EEVblog 121GW Multimeter Issues

[hwj-d]

...
The SD data logging functionality is in the meter, is advertised, and is not working at the moment.

Hm, it is working, but with an issue, in the moment. This meter is rel fresh and need some community support, as it is defined to be. Your longtime logging is a little worst case scenario, even for meters like the Keithley DMM6500, imho. Maybe you need an other meter for that, or look inside the firmware by yourself (don't know it is open source - hackable too).
Wish good luck.   

[ggchab]

Working but with random problem ? What does it mean for a logging device 

The firmware is not open source. I would not expect too much "community support". Dave has a copy of the source code (with a non disclosure agreement) but will he ever use it? Only reading and understanding it could take weeks! ... And UEI does not seem to work hard on the 121GW firmware.

I hope I'm wrong 

[jchw4]

I do use two 121GW for few months, but recently started doing some extra data logging intensively (which is the reason I got this meter really), and unfortunately it is broken in many ways. I tested both, 1.57, and 2.00 fw.

• SERIOUS: After some time of data logging to SD card, it stops updating values in the file. LCD is showing proper values, and says that new records are added to the file. However the file only starts having 3 values cycling forever (flat line forever). This can start after 1000 records, or after 10000, or somewhere in between, hard to say. If I do a data logging for a 3 days, it is 100% certain my data file will be ruined, and experiment waste of time, and possibly money. The worst part is you only learn about it later, when you read files back on a computer. Example at the bottom.

I noticed this too, but decided that I probably had another problem!
Looking at my old file with 204179 entries, here is number of different values per every 10000 entries:
(first column = [record number/10000], second column = number of unique values among all records with the same  [record number/10000])

Code: [Select]

0 3959
1 219
2 54
3 44
4 54
5 43
6 28
7 11
8 11
9 11
10 11
11 11
12 11
13 11
14 11
15 11
16 11
17 11
18 11
19 11
20 11

If I reduce the divisor to 100, I get the same histogram size (last 10 entries):

Code: [Select]

...
2032 11
2033 11
2034 11
2035 11
2036 11
2037 11
2038 11
2039 11
2040 11
2041 11

Looking at my other files I also see that set of logged values becomes fixed earlier or later.

• CONSIDERABLE: It looks like a there is a limit of 600000 records, after which the logging automatically stops. Please rise it at least 100 times, as I don't see much reason to have it, and it is somehow limiting.

+1

• SERIOUS: When having Interval of 0, the sampling rate is very non uniform, sometimes doing 1 per second, sometimes 10 per second, sometimes stopping for some time. And because records don't have timestamp, it makes data logging at interval 0, totally useless for anything, there is no way to reliably use them for anything. This comes back to the 4th point above.

It seems that the meter likely blocks while writing to SD card. By turning logging quickly on and off several times I managed to block DMM completely for a few (about 10) seconds.
I even started thinking that it died, but then it unblocked on its own, so I am pretty sure writes to SD card are synchronous.

There are some other minor things that do bother me about the 121GW, but really, the data logging issues are the biggest problem right now, making it not usable at the moment.

I bought it specifically for logging too. ;(

[bicycleguy]

These are some templates I use with Libre Office (free office suite).  They make graphs of your data with time based on the start of the file and the interval.  For no interval it uses a guess of .19 seconds.  This is a plot of @Movax data

The 'How to save 121GW data to LibreOffice.txt' file I keep on the SD card to remind me how to do.
Note: First the .ods files need to be renamed to .ots  then opened in Libre Office and saved as templates.
The details on key strokes are Mac based, the windows or linux will be similar.

[movax]

...
The SD data logging functionality is in the meter, is advertised, and is not working at the moment.

Hm, it is working, but with an issue, in the moment. This meter is rel fresh and need some community support, as it is defined to be. Your longtime logging is a little worst case scenario, even for meters like the Keithley DMM6500, imho. Maybe you need an other meter for that, or look inside the firmware by yourself (don't know it is open source - hackable too).
Wish good luck.   

I don't think firmware is open source. I tried to find it, and debug, add some features, but can't find it anywhere.

[movax]

Just to visualize some issues in a graphic form, about my previous post.

[movax]

• SERIOUS: When having Interval of 0, the sampling rate is very non uniform, sometimes doing 1 per second, sometimes 10 per second, sometimes stopping for some time. And because records don't have timestamp, it makes data logging at interval 0, totally useless for anything, there is no way to reliably use them for anything. This comes back to the 4th point above.

It seems that the meter likely blocks while writing to SD card. By turning logging quickly on and off several times I managed to block DMM completely for a few (about 10) seconds.
I even started thinking that it died, but then it unblocked on its own, so I am pretty sure writes to SD card are synchronous.

