Author Topic: EEVblog 121GW Multimeter Issues  (Read 669935 times)

0 Members and 2 Guests are viewing this topic.

Offline IanB

  • Super Contributor
  • ***
  • Posts: 11790
  • Country: us
Re: EEVblog 121GW Multimeter Issues
« Reply #450 on: August 06, 2018, 11:40:40 pm »
This behavior to me is completely crazy. I don't quite understand the code behind the rules or what purpose they serve, but they are clearly there. I have no idea why they designed it like this.

What's worse, is the behavior appears and acts completely crazy. So much so, that I don't think I would be prepared to use my 121GW for resistance measurement. That function for me is effectively absent from the meter.

Resistance measurement should either be "slow and sure", or not at all.

I don't care a fig if the autoranging is slower than other meters as long as it actually works.

(Having re-read my post, I should clarify. It would of course be great if the auto-ranging for resistance were "fast and sure", but if it is a choice between "fast and unreliable" or "slow and sure", then I want slow and sure. Messing up the functionality in order to win some hypothetical speed race with other meters is not a good way to proceed.)
« Last Edit: August 07, 2018, 07:27:32 am by IanB »
 

Offline tpw_rules

  • Regular Contributor
  • *
  • Posts: 50
Re: EEVblog 121GW Multimeter Issues
« Reply #451 on: August 07, 2018, 01:02:21 am »
I am not suggesting that the original firmware was bug free by any means.  It makes me wonder if they (UEI) can't produce firmware for the meter, what's the future plans.  It's painfully obvious they are unable to test the code they release.   If you are going to charge a premium for it, it should at least perform as well as other meters in that price class and I don't see that happening today.

I'm about to start my own open source firmware project for all the bugs and idiosyncrasies I have pointed out. I have lots of very unkind words regarding the firmware design and implementation, which I will not repeat here because I don't believe them to be useful. Instead, I will aim to correct them in my own code and perhaps lead by example.

However! I am starting open source firmware only because I want to. I think it's unfair for me (and whoever will contribute in the future) to have to provide all of the features that are promised, and perhaps to influence future UEI meters. I think it's even more unfair for owners of the meter to have to rely on the free time of some bored college kid (and the free time of future contributors) to make their ~$400 instrument usable.

The other great loss of an open source firmware is that UEI has presumably acres of expensive testing, characterization, and calibration equipment they can use to make the meter the best it can be (whether they use it is another matter). An open source firmware would have to rely on the generous effort of people like joeqsmith and others in this thread who have access to similar equipment in order to make the guarantees that make the equipment worth its price. Will this be practical? I guess we will soon see.
 
The following users thanked this post: Andrew McNamara

Offline tsman

  • Frequent Contributor
  • **
  • Posts: 599
  • Country: gb
Re: EEVblog 121GW Multimeter Issues
« Reply #452 on: August 07, 2018, 03:17:47 am »
The card is a pain for logging.  I have left the meter apart since starting this little side project.  I don't have long enough finger nails to reach the card and my fingers are too large.  So needle nose pliers.
A small piece of tape on the card as a pull tab makes it so much easier to extract the card.

It makes me wonder if they (UEI) can't produce firmware for the meter, what's the future plans.  It's painfully obvious they are unable to test the code they release.
I wondering if it is because it is a STM32 inside and they're having issues with development + testing. The 121GW development video mentions that it originally a PIC with a choice of a MSP430 as well but UEI changed it for a STM32 later on.
« Last Edit: August 07, 2018, 03:29:35 am by tsman »
 

Offline Andrew McNamara

  • Regular Contributor
  • *
  • Posts: 76
  • Country: au
Re: EEVblog 121GW Multimeter Issues
« Reply #453 on: August 07, 2018, 06:17:14 am »
The other great loss of an open source firmware is that UEI has presumably acres of expensive testing, characterization, and calibration equipment they can use to make the meter the best it can be (whether they use it is another matter). An open source firmware would have to rely on the generous effort of people like joeqsmith and others in this thread who have access to similar equipment in order to make the guarantees that make the equipment worth its price. Will this be practical? I guess we will soon see.

