EEVblog 121GW Multimeter Issues

[Dr. Frank]

Hello,

I finally analysed the Ohm function.
The 500k and 5M ranges are especially unstable, sometimes giving an unreadable display.
But sometimes, also the 50 Ohm range is unstable on the last two digits, but may stabilize after some 'warmup'.

The 50M range sampling rate is a bit slower, and therefore delivers more stable readings, which makes sense for these high values.
Anyhow, it seems that in this range, there also exist some disturbance, which cause a virtually static shift of the reading (gain error).

I have digitized the probing current of the 121GW by means of my 34465A.
All currents from 470µA down to 20nA, were absolutely stable apart from some amplifier noise.

When I digitized the voltage across the D.U.T. resistor, I encountered a huge 50Hz signal, 100mVpp @ 700mV DC, which might be picked up by the mains environment (Germany = 50Hz). The 121GW seemed to be extremely sensitive to that 50Hz pickup, although its A/D should suppress the mains frequencies of either 50Hz, or 60Hz.

I have the suspicion, that the sampling rate especially in Ohms mode is not selected correctly, for 60Hz only, so that this would be stable in U.S. or other countries with 60Hz mains frequency, but unstable here in Europe.

Do other owners here in EU see that same effect? 

How can I check inside the guts of the 121GW, which mains frequencies are suppressed?

Frank

[Zucca]

Interesting,

Can you repeat the experiment with the R DUT in a closed box and by using coax BNC cables to the 121GW, surely you have some of those adapters:

[HKJ]

In my setup it is stable reading 100kOhm and 10Mohm, but jumps up to 1% up/down on 1Mohm:

[firewalker]

AC+DC Mode while measuring 230VAC Mains do not shows anything to me: autorange goes in loop?

 

True. But if you allow AC+DC to settle to zero and then measure, it works.

Do you know if in AC+DC mode it can display the AC or DC component in the second display?

Alexander.

[Dr. Frank]

Interesting,

Can you repeat the experiment with the R DUT in a closed box and by using coax BNC cables to the 121GW, surely you have some of those adapters:

I could, but I won't, for logical reasons.

It is crystal clear for me, that the 121GW is susceptible to 50Hz, as the variation is smaller (but well visible), when short cables are attached, but it goes wild, when I additionally connect longer cables towards the 34465A, and latter records that high 50Hz hum.


I also encounter the same behavior, and same magnitude, like HKJ, i.e 100k and 10M relatively stable, but 1M is jumping on the 1% scale.. my BM869 in comparison is rock solid in all these ranges.

Perhaps others might also confirm that finding...

[Zucca]

Got it, yes it's logical. The 50-60Hz will always be present picked up by the cables in a normal DMM curstomer user case.
With BNC we could reduce the problem but it's not the solution.

I looked up the frontend:
http://www.hycontek.com/wp-content/uploads/DS-HY3131_EN.pdf

It's not very easy to understand, I suppose you should play with the pre-filter to get rid of the 50Hz noise and generate a clean AD1FP and a AD1FN signals.
Between FTP and FTN there is a 27nF C21.

It's getting compicated.

Maybe a simple aluminum foil around the 121GW case? Meh... nevermind.

Do we have the BM869 schematics? A simple frontend comparison could help...

[Dr. Frank]

I don't remember the pre-production meter being all that unstable. 

Don't you have 60Hz mains, by chance?

[exe]

Could you please remind me how 50/60Hz filtering works? I thought there are only three ways of get rid of ripple: 1) synchronize measurements to mains frequency 2) set integration time (or whatever) equal to one period of sine. 3) low-pass filter, but it may make DMM unresponsive and slow to settle.

[Zucca]

Could you please remind me how 50/60Hz filtering works? I thought there are only three ways of get rid of ripple: 1) synchronize measurements to mains frequency 2) set integration time (or whatever) equal to one period of sine. 3) low-pass filter, but it may make DMM unresponsive and slow to settle.

1) Impossible on a battery device with no main input.
2) With 300ms integration time, you get 5@60Hz or 6@50Hz cycle. This could be a winner.
3) as you said, no.

[exe]

1) Impossible on a battery device with no main input.

It should be possible to sense it remotely, shouldn't it? Although, would eat internal space and increase power consumption...

[EEVblog]

Any chance of detecting lower than 12V on Low Z mode in the future firmware upgrade?

I recall asking this a long time back and they said maybe not, I don't recall the reason.
Will have to ask again.

