EEVBlog 121GW Discussion thread

[jopie]

Got my meter this morning, I live in the Netherlands.   

I followed the package through the DHL website for 17 days after I got their tracing code. For some reason it was stuck in a DHL distribution centre for nearly 2 weeks, which was quite exciting...

[M4x]

Happy to read your feedback here.

As the official EEVBlog distributor, we now also have the 121GW available for purchase in Europe:
https://www.welectron.com/EEVBlog-121GW-Data-Logging-Multimeter-with-Bluetooth

Thanks for pointing that out! Too bad this information got to me too late (I've followed Daves "limited amount of meters in stock" e-mail and ordered directly from him.) The meter is sitting in Frankfurt right now...

[EEVblog]

FYI, Welectron in Germany have some meters in stock:
https://www.welectron.com/EEVBlog-121GW-Data-Logging-Multimeter-with-Bluetooth

[M4x]

Very funny ;-)

[eV1Te]

My review of the meter: https://lygte-info.dk/review/DMMEEVBlog%20121GW%20UK.html

Interesting review, I found the input impedance measurements especially surprising. The fact that the meter has different input impedances in different mode and ranges would make it incompatible with high-voltage probes etc. that rely on having exactly 10 MOhm for voltage division.

Having an input impedance as low as ca 2 kOhm for frequency measurement makes limiting for many things. For example measuring the frequency at different points of a 555 timer (or other oscillator) circuit without knowing about the low impedance could result in many strange readings (different frequency at different points, since the meter would affect the circuit by a lot).

[HKJ]

Having an input impedance as low as ca 2 kOhm for frequency measurement makes limiting for many things. For example measuring the frequency at different points of a 555 timer (or other oscillator) circuit without knowing about the low impedance could result in many strange readings (different frequency at different points, since the meter would affect the circuit by a lot).

Nearly all multimeters have that low input impedance on the frequency input (I measure input impedance for different modes in all my reviews).

[jancumps]

Is there any uncertainty? Something you can't take into account while measuring?

[HKJ]

Is there any uncertainty? Something you can't take into account while measuring?

If it is my impedance measurement you ask about, I am using a SMU to supply (DC) voltage and measure current, it is fairly sensitive to noise and I use a shielded cable for the test. There is a few meters that do not terminate the input at 0 volt, but at maybe 2V.

[Dr. Frank]

I have reported https://www.eevblog.com/forum/testgear/eevblog-121gw-discussion-thread/msg1687298/#msg1687298. that the 50MOhm range of my new 121GW was out of specification, by 1.4 ..2%.
Today, I have re-adjusted this range.
At first, the manual was wrong, as a 50 MOhm reference resistor is required, instead of a 40 MOhm one.

Then, due to higher leakage currents over protection diodes at 50 MOhm, compared to 10 or 20 MOhm, this range is quite non-linear, like I have found with my BM869.

So the readings at 10M and 19M are about 0.2% higher than nominal, when the 121GW is correctly calibrated at 50.000MOhm.

To compensate for this effect, it makes sense to use a 50.05 MOhm reference instead, so that at 50MOhm, the nominal reading is 0.1% low, but at lower values, like 10MOhm, the reading is only 0.1% high.

So the 50MOhm range is now very precise and stable, and also is not affected by mains disturbances.
The 5MOhm range though, is noisy, due to lacking 50/60Hz suppression, and probably a missing filter capacitor.
I could silence this range, even with strong interference from mains, by a parallel low leakage foil capacitor, 100nF, MKP.
I still hope for FW change by UEI, to correct both features.

Frank

[tpw_rules]

I have reported https://www.eevblog.com/forum/testgear/eevblog-121gw-discussion-thread/msg1687298/#msg1687298. that the 50MOhm range of my new 121GW was out of specification, by 1.4 ..2%.
Today, I have re-adjusted this range.
At first, the manual was wrong, as a 50 MOhm reference resistor is required, instead of a 40 MOhm one.

Then, due to higher leakage currents over protection diodes at 50 MOhm, compared to 10 or 20 MOhm, this range is quite non-linear, like I have found with my BM869.

So the readings at 10M and 19M are about 0.2% higher than nominal, when the 121GW is correctly calibrated at 50.000MOhm.