UEI are seriously missing a trick here - Open Sourcing their code would make this meter into a unique platform, and the extra eyeballs would quickly solve these niggling firmware issues. They're likely worried about clones, but I doubt keeping the source secret will slow the cloners down much - they need to have more confidence in the value of their offering.
 

Offline IanJ

  • Supporter
  • ****
  • Posts: 1580
  • Country: scotland
  • Full time EE biz & Youtuber
    • IanJohnston.com
Re: EEVblog 121GW Multimeter Issues
« Reply #454 on: August 07, 2018, 06:32:43 am »
The other great loss of an open source firmware is that UEI has presumably acres of expensive testing, characterization, and calibration equipment they can use to make the meter the best it can be (whether they use it is another matter). An open source firmware would have to rely on the generous effort of people like joeqsmith and others in this thread who have access to similar equipment in order to make the guarantees that make the equipment worth its price. Will this be practical? I guess we will soon see.

UEI are seriously missing a trick here - Open Sourcing their code would make this meter into a unique platform, and the extra eyeballs would quickly solve these niggling firmware issues. They're likely worried about clones, but I doubt keeping the source secret will slow the cloners down much - they need to have more confidence in the value of their offering.

My opinion is that it's probably less about clones and more that UEI probably have quite a lot of routines and generic blocks of code in the meter that they use or want to use in forthcoming products and don't want to give those ideas away...…...I couldn't blame them for that. I think the fully open source idea would only have happened if it had been designed that way from the start......but then price point and timing would have been affected.

Ian.
Ian Johnston - Manufacturer of the PDVS2mini & author of the free WinGPIB app.
Website & Online Shop: www.ianjohnston.com
YT Channel (electronics repairs & projects): www.youtube.com/user/IanScottJohnston, Twitter (X): https://twitter.com/IanSJohnston
 

Offline Andrew McNamara

  • Regular Contributor
  • *
  • Posts: 76
  • Country: au
Re: EEVblog 121GW Multimeter Issues
« Reply #455 on: August 07, 2018, 07:01:38 am »
My opinion is that it's probably less about clones and more that UEI probably have quite a lot of routines and generic blocks of code in the meter that they use or want to use in forthcoming products and don't want to give those ideas away...…...I couldn't blame them for that. I think the fully open source idea would only have happened if it had been designed that way from the start......but then price point and timing would have been affected.

Maybe, although tpw_rules's analysis suggests coding is not their strength...  :(

My feeling is the UEI could benefit enormously from an Open Source code base for this meter - if they are unable to release their code for whatever reason, an alternative would be for them to actively participate in the development of an Open Source alternative code base, and to seriously consider using this as the shipping code when it's ready. At the very least, stating that they support the effort would win them some points.
 

Offline IanB

  • Super Contributor
  • ***
  • Posts: 11790
  • Country: us
Re: EEVblog 121GW Multimeter Issues
« Reply #456 on: August 07, 2018, 07:24:02 am »
I wondering if it is because it is a STM32 inside and they're having issues with development + testing. The 121GW development video mentions that it originally a PIC with a choice of a MSP430 as well but UEI changed it for a STM32 later on.

I doubt that it should be that. Software engineering is a generic skill, just like hardware engineering. Apparently many weaknesses in the code relate to structure and organization, which would point to poor design skills.
 

Offline nugglix

  • Regular Contributor
  • *
  • Posts: 209
  • Country: de
Re: EEVblog 121GW Multimeter Issues
« Reply #457 on: August 07, 2018, 07:37:01 am »
I doubt that it should be that. Software engineering is a generic skill, just like hardware engineering. Apparently many weaknesses in the code relate to structure and organization, which would point to poor design skills.