[EEVblog]

Hello,

I finally analysed the Ohm function.
The 500k and 5M ranges are especially unstable, sometimes giving an unreadable display.
But sometimes, also the 50 Ohm range is unstable on the last two digits, but may stabilize after some 'warmup'.

The 50M range sampling rate is a bit slower, and therefore delivers more stable readings, which makes sense for these high values.
Anyhow, it seems that in this range, there also exist some disturbance, which cause a virtually static shift of the reading (gain error).

I have digitized the probing current of the 121GW by means of my 34465A.
All currents from 470µA down to 20nA, were absolutely stable apart from some amplifier noise.

When I digitized the voltage across the D.U.T. resistor, I encountered a huge 50Hz signal, 100mVpp @ 700mV DC, which might be picked up by the mains environment (Germany = 50Hz). The 121GW seemed to be extremely sensitive to that 50Hz pickup, although its A/D should suppress the mains frequencies of either 50Hz, or 60Hz.

I have the suspicion, that the sampling rate especially in Ohms mode is not selected correctly, for 60Hz only, so that this would be stable in U.S. or other countries with 60Hz mains frequency, but unstable here in Europe.

Do other owners here in EU see that same effect? 

How can I check inside the guts of the 121GW, which mains frequencies are suppressed?

Frank

I don't recall seeing that here (we use 50Hz too).
We'll have to do more testing to see if it can be replicated.

[firewalker]

The manual states:



But this is not the case. It will remember the last mode.

Alexander.

[dcac]

Hello,

I finally analysed the Ohm function.
The 500k and 5M ranges are especially unstable, sometimes giving an unreadable display.
But sometimes, also the 50 Ohm range is unstable on the last two digits, but may stabilize after some 'warmup'.

The 50M range sampling rate is a bit slower, and therefore delivers more stable readings, which makes sense for these high values.
Anyhow, it seems that in this range, there also exist some disturbance, which cause a virtually static shift of the reading (gain error).

I have digitized the probing current of the 121GW by means of my 34465A.
All currents from 470µA down to 20nA, were absolutely stable apart from some amplifier noise.

When I digitized the voltage across the D.U.T. resistor, I encountered a huge 50Hz signal, 100mVpp @ 700mV DC, which might be picked up by the mains environment (Germany = 50Hz). The 121GW seemed to be extremely sensitive to that 50Hz pickup, although its A/D should suppress the mains frequencies of either 50Hz, or 60Hz.

I have the suspicion, that the sampling rate especially in Ohms mode is not selected correctly, for 60Hz only, so that this would be stable in U.S. or other countries with 60Hz mains frequency, but unstable here in Europe.

Do other owners here in EU see that same effect? 

How can I check inside the guts of the 121GW, which mains frequencies are suppressed?

Frank

I don't recall seeing that here (we use 50Hz too).
We'll have to do more testing to see if it can be replicated.

I've noticed this behavior too, I'm in Sweden, 50Hz mains.

I compared with my U1252B which has 15 fps update rate. I put it next to the 121gw using equal length test leads laying in more or less the same path on the table measuring same valued 1% metal film resistors.

10M - seemed equally stable on both meters, although very very slow update rate on 121gw about 1-2 fps.

1M - U1252 was only flickering on 1 count - but 121gw was very unstable, recorded min = 0.9674, max = 1.0261 over about 10 sec time.

100K - seemed nice and stable on both meters.

51K - 121gw was unstable again, not nearly as much though as 1M and the reading was more like slowly crawling up and down, recorded min = 50.566, max 50.667.

[Dr. Frank]

As an example, here's one log of 1000 samples, 5M range, 1.9000 MOhm from the 5450A, short cables, 'noise' is 0.4%RMS or 1.3% PP.
That's not usable at all.

Frank

PS: I've verified, that the correct XTAL of 4.9152MHz is assembled, and also re-soldered these bad junctions J1 / J2 (wrecking my soldering tip by the lead-free stuff) - no change at all.
UEI manufacturing also burnt the edge of one coil a bit..   

[IanB]

Assuming you feel the meter is sensitive to something going on with the mains...

My meter appears to be quite sensitive to electrostatic fields. For example if I move my fingers near some of the buttons like Hold or Rel when in certain ranges (e.g. some resistance ranges) the display reading jumps around a lot.

I also find a similar effect with the test leads. Moving my hand near the test leads or moving the test leads around will also cause the reading to be unstable.

I previously mentioned that trying to measure a 10 M resistance is unsuccessful. Either the meter locks on the wrong reading, or when it does settle it reads ~9.6 M rather than 10 M.

Similarly, the lowest ohm range has a 0.000 resolution, but I cannot successfully make use of those digits. The reading is very unstable and will not settle reliably or repeatably.

I would have to say right now that the resistance ranges are compromised and not easy to trust.

[firewalker]

Are those issues firmware addressable?

Alexander.

[MadModder]

Firmware 1.22
I logged my mains voltage for 1 hour plus a few minutes. (MEM button) Logging interval 1 second.
I stopped the logging when the display said 4049 data points.

Issue 1: When reviewing the data on the instrument, it only steps from 1 to 60 and then back to 1 (maybe it should be like that, I don't remember), and the decimal point is in the wrong place. It says 2350.4V for example, not 235.04V.
Issue 2: When loading the data off of the SD-card, the data is only 3990 lines. Not 4049. They're all consecutive. And the decimal point is in the right place.
Issue 3: This is really weird. Look at the attached diagrams. I find the regularity in the right half highly unlikely. Exactly 60 second periods. I must log for longer to see what that looks like.

[DavidDLC]

Firmware 1.22
I logged my mains voltage for 1 hour plus a few minutes. (MEM button) Logging interval 1 second.
I stopped the logging when the display said 4049 data points.

Issue 1: When reviewing the data on the instrument, it only steps from 1 to 60 and then back to 1 (maybe it should be like that, I don't remember), and the decimal point is in the wrong place. It says 2350.4V for example, not 235.04V.
Issue 2: When loading the data off of the SD-card, the data is only 3990 lines. Not 4049. They're all consecutive. And the decimal point is in the right place.
Issue 3: This is really weird. Look at the attached diagrams. I find the regularity in the right half highly unlikely. Exactly 60 second periods. I must log for longer to see what that looks like.

I reported something similar, I mentioned that it may be that the reading on the background is handling the decimal point, when you are looking at your stored data.

David DLC

[MadModder]

Yes, I think you're right on that point. The decimal point moves around all over if selecting ohms range and pressing MEM before the auto sense has settled.

[fubar.gr]

Don't know if it is mentioned already, but I got mine today and I noticed some segments on the screen turning on as I peeled off the protective film.

It is quite reproducible, you can see the video here:

https://twitter.com/FubarGR/status/1022971753340039169

Has anyone else noticed that?

[Muttley Snickers]

The preproduction meter is also sensitive to the proximity of my hand, similar to that Gossen I looked at but not to that extent.

I commented a while back in another thread about the Keysight U1282A multimeter also acting in a similar manner with the readings dramatically fluctuating if the users hand was moving in close proximity to the meter.

It was mentioned elsewhere that these meters use the same chipset so perhaps it could be an inherent problem regardless of the filtering or front end configuration, I don't have a 121GW to compare against but perhaps somebody else does.

[mattselectronics]

Hello,
I have an issue with the power measurement.
The meter is still reading too high voltage values when it is in the VA range.
If I switch it to the voltage range, everything is as expected.
My Firmware Version is 1.22
Look at this Video at 5:40:
https://youtu.be/QpeivpCtZSk

[joeqsmith]

As an example, here's one log of 1000 samples, 5M range, 1.9000 MOhm from the 5450A, short cables, 'noise' is 0.4%RMS or 1.3% PP.
That's not usable at all.

Frank

PS: I've verified, that the correct XTAL of 4.9152MHz is assembled, and also re-soldered these bad junctions J1 / J2 (wrecking my soldering tip by the lead-free stuff) - no change at all.
UEI manufacturing also burnt the edge of one coil a bit..   

I attempted to repeat your test the best I could.  I am using a few Caddock TK series resistors to make up roughly 1.9M.   The two plots shown are my HP34401A (white) compared with the pre-production 121GW (red).   I have also included the CSV file off the meter (renamed to .txt to allow uploading).  The recording interval is 1 second for both meters.

Looking at the schematic, nothing pops out other than you should have a better reference.  A bit odd you don't see the problem in every range. 

Added your data set for completeness.

[exe]

Don't know if it is mentioned already, but I got mine today and I noticed some segments on the screen turning on as I peeled off the protective film.

I'm no expert, but this looks pretty normal to me. Static electricity powers up segments, I'd consider this normal. Probably, all LCDs are susceptible to some extent.