those data are usually not available when the specs are written.

[Phil1977]

I assume something more stupid:

The design uses a HY3131 DMM IC - generally not a bad choice. This chip is specified for 50.000 counts and has 17Bit Output (65536 values + Polarity).

Owon seems to extract 5000counts more while (probably) still in spec - that´s quite usual overexceeding of the datasheet specs.

The new Chinese clone has tickled another 5000 counts out of the IC - so it´s 60000. For marketing stupidity it counts positive and negative range together, resulting in 120000counts.

I´ve seen this a lot on Chinese platforms 

[coromonadalix]

you have 2 other candidates, a Cirrustek, Fortune semiconductor    and the known Hycon

if you search dmm asics or dmm front ends  you may find the good one ?  i only know a 100K count chip ?   unless a new one came out ?

[floobydust]

From XDM1041 board pics, it's using the HY3131.
I think they are just using decimation to get an extra digit out of it, like a few other manufacturers do, getting 500,000 counts but slower.
XDM1041 has provisions for either of two ref IC's, usually U17 is populated SOT23-6, optional U6 SOIC-8 possibly for a MAX6190 which is typical.
edit - the SOIC-8 U6 I see only two pins connected so it is a 1.2xV shunt reg. Pin 8+, pin 4 GND. Pinout matching LM385D-1.2, AD589, ICL8069 etc.
Unknown shunt reg U17 SOT23-6 pin 1+2 GND, pin 3 +.

[Muck]

XDM1041 FW v4.5.0


Does anyone have the original calibration constants (COEF) for 5 VDC and 50 VDC for FW v4.5.0 ?

Mine was deleted due to an incorrect action.

[Muck]

OS for XDM1041, got it from a colleague, doesn't work well for me. 

My device is no longer working properly. Maybe someone else can use the OS. 


 OS--XDM1041_OS_v4.7.0.7z (74.55 kB - downloaded 37 times.)

[betelgeux]

XDM1251 teardown images (part1), if anyone curious.

[betelgeux]

XDM1251 teardown part2 images

[Aldo22]

Today I got a used XDM1241 for CHF 45.
It's in very good condition, has 1000V fuses, and is running firmware version 4.2.0.
Does it make sense to ask Owon for a newer firmware?

[coromonadalix]

ain't broke   do nothing,   if you see some   do check 1st  the update logs if there is one, ask Owon ... 

and or check others members / threads,  if they have installed it, and it got better  or worse ?

[Aldo22]

ain't broke   do nothing, 

Yeah, you're absolutely right.
4.2.0 is already a pretty advanced version number, and you never know what has changed anyway (no changelog).

Otherwise, I like this little guy. I have the battery-powered version.
The accuracy and the display are very good.
The True RMS bandwidth is also much better than specified (1 kHz). I measure -3 dB at 8.4 kHz sine. Up to 4 kHz sine, the error is still under 1%.
I also like the Max-Min-Average display on the same screen as the normal measurement. SCPI works, too.
It has a few quirks, but for the price I paid for it used, it’s really a bargain. And it’s cute... 

[Aldo22]

I wrote a small script in Python this evening to read out the internal data (from the data record function).
It’s possible something like this already exists—I have no idea.
The script has no checks and options; it simply downloads the data and saves it as a CSV file (owon_record.csv).
You need to specify the serial port as a parameter.
This function might require a specific minimum firmware version.

The result looks something like this:

Freq;1.0100;kHz
Freq;1.0110;kHz
Freq;1.0120;kHz
Freq;1.0110;kHz
Freq;1.0100;kHz
Freq;1.0100;kHz
etc. etc.

CSV data imported into PulseView.
1 Hz sine wave (interval: 0.015 s, points: 1000)