See  https://en.wikipedia.org/wiki/Conway%27s_law
 

Offline tpw_rules

  • Regular Contributor
  • *
  • Posts: 50
Re: EEVblog 121GW Multimeter Issues
« Reply #458 on: August 07, 2018, 03:09:12 pm »
Here's the behavior in v1.22. It's the same code as v1.02, except the "mode/range change delay" in ohms is now 0, i.e. the meter will perform an autorange on the next sample. This figure was 1 or 2 in v1.02, depending on range. To replace that, there is now an "autorange lock" function called right before the autorange change happens. Rule A: If not in ohms mode, the change is allowed. Rule B: If in ohms mode and changing to a lower range, deny changes for 1 second, then allow all changes. Rule C: If in ohms mode and changing to a higher range, deny the change, then allow one to happen in 1 sample. This leads to some bizarre behavior. If changing from a lower to a higher range, the meter will switch ranges quickly as expected. However, if changing from a higher to a lower range, Rule B above will force the meter to range down as quickly as possible, until either it's at the lowest range or a new ADC measurement arrives. Then it will use each sample to decide to range up, and eventually end up at the desired range. This behavior can easily be demonstrated by starting with open leads and connecting the meter to something like a 100k resistor.

Hot off the presses is the v1.25 release, which claims to fix this behavior, and in fact does. They've reinstated the mode/range change delay, but kept all the autorange lock stuff. There seems to be a lot more code related to the HOLD feature when in ohms mode and 50M range, maybe that fixes the "loop"? It's so difficult to understand.

They also tried to fix the noisy 5M range, not by changing the hardware configuration like this thread tried, but by cranking up the smoothing way past reasonable. As I mentioned before, the displayed value was the rolling average over the last 10 (I misunderstood and wrote 20) readings, and the averaging would reset if there was a delta of more than 20 counts. Now, it's the last 50, and the averaging resets with a delta of 300.
 

Offline Caelarius

  • Newbie
  • Posts: 7
  • Country: nl
Re: EEVblog 121GW Multimeter Issues
« Reply #459 on: August 07, 2018, 05:04:22 pm »
I've done a complete reverse engineering of the v1.02 firmware, which is how I know basically all the information I've posted in this thread.
how.

Ah, that was you? Awesome work that! I read the post at that time, but forgotten who did that in the meantime  :)

I pored over my disassembly for anything that could explain this bug but I don't see anything. The fact that it's writing bogus data but at a consistent interval is really weird. As you can see, someone else reported something similar. It's 100% a fixable software issue, except for maybe cosmic rays or damage to the EEPROM. Could you reprogram your meter with the v1.02 firmware and see if the issue is still there? I understand it a little more than v1.22. If you can't reproduce the issue, then I can look at what's changed between the versions for clues. Is there any chance whatsoever you own an ARM SWD CMSIS-DAP programmer?

There is a way to test if the EEPROM became write protected. Set up your experiment, start logging, wait however long, then hold the MEM button to turn off logging. Without turning off the meter, turn the dial to your favorite selection and note what mode is shown on the screen. Now, press the MODE key to change modes, and note what mode is now shown. Wait a few seconds, turn the dial to another selection, and then back to your favorite. If the mode shown is the one before you pressed the MODE key, the EEPROM is going into write protection somehow.

I'll give that a go later. I needed my meter for different purposes today, so haven't had the chance to go through some reproducing scenario's like I mentioned before. It's currently logging again. After that, I'll give firmware 1.25 and 1.02 a go to see how they behave. I'll try to scrounge up some time for that in the upcoming days.
 

Offline joeqsmith

  • Super Contributor
  • ***
  • Posts: 11630
  • Country: us
Re: EEVblog 121GW Multimeter Issues
« Reply #460 on: August 07, 2018, 05:10:16 pm »
Hot off the presses is the v1.25 release, which claims to fix this behavior, and in fact does.
I am not too surprised to hear this. 

I noticed I have cracked the SD card connector.  I am surprised none of the pins nave broke yet as many times as I have cycled the card in and out.   I always saw this as a mechanically weak point and thought they would do something to try and make it more robust in the final design but it appears to still be the same. 

Something to be aware of.  It's not Fluke tough and needs to be treated like a princess.     
 
The following users thanked this post: Marco1971

Offline Dr. Frank

  • Super Contributor
  • ***
  • Posts: 2377
  • Country: de
Re: EEVblog 121GW Multimeter Issues
« Reply #461 on: August 07, 2018, 05:47:30 pm »
Yep, FW 1-25 fixed several oddities of the Ohm mode.