I don't necessarily care why it is broken. If it doesn't work reliably, or have no timestamps to make it usable, it should be removed from options. I know it is likely due to SD library, and buffering, and possibly latency issues when communicating with SD library. It could (maybe) be fixed with double buffering (i.e. accumulating results to a secondary buffer, via interrupts, while the library is flushing buffers or FAT related stuff and waiting for sd to acknowledge writes. However, there is no point speculating why it is broken. The fact is, it is broken.

[Brumby]

...
The SD data logging functionality is in the meter, is advertised, and is not working at the moment.

Hm, it is working, but with an issue, in the moment. This meter is rel fresh and need some community support, as it is defined to be. Your longtime logging is a little worst case scenario, even for meters like the Keithley DMM6500, imho. Maybe you need an other meter for that, or look inside the firmware by yourself (don't know it is open source - hackable too).
Wish good luck.   

I don't think firmware is open source. I tried to find it, and debug, add some features, but can't find it anywhere.

No.  The firmware is definitely NOT open source.

Some enterprising souls took a copy of the binary and worked their way through it as object code, making some changes directly to the binary file.  This is difficult enough for code where the environment is familiar - but for a multimeter where hardware configuration is essentially an unknown, making any useful changes is quite the challenge and generally limited to minor tweaks.  Making wholesale changes or adding new features this way is pretty close to impossible.

The best avenue for change is to make your case here and see if it gets escalated up the chain.  Be warned though, don't expect changes to be forthcoming quickly.  The manufacturer will have other programming tasks and once someone gets onto it, there can be challenges in the coding work itself.  As part of this, a serious effort will be required for testing - and these steps invariably end up as a cycle which is worked through multiple times until you have a stable and functional piece of code.  As such, changes can sometimes take several weeks of development effort before new firmware is released.

[grzeartur]

Hello!
I've measured current consumption of mine 121GW. While it is turned on, it consumes ~5,5mA (voltage measurement). But after 30 minutes, when it goes sleep is still consumes 3mA(!). Can anyone of you measure current consumption is sleep mode? I wonder, whether mine multimeter is damaged or sleep mode is very poor in all of them. I have firmware 2.0.

[sirtet]

Darn, there are so many issues around data logging on this thing...
Was also one of my main reasons to get it.
But it's crazy to get to Data, screw open instead of BT File access.
No timestamps, instable on long measurements...
What's a better logging meter at around that price range? Or, how far up to get something better?
Any recommendations?

[hwj-d]

Darn, there are so many issues around data logging on this thing...
Was also one of my main reasons to get it.
But it's crazy to get to Data, screw open instead of BT File access.
No timestamps, instable on long measurements...
What's a better logging meter at around that price range? Or, how far up to get something better?
Any recommendations?

Hm, what do you expect from "this price range" on the market?

[CDaniel]

All the issues are happening because the multimeter was released very unfinished as "upgradable" ... but there is little incentive to work hard at something after you received the money and people still buy it , after all , this is no big surprise . And the task was probably largely underestimated , programming STM32 micro is not so easy even when you are a professional , supposedly

[sirtet]

Hm, what do you expect from "this price range" on the market?
[/quote]
Yes... i don't know the market well enough.
But someone quite going into the details like Dave, i expected every feature to be reasonably usable. SD under screws is... well, not thought through at all. Seems like a marketing gimmick, unless it was thought to be used with BT file access, which was maybe cut out due to no time/money to finish it, is what i guess.

after Dave multiple comments about his intent for this thread

Yes, i was unsure what this monster thread is meant for and what not.
My bad, Dave said it eg. on Page 1 comment 2.
And now i realize, i wanted to post on the general 121gw thread, but was too dumb to get that right...

[Brumby]

But someone quite going into the details like Dave, i expected every feature to be reasonably usable. SD under screws is... well, not thought through at all. Seems like a marketing gimmick, unless it was thought to be used with BT file access, which was maybe cut out due to no time/money to finish it, is what i guess.

Your criticism is based on ignorance.  Many of us have been following the development of the 121GW for quite some time.  The SD card addition came late in that development and a lot of the features which are now being exploited came along as "feature creep".  This alone means (in my book) criticism needs to be tempered.  I expect there will be some further firmware revisions that will address these issues, but they were not in the original design brief - so I would advise some patience.  The key functionality as a DMM is what is most important and that is becoming mature.

As for "SD under screws" - there is a very important safety reason for not trying to get around that.  Since the SD is electrically connected to the measuring circuitry, having easy access to the SD card would mean having easy access to the measuring circuitry and that would immediately negate your Cat ratings!!


BTW - These things have been discussed before.

[Muttley Snickers]

Your criticism is based on ignorance.  Many of us have been following the development of the 121GW for quite some time. (snip)...BTW - These things have been discussed before.