To compensate for this effect, it makes sense to use a 50.05 MOhm reference instead, so that at 50MOhm, the nominal reading is 0.1% low, but at lower values, like 10MOhm, the reading is only 0.1% high.

So the 50MOhm range is now very precise and stable, and also is not affected by mains disturbances.
The 5MOhm range though, is noisy, due to lacking 50/60Hz suppression, and probably a missing filter capacitor.
I could silence this range, even with strong interference from mains, by a parallel low leakage foil capacitor, 100nF, MKP.
I still hope for FW change by UEI, to correct both features.

Frank

The meter has a bonus 50MΩ nonlinearity calibration mode, which seems undocumented and I can't claim to understand. Regardless, maybe you can back up your calibration and give the following procedure a try:

• Calibrate the 50M offset as documented.
• Calibrate the 50M gain with a 50.000M resistor as documented.
• Enter calibration mode in the 50M range and connect a 30.000M (I think...) resistor. Hold MEM (instead of pressing it) to perform nonlinearity calibration.

This procedure just changes the influence of a hardcoded nonlinearity offset table. That table might be wrong, or it might not accurately capture the error of all meters. It also could be plausible that you use a 40.000M resistor as indicated in the manual in place of the 30.000M resistor, or you use a 40.000M resistor for both steps. I'm not entirely sure.

The 5M range also has a nonlinearity correction applied, but it cannot be calibrated.

[Dr. Frank]

Dr Frank, if you have a way to run some sort of temperature study on this meter,  I would be very interested in seeing you install TPW_rules's patched version of 1.02 and trying it out.   The prototype has always drifted pretty badly which I assume is why they changed the reference circuit.    It sounds like you are already trying to find the sweet spots to use for the alignment to bring it in tighter.   I wonder how the TC matches what is specified in the manual.

Well, I don't need FW 1-02 to recognize a very high T.C. on the 50M Ohm range.
We have a heat wave since weeks here in Germany, outside it's 35..38°C every day, so meanwhile even my cal lab is  affected, the basement and the 121GW currently have 24°C .. 24.5°C instead of the nominal 21.5°C, and I even get 29°C inside the instrument, when I take it from my room on the first floor down to the lab.

This 5°C difference causes maybe 0.5% change in the 50M Ohm range, so the 10M reference resistor seems to be of very bad quality... the voltage reference should not play a role, as Ohm will be realized in ratio mode, but I may check DCV separately (did not notice such a high T.C., though).

That explains, why UEI has specified such a bad accuracy of 1.3%, but w/o declaration of  temperature range.

Frank

[racemaniac]

I just received tracking info for mine, and am about to leave on vacation for 2 weeks >_<
Anyone got an idea how to arrange this with DHL, or will it just end up going back to sender, and i'll have to arrange a next shipment?

A bit impractical to just send them out in the middle of holiday season without asking if people can receive it

Ended up being lucky XD
i still had 1 day to get it from the post office when i returned from vacation XD.
(another kickstarter arrived one day earlier and wasn't so lucky)
What's up with kickstarters being like "you know what, after all this time, let's send it in the middle of holiday season XD".
The undelivered rates must be higher these months.

[jancumps]

FYI, Welectron in Germany have some meters in stock:
https://www.welectron.com/EEVBlog-121GW-Data-Logging-Multimeter-with-Bluetooth

I ordered one from Welectron.Good that there are suppliers wihin the European trade zone.
I received shipping notification yesterday.

[M4x]

FYI, Welectron in Germany have some meters in stock:
https://www.welectron.com/EEVBlog-121GW-Data-Logging-Multimeter-with-Bluetooth

I ordered one from Welectron.Good that there are suppliers wihin the European trade zone.
I received shipping notification yesterday.

Good for you! Mine is still in Frankfurt since August 1 "Arrived at transit facility"

[Dr. Frank]

OK,
here are my brief quantitative T.C. measurements:

50M: 9.965M @ 24.0°C; 10.503M @ 32.4°C    => 0.64 %/°C  (I = 20nA)
5M: 1.0003M @ 24.0°C; 1.0061M @ 32.3°C    => 0.07 %/°C  (I=200nA)
500k: 99.97k @ 25.6°C; 100.04k @ 32.2°C    => 0.01 %/°C  (I= 1.37 .. 2.2µA)

You know what?
No usual resistor technology, even at 10M, has a T.C. of 6400 ppm/K, you get these components easily at < 200ppm/K, in Thick Film technology.
And that change correlates with the range..

Therefore, this enormous change over temperature in 50M range, and 1/10 of that in 5M range, is caused by a (parasitic) change of the measuring current, that would be about 130pA/°C for both, and that makes sense, if it is a change of bias currents, like caused by the HY3131's amplifier, CMOS MUX, or protection circuitry.
If you select DCV, 50mV range, and leave the input open, you'll get a reading of about 1.700mV, which means that the HY3131 has about 170pA bias current..which also should be strongly dependent on temperature (not tested yet).. but it's in the same ballpark.

Don't know, if the BM869 is better in that aspect, but I assume so, because that uses a higher test current (100nA) in the 50M range.

The 10V DC range is practically not affected by this temperature change, maybe < 2 counts, which implies < 3 ppm/K

Frank

[IanB]

The most sensitive part appears to be the HY3131 itself.  I'm not making the part even warm enough to detect the change in temp with my fingers.   

Curious. If you use the meter to measure mains voltage on Low-Z for a few seconds and then try to take some readings on other ranges, some of those readings will be way off. Could it be that heat generated inside the meter case is responsible for some of this?

[PA0PBZ]

Are you maybe heating up a PTC? What range(s) and how long until things go back to normal?

[IanB]

To be clear, I do not have the current version of the meter.   Any idea how much "way off" is and what modes?  Could you post some details?

I posted about it before, and my observation was brushed off with comments like "expected", "PTC", and so on.

Are you maybe heating up a PTC? What range(s) and how long until things go back to normal?

The heat is what I'm curious about.

For me I think it was some of the resistance ranges, but someone else mentioned that temperature measurements might go crazy too.

Sorry, it's not easy for me to go back and check right now. Here's a post I made way ago about this:

How to make the meter show weird readings:

1. Put the meter on the Low Z range
2. Measure the mains voltage (let's say 120.0 V AC)
3. Put the meter on the Resistance range
4. Short the probes

You may observe an anomalous reading of about 3 ohms, slowly declining towards zero.

On other ranges (like DC mV) you may observe anomalous readings too.

It's as if a capacitor is charged up and is taking time to discharge back to zero volts. (Or did something heat up and it's taking time to cool down?)

[PeterL]

Just a note for Dave or David:
The review section on the products page for this meter is starting to attract spam. You might want to take measures to prevent these (I'm not sure if email confirmation is required), or do some moderation.

[Messtechniker]

Planning to join the esteemed EEVBlog 121GW owners club after the next hardware revision or in one years time or so, after the firmware has matured a bit. I really don't need a further DMM, but this voltnut thing seems to be biting me a bit.
Anyway: thanks to all helping to improve this DMM and iron out bugs and idiosyncrasies.

[jancumps]

FYI, Welectron in Germany have some meters in stock:
https://www.welectron.com/EEVBlog-121GW-Data-Logging-Multimeter-with-Bluetooth

I ordered one from Welectron.Good that there are suppliers wihin the European trade zone.
I received shipping notification yesterday.

Good for you! Mine is still in Frankfurt since August 1 "Arrived at transit facility"

It arrived today. I ordered the 4th from Welectron. I received a first notification from the post services on the 6th, arrived in Belgium the 8th and at my door the 10th.

[M4x]

I've ordered directly com Dave :-(

[firewalker]

How do you use the hook-n-loop fasteners of the case?

Alexander.

[IanB]

More importantly, why do this and other Brymen cases have the bizarre half-pouch in the cover? Anything you try to put in there immediately falls out when you open the lid. It would be much preferable to have a full size pouch in the lid with a velcro tab to keep it closed.

(In fact, one reason to consider having a sewing machine in a home workshop is to be able to make cases, covers and bags to your own specifications and not have to rely on finding something off the shelf that is never quite right.)

[firewalker]

There was a python script for communication with the meter under GNU/Linux? Do I remember correctly? Any links?

Alexander.