The 5 MOhm is far more stable, and  it is not that susceptible to 50Hz mains influence any more.
Could do even better, if a capacitor is paralleled.
But I would call that a Fix.

Also, auto-ranging seems to work fine now, no false intermediate readings any more, and auto ranging speed seems to be the same as before.

Frank
 

Offline darik

  • Regular Contributor
  • *
  • Posts: 54
  • Country: us
Re: EEVblog 121GW Multimeter Issues
« Reply #462 on: August 07, 2018, 06:44:28 pm »
Also, auto-ranging seems to work fine now, no false intermediate readings any more, and auto ranging speed seems to be the same as before.

With 1.25 I am still consistently picking up a false intermediate reading in the 50 MOhm range when I short the probes, pretty much exactly as it did with 1.22.
 

Offline joeqsmith

  • Super Contributor
  • ***
  • Posts: 11630
  • Country: us
Re: EEVblog 121GW Multimeter Issues
« Reply #463 on: August 07, 2018, 07:09:00 pm »
Plots comparing the various versions of firmware.   Again, this is the same resistor, same location on the bench, same batteries. 

For 1.25, tried the tape eraser test.   I zoomed into the data and turned off everything except version 1.0, 1.22 and 1.25.   

I also added the autorange speed test to the previous video. 

https://youtu.be/q4h4ieNCN0I

Offline joeqsmith

  • Super Contributor
  • ***
  • Posts: 11630
  • Country: us
Re: EEVblog 121GW Multimeter Issues
« Reply #464 on: August 07, 2018, 08:03:26 pm »
Same data zoomed in a little further with thicker lines to show a better contrast between the data sets. 

It's strange that on every release the noise seems to get worse.  I can go back to an old version and it will calm back down so it does not appear to be anything with the test setup.   It appears to be very reproducible but perhaps something else is going on. 

Offline joeqsmith

  • Super Contributor
  • ***
  • Posts: 11630
  • Country: us
Re: EEVblog 121GW Multimeter Issues
« Reply #465 on: August 07, 2018, 09:45:04 pm »
Just for transparency sake and to provide you with some data on repeatability after running the sweep using the new 1.25 firmware I reinstalled  TPW_rules's modified version of 1.02.  Picture showing the test setup. Basically meter sitting on the bench with the CADDOCK TK resistor installed.   Nothing fancy.   This time, I did however at 2450 data points swept around the meter with the tape eraser like I did with 1.25.

The plots show the two data sets using the 1.02 firmware.  One from yesterday and the other just now.  The office temperature again is not constant.   The meter appears almost immune to the tape eraser.   

I included the latest firmware in blue for a reference.  I have to conclude that the firmware is worse than ever in regards to the noise level and sensitivity to the tape eraser.   If anyone feels that the test is invalid for some reason or has some suggestion for a better way to test it, let me know.

Offline DavidDLC

  • Frequent Contributor
  • **
  • Posts: 755
  • Country: us
Re: EEVblog 121GW Multimeter Issues
« Reply #466 on: August 08, 2018, 04:38:38 am »
Fw 1.25

1.- I also see intermediate values on the Ohm range.

2.- When looking at the mem values, the background measurements still handles the decimal point, very easy to reproduce:
- Record some values
- Set the meter to ohm
- View the recorded data
- Immediately short the probes

You will see the problem.


3.- The temperature measurement still goes crazy for some time while measuring the mains frequency


More coming

David DLC

 

Offline firewalker

  • Super Contributor
  • ***
  • Posts: 2450
  • Country: gr
Re: EEVblog 121GW Multimeter Issues
« Reply #467 on: August 08, 2018, 07:56:36 am »
The also doesn;t default to the "white mode" when turned off.

Alexander.
Become a realist, stay a dreamer.

 

Offline Dr. Frank

  • Super Contributor
  • ***
  • Posts: 2377
  • Country: de
Re: EEVblog 121GW Multimeter Issues
« Reply #468 on: August 08, 2018, 09:52:13 am »
DavidDLC and joeqsmith,

concerning the noise, are you referring to data logged to SD card?

The reading on the display is now stable, compared to 1-22, and as far I could determine.