Many here would not be aware of the discussions which took place in the supporters lounge even prior to the first prototype, this includes the addition of the SD card.

[Brumby]

I did not mean "Ignorance" in the derogatory sense - just describing a "lack of knowledge".

Also, while early discussions on such subjects were limited to those with access to the Supporters area, they have been mentioned elsewhere since.  To be fair, though, I must concede that mentions (such as the late addition of the SD card) will be buried in broader discussion, but the issue of compromising Cat ratings is one of general knowledge.

However, this does rest on the presumption that people understand the basics of Cat rating ... and not everyone will have that knowledge when they enter a discussion.  To those people I say: Pay attention and understand what Cat ratings are about - it could save someone's life.

[CDaniel]

The fact is it doesn't have an easy way to see/download the data logging . This is a fair criticism . You should have pointed out clearly this deficiency at the "supporters area" early discusions 

[beanflying]

The fact is it doesn't have an easy way to see/download the data logging . This is a fair criticism . You should have pointed out clearly this deficiency at the "supporters area" early discusions 

Do you really go out of your way to be an ASSHAT?

Data being logged to an internal card at speed that is the same as on the display is not an 'issue' and you once again are being a TROLL !

[sirtet]

I wouldn't say CDaniel is trolling, but ok, to me his tone is rather tuning down than up.
Also, thanks @joeqsmith.

Well, i am an "outsider" who bought the meter based on how i judged Dave from seeing his DMM Review videos, and from the Kickstarter, which prominently says

A fully custom EEVblog Electronics Multimeter designed by Dave & UEI, with many exclusive features like SD card logging & Bluetooth!

This gave me the impression Datalogging is not a feature creep thing, but one of the more notable features. And as such, well-thought-out and executed.
With CAT rating, makes sense the SD is under the hood, vid did show that, but log-access only via physical makes it really a not-so great feature. also, logs without timestamp degrade the feature a bit more, as the manual AFAIK does not state a precision/torerance for the log intervals.
I did not dig much into what i can do better with the app, as in my case, it's not practically possible to have my phone or another device along with the meter.

Side-note: the CSV says 0 sec. interval instead of 0.2 when selecting 0... Reported that via Issue-link.

[CDaniel]

The fact is it doesn't have an easy way to see/download the data logging . This is a fair criticism . You should have pointed out clearly this deficiency at the "supporters area" early discusions 

Do you really go out of your way to be an ASSHAT?

Data being logged to an internal card at speed that is the same as on the display is not an 'issue' and you once again are being a TROLL !

I"m a troll for fanboys like you , the logging has issues not just the physical acces , a pain in the butt because they couldn"t do it wirelessly or add a normal computer interface . Whatever ...

[dr.diesel]

I'm not entering the war here, just commenting on my experience with the bluetooth logging...

I have tried to log my solar powered golf cart dozens of times and have not once had the logging last an entire day-light span.  Sometimes it will log for a few hours, sometimes less than 5 minutes before updates just stop, Android tablet on plug power 5' from the 121.

Note:  I have not tried with the 2.0 FW update, have since moved to a different meter.

[tautech]

I'm not entering the war here, just commenting on my experience with the bluetooth logging...

I have tried to log my solar powered golf cart dozens of times and have not once had the logging last an entire day-light span.  Sometimes it will log for a few hours, sometimes less than 5 minutes before updates just stop, Android tablet on plug power 5' from the 121.

Note:  I have not tried with the 2.0 FW update, have since moved to a different meter.

Time for a signature change ?

[beanflying]

The fact is it doesn't have an easy way to see/download the data logging . This is a fair criticism . You should have pointed out clearly this deficiency at the "supporters area" early discusions 

Do you really go out of your way to be an ASSHAT?

Data being logged to an internal card at speed that is the same as on the display is not an 'issue' and you once again are being a TROLL !

I"m a troll for fanboys like you , the logging has issues not just the physical acces , a pain in the butt because they couldn"t do it wirelessly or add a normal computer interface . Whatever ...

As you have spent maybe half of your posts TOTAL on this forum whining bitching and moaning about this meter, the manufacturer and the distributor you are an ASSHAT ! Stop being a serial SOOK

[dcac]

It’s too bad Fanboys and Trolls are such a major issue with the 121gw as their posting seems to drown the real hardware and firmware problems being reported.

[1anX]

It’s too bad Fanboys and Trolls are such a major issue with the 121gw as their posting seems to drown the real hardware and firmware problems being reported.

I agree 

A truly inspired meter design by Dave, but (IMHO) suffers from really poor implementation and support!
I too have moved on to other meters that do what I want in a manner as described in their manual and specification!