Frank
 

Offline joeqsmith

  • Super Contributor
  • ***
  • Posts: 11630
  • Country: us
Re: EEVblog 121GW Multimeter Issues
« Reply #469 on: August 08, 2018, 04:23:44 pm »
DavidDLC and joeqsmith,

concerning the noise, are you referring to data logged to SD card?

The reading on the display is now stable, compared to 1-22, and as far I could determine.

Frank
In my case, I am always referring to the data collected on the SD card.  I have no desire to watch the LCD for 40 minutes and try to manually pick out any sort of trend.   

Again, keep in mind this is not the same hardware you are testing.  I have damaged the meter three times and have changed some of the hardware in an attempt to make it more robust.  I have also completely realigned the meter.   Maybe there is some other difference.   

I would expect with your setup you should easily replicate anything I show.   If you are seeing something different there must be something we have not accounted for.   

I did make a short video for you.   I start out with the modified 1.02 firmware.  I just let it run for a few minutes.  You can see the drift as the meter warms up.   Again, I assume this is caused by the Hycon chip they chose.    I always found it strange they used that part after Dave's review of the Keysight meter.    Anyway,  I then reprogrammed the latest firmware and let the meter run for 15 minutes or so to stabilize.   I then captured some video while it continued to run for several more minutes.   

One point of interest is that I understood  the kickstart to have roughly 2000 participants and that the vast majority have been delivered.  I assume very few people have firmware 1.0 installed.   Your recent post where you used the 1.9M was really brought my attention to this potential problem.  It does not seem to be common.  So, it's possible that people are just trusting the meter is fine.   Perhaps they really have no need for a higher class meter and bought it for other reasons.    Another possibility is they are not seeing the problem.   If it's the later, then why.   Also,  why would your meter all of a sudden behave different from mine.    Mine seems to reproduce based on the firmware but then again, as many tests as I have ran, we still don't have a lot of data showing how repeatable my results are.  We may just be chasing our tails right now.  It may be a case where we just have to wait for a next generation of the meter (assuming they go ahead with it) and I can have a look at the latest hardware. 

******************
Pulled the video for now.  I have ran a much larger test and put together a video explaining some of the data I have collected and will include this short clip as part of it.   

****  Updated with additional testing and description ****
https://youtu.be/RWqP0gM_Qf0
« Last Edit: August 08, 2018, 09:18:40 pm by joeqsmith »
 

Offline IanB

  • Super Contributor
  • ***
  • Posts: 11790
  • Country: us
Re: EEVblog 121GW Multimeter Issues
« Reply #470 on: August 08, 2018, 06:11:12 pm »
One point of interest is that I understood  the kickstart to have roughly 2000 participants and that the vast majority have been delivered.  I assume very few people have firmware 1.0 installed.   Your recent post where you used the 1.9M was really brought my attention to this potential problem.  It does not seem to be common.  So, it's possible that people are just trusting the meter is fine.

I bought the meter out of curiosity, and because it might have some useful features.

I have not had time to give it a comprehensive evaluation, but I have decided for now that I don't trust any of the resistance ranges.

So I for one don't think the meter is "fine". So far, it passes as a curiosity, but not as a working tool. For a working tool I trust my Brymen meters.
 

Offline darik

  • Regular Contributor
  • *
  • Posts: 54
  • Country: us
Re: EEVblog 121GW Multimeter Issues
« Reply #471 on: August 08, 2018, 06:56:28 pm »
One point of interest is that I understood  the kickstart to have roughly 2000 participants and that the vast majority have been delivered.  I assume very few people have firmware 1.0 installed.   Your recent post where you used the 1.9M was really brought my attention to this potential problem.  It does not seem to be common.  So, it's possible that people are just trusting the meter is fine.

I bought the meter out of curiosity, and because it might have some useful features.

I have not had time to give it a comprehensive evaluation, but I have decided for now that I don't trust any of the resistance ranges.

So I for one don't think the meter is "fine". So far, it passes as a curiosity, but not as a working tool. For a working tool I trust my Brymen meters.

I think most people haven't had the time and inclination to do this kind of evaluation, so who knows how common it is.

I got mine for the low burden and as a curiosity. I personally would have been shocked if it didn't have some problems, I'm a software engineer, I know what version 1.0 is like. I just hope the hardware design is sound enough that the roughest edges can be taken off with firmware updates.
 

Offline Dr. Frank

  • Super Contributor
  • ***
  • Posts: 2377
  • Country: de
Re: EEVblog 121GW Multimeter Issues
« Reply #472 on: August 09, 2018, 06:03:06 am »
In my case, I am always referring to the data collected on the SD card.  I have no desire to watch the LCD for 40 minutes and try to manually pick out any sort of trend.   

OK, and I was referring to the stability on the display, where you need a couple of seconds only to see, that it's now stable to the last digit
The SD card logging, though, contains completely other data, 50 times more noisy, and a different average value, see graph.
That's not usable, analysis from disassembling by tpw_rules explains the whole story.

Again, keep in mind this is not the same hardware you are testing.  I have damaged the meter three times and have changed some of the hardware in an attempt to make it more robust.  I have also completely realigned the meter.   Maybe there is some other difference.   

I would expect with your setup you should easily replicate anything I show.   If you are seeing something different there must be something we have not accounted for.   

Well, I don't think, that this makes a qualitative difference, as I in fact see the same behavior.

I did make a short video for you.   I start out with the modified 1.02 firmware.  I just let it run for a few minutes.  You can see the drift as the meter warms up.   Again, I assume this is caused by the Hycon chip they chose.    I always found it strange they used that part after Dave's review of the Keysight meter.    Anyway,  I then reprogrammed the latest firmware and let the meter run for 15 minutes or so to stabilize.   I then captured some video while it continued to run for several more minutes.   

Thanks a lot! Very instructive, and confirms that for recent firmware 1-25, the logged data is noisy, in contrast to the displayed data.

One point of interest is that I understood  the kickstart to have roughly 2000 participants and that the vast majority have been delivered.  I assume very few people have firmware 1.0 installed.   Your recent post where you used the 1.9M was really brought my attention to this potential problem.  It does not seem to be common.  So, it's possible that people are just trusting the meter is fine.   Perhaps they really have no need for a higher class meter and bought it for other reasons.    Another possibility is they are not seeing the problem.   If it's the later, then why.   

Well, only few people of this Kickstarter project have the means to check the 121GW in depth, and even less might go into detail, and just trust the manufacturer, with Dave as the initiator in the background.

Nevertheless, it's very positive, that several members of  the eevblog community take their time to make a deep analysis of the faults and traps of this eevblog DMM, and especially for me it's a great pleasure to cooperate with you and tpw_rules on this special problem.

Btw.: I signed up for this 121GW, because I wanted to have THE 'Dave-O-Meter', which is simply iconic.
Initially, I did not expect this instrument to be perfect.
So I also expected to have some fun in detecting errors in the instrument, and in the documentation.

Frank
« Last Edit: August 09, 2018, 06:22:52 am by Dr. Frank »
 

Offline dcac

  • Frequent Contributor
  • **
  • Posts: 336
Re: EEVblog 121GW Multimeter Issues
« Reply #473 on: August 09, 2018, 10:40:42 am »
I've contacted Hycontek about the HY3131 Configuration Application Notes (APD-DMM003) that's mentioned on their site, and they've kindly emailed it to me.

Maybe it has some useful information for 121gw-re project. I noticed it's not yet included in the pdfs folder so I thought I post it here (had to zip it as it was >1MB):




 
The following users thanked this post: Dr. Frank, exe

Online EEVblog

  • Administrator
  • *****
  • Posts: 37661
  • Country: au
    • EEVblog
Re: EEVblog 121GW Multimeter Issues
« Reply #474 on: August 09, 2018, 11:36:35 am »
I wondering if it is because it is a STM32 inside and they're having issues with development + testing. The 121GW development video mentions that it originally a PIC with a choice of a MSP430 as well but UEI changed it for a STM32 later on.

Correct. It was changed from a PIC to an STM32 at their request. I chose the final STM part used, but IIRC they said they hadn't done STM dev before that. We don't have visibility on their development environment and how they use it.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf