Author Topic: EEVBlog 121GW Discussion thread  (Read 770112 times)

0 Members and 3 Guests are viewing this topic.

Offline Iagash

  • Regular Contributor
  • *
  • Posts: 69
  • Country: de
Re: EEVBlog 121GW Discussion thread
« Reply #450 on: March 22, 2018, 04:03:35 pm »
If I change the input to the meter for negative current, the meter displays the negative current correctly, but the app still displays the same wrong positive current.

This is a bug in your app, Seppy. I get the correct sign value for the sub-display via bluetooth.

Longterm log is now running.
 

Offline Elasia

  • Frequent Contributor
  • **
  • Posts: 726
  • Country: us
Re: EEVBlog 121GW Discussion thread
« Reply #451 on: March 22, 2018, 07:05:28 pm »
Thx.

Do you know if there is any channel / information source that I can subscribe to see when multimeters will be available ?

Sign up for the newsletter
http://www.eevblog.com/newsletter/

Many people still ask, and I don't know why people seaming don't quite understand that I have to deliver the Kickstarter units first before I sell to the general public.

Simple... because its shiny!  .....  Squirrel!  That and they most likely don't comprehend your order book nor mass production and distribution of custom small batch electronics.
 

Offline Iagash

  • Regular Contributor
  • *
  • Posts: 69
  • Country: de
Re: EEVBlog 121GW Discussion thread
« Reply #452 on: March 22, 2018, 09:10:07 pm »
Ok, Version 1.10 is definitely a step forward but it still contains errors despite a correct checksum.

There is a pattern, when switching the second display from current to voltage in DC VA-mode the decimal point is wrong:
Code: [Select]
--
1521739280.909 2023-08 02342    Power DC (VA)  1966.3 mVA       DCA  205.12    0% f:  5%
1521739281.179 2023-08 02342    Power DC (VA)  1966.3 mVA       DCV  21.863    0% f:  5%
1521739281.517 2023-08 02342    Power DC (VA)  1966.3 mVA       DCV  2.1863    0% f:  5%
--
1521739349.017 2023-08 02342    Power DC (VA)  1967.2 mVA       DCA  204.99    0% f:  5%
1521739349.287 2023-08 02342    Power DC (VA)  1967.2 mVA       DCV  21.855    0% f:  5%
1521739349.557 2023-08 02342    Power DC (VA)  1967.2 mVA       DCV  2.1855    0% f:  5%
--
1521739615.442 2023-08 02342    Power DC (VA)  1966.9 mVA       DCA  204.99    0% f:  5%
1521739615.712 2023-08 02342    Power DC (VA)  1966.9 mVA       DCV  21.862    0% f:  5%
1521739615.914 2023-08 02342    Power DC (VA)  1967.1 mVA       DCV  2.1862    0% f:  5%
--

But sometimes it stays wrong:
Code: [Select]
--
1521741169.774 2023-08 02342    Power DC (VA)  1966.2 mVA       DCA  205.10    0% f:  5%
1521741170.044 2023-08 02342    Power DC (VA)  1966.3 mVA       DCV  21.866    0% f:  5%
1521741170.584 2023-08 02342    Power DC (VA)  1966.3 mVA       DCV  21.866    0% f:  5%
1521741170.854 2023-08 02342    Power DC (VA)  1966.3 mVA       DCV  21.866    0% f:  5%
1521741171.192 2023-08 02342    Power DC (VA)  1966.3 mVA       DCV  21.866    0% f:  5%
1521741171.327 2023-08 02342    Power DC (VA)  1966.3 mVA       DCV  21.866    0% f:  5%
1521741171.597 2023-08 02342    Power DC (VA)  1966.7 mVA       DCV  21.857    0% f:  5%
1521741171.799 2023-08 02342    Power DC (VA)  1966.7 mVA       DCV  21.857    0% f:  5%
1521741172.070 2023-08 02342    Power DC (VA)  1966.7 mVA       DCV  21.857    0% f:  5%
1521741172.272 2023-08 02342    Power DC (VA)  1966.7 mVA       DCV  21.857    0% f:  5%
1521741172.541 2023-08 02342    Power DC (VA)  1966.7 mVA       DCV  21.857    0% f:  5%
1521741172.880 2023-08 02342    Power DC (VA)  1966.7 mVA       DCV  21.857    0% f:  5%
1521741173.352 2023-08 02342    Power DC (VA)  1966.7 mVA       DCA  205.09    0% f:  5%
--

Sometimes I even get a wrong mode:
Code: [Select]
--
1521743502.116 2023-08 02342    Power DC (VA)  1968.8 mVA       DCA  204.66    0% f:  5%
1521743502.589 2000-00 00040 Voltage LowZ (V) -0.61975 V    48% f:  5%
1521743502.859 2023-08 02342    Power DC (VA)  1968.7 mVA       DCA  204.66    0% f:  5%
--

I collected arround 44k log records around 2k2 or 5% have a wrong checksum.

From the remaning 42k loglines with correct checksum 155 are broken, having either a wrong mode (66) a wrong decimal point(87) and 2 have some other parse error.
 

Offline HKJ

  • Super Contributor
  • ***
  • Posts: 2899
  • Country: dk
    • Tests
Re: EEVBlog 121GW Discussion thread
« Reply #453 on: March 22, 2018, 09:16:10 pm »
From the remaning 42k loglines with correct checksum 155 are broken

A XOR is not guaranteed to catch more than one byte errors, once in a while a multi byte error will slip through.
 

Offline Iagash

  • Regular Contributor
  • *
  • Posts: 69
  • Country: de
Re: EEVBlog 121GW Discussion thread
« Reply #454 on: March 22, 2018, 09:43:35 pm »
Another error:

The bargraph display in the app doesn't match the display on the meter.
I also get "strange" values via bluetooth, but the document about the BLE format doesn't describe how the values should be interpreted and the gitlab repo of the app is not uptodate.

I question the sending of the bargraph data via BLE at all, since the bargraph is meant for fast updates, which seems useless to me via BLE with 4-5 updates per second we see now.

I would prefer if the meter would send the values it has but doesn't display on the meter, like temperature and sending always current and voltage in VA mode.
But this is not a bug, more a question of preference and usage.
 

Offline Iagash

  • Regular Contributor
  • *
  • Posts: 69
  • Country: de
Re: EEVBlog 121GW Discussion thread
« Reply #455 on: March 22, 2018, 10:08:57 pm »
From the remaning 42k loglines with correct checksum 155 are broken

A XOR is not guaranteed to catch more than one byte errors, once in a while a multi byte error will slip through.

Yes of course, but I'm still very glad to be able to throw out the guessing of digits.
The remaining errors can be found quite easily by checking the log lines with grep and awk.

The goal would be getting flow-control right between the CPU and the BLE module, then even no checksum would be needed as FrankBuss pointed out.

The errors with the decimal point are probably another bug in the firmware that has nothing to do with the wrong checksums.
 

Offline Seppy

  • Supporter
  • ****
  • Posts: 189
  • Country: au
  • Curious
Re: EEVBlog 121GW Discussion thread
« Reply #456 on: March 23, 2018, 12:32:16 am »
Ok, Version 1.10 is definitely a step forward but it still contains errors despite a correct checksum.

There is a pattern, when switching the second display from current to voltage in DC VA-mode the decimal point is wrong:
Code: [Select]
--
1521739280.909 2023-08 02342    Power DC (VA)  1966.3 mVA       DCA  205.12    0% f:  5%
1521739281.179 2023-08 02342    Power DC (VA)  1966.3 mVA       DCV  21.863    0% f:  5%
1521739281.517 2023-08 02342    Power DC (VA)  1966.3 mVA       DCV  2.1863    0% f:  5%
--
1521739349.017 2023-08 02342    Power DC (VA)  1967.2 mVA       DCA  204.99    0% f:  5%
1521739349.287 2023-08 02342    Power DC (VA)  1967.2 mVA       DCV  21.855    0% f:  5%
1521739349.557 2023-08 02342    Power DC (VA)  1967.2 mVA       DCV  2.1855    0% f:  5%
--
1521739615.442 2023-08 02342    Power DC (VA)  1966.9 mVA       DCA  204.99    0% f:  5%
1521739615.712 2023-08 02342    Power DC (VA)  1966.9 mVA       DCV  21.862    0% f:  5%
1521739615.914 2023-08 02342    Power DC (VA)  1967.1 mVA       DCV  2.1862    0% f:  5%
--

But sometimes it stays wrong:
Code: [Select]
--
1521741169.774 2023-08 02342    Power DC (VA)  1966.2 mVA       DCA  205.10    0% f:  5%
1521741170.044 2023-08 02342    Power DC (VA)  1966.3 mVA       DCV  21.866    0% f:  5%
1521741170.584 2023-08 02342    Power DC (VA)  1966.3 mVA       DCV  21.866    0% f:  5%
1521741170.854 2023-08 02342    Power DC (VA)  1966.3 mVA       DCV  21.866    0% f:  5%
1521741171.192 2023-08 02342    Power DC (VA)  1966.3 mVA       DCV  21.866    0% f:  5%
1521741171.327 2023-08 02342    Power DC (VA)  1966.3 mVA       DCV  21.866    0% f:  5%
1521741171.597 2023-08 02342    Power DC (VA)  1966.7 mVA       DCV  21.857    0% f:  5%
1521741171.799 2023-08 02342    Power DC (VA)  1966.7 mVA       DCV  21.857    0% f:  5%
1521741172.070 2023-08 02342    Power DC (VA)  1966.7 mVA       DCV  21.857    0% f:  5%
1521741172.272 2023-08 02342    Power DC (VA)  1966.7 mVA       DCV  21.857    0% f:  5%
1521741172.541 2023-08 02342    Power DC (VA)  1966.7 mVA       DCV  21.857    0% f:  5%
1521741172.880 2023-08 02342    Power DC (VA)  1966.7 mVA       DCV  21.857    0% f:  5%
1521741173.352 2023-08 02342    Power DC (VA)  1966.7 mVA       DCA  205.09    0% f:  5%
--

Sometimes I even get a wrong mode:
Code: [Select]
--
1521743502.116 2023-08 02342    Power DC (VA)  1968.8 mVA       DCA  204.66    0% f:  5%
1521743502.589 2000-00 00040 Voltage LowZ (V) -0.61975 V    48% f:  5%
1521743502.859 2023-08 02342    Power DC (VA)  1968.7 mVA       DCA  204.66    0% f:  5%
--

I collected arround 44k log records around 2k2 or 5% have a wrong checksum.

From the remaning 42k loglines with correct checksum 155 are broken, having either a wrong mode (66) a wrong decimal point(87) and 2 have some other parse error.

It does appear that you have the wrong app, try the other app avaliable on the app store. This is not the format that my app logs, so yeah it must be the wrong app.

https://play.google.com/store/apps/details?id=eevblog.x121gw&hl=en

Update 1:
I have reproduced the voltage reading error in VA mode, am investigating whether the packet is the cause or the app, will update this post soon.

Update 2:
There is an issue with the voltage display (in the app on the sub display) in the VA mode, it appears to behave like a signed/unsigned issue with the formation of the packet, this means that for voltages above (32768 - 1 ish +- the calibrated offset) the value appears to result in invalid readings. It is possible still that the App has caused the issue but it is looking quite unlikely.
This will be resolved as soon as possible.

Note: The decimal position for the mVA appears also to be reported incorrectly in the packet (only those modes) these issues will also be dealt with soon.
« Last Edit: March 23, 2018, 06:32:47 am by Seppy »
 

Offline Iagash

  • Regular Contributor
  • *
  • Posts: 69
  • Country: de
Re: EEVBlog 121GW Discussion thread
« Reply #457 on: March 23, 2018, 12:41:25 pm »

It does appear that you have the wrong app, try the other app avaliable on the app store. This is not the format that my app logs, so yeah it must be the wrong app.

https://play.google.com/store/apps/details?id=eevblog.x121gw&hl=en

Yes Seppy, it's the wrong app, I'm sorry. The output is from my BLE linux script which I adapted to the new format according to the BLE documentation.

I can't do extended logging with your app, since I have only one BLE capable Android device and no BLE capable Windows and not even a Windows with a Microsoft store. The Android phone is my phone and has to follow me around. So I can't leave it untouched next to the meter for long. I just use it to compare the values displayed by the app to the output of my script.

I also had the problem with your android app, that I couldn't save the log when I left it running once over night. When pressing save the app just froze for some seconds and then continued. Maybe my phone has too little memory.
 

Offline Iagash

  • Regular Contributor
  • *
  • Posts: 69
  • Country: de
Re: EEVBlog 121GW Discussion thread
« Reply #458 on: March 24, 2018, 10:50:11 am »
I added plausibility checks back into my script. I'm checking the year and month the meter is transmitting, since it seems to be always the same.
Then I added printing the data of invalid records. The log below shows the invalid data between two correct records. As you can see, they don't happen very often.

Maybe you can check, if you can also see this pattern with your app. The last column shows the percentage of records with invalid xor checksum. So around 5-6% of the records are broken. I have not yet analyzed these records.

Code: [Select]
20180324-10:07:40.176 2023-08 02342    Power DC (VA)  2005.9 mVA       DCV  2.0987    0% f:  5%
20180324-10:07:40.648 error: unparsable record: Invalid year (0)
20180324-10:07:40.648 error: data: f2 00 00 01 40 00 d5 f2 17 80 23 42 18 00 4e 60 01 04 51
20180324-10:07:40.783 error: skiped 7 input bytes
20180324-10:07:40.783 2023-08 02342    Power DC (VA)  2006.4 mVA       DCV  2.0978    0% f:  5%
--
20180324-10:08:42.006 2023-08 02342    Power DC (VA)  2006.2 mVA       DCV  2.0978    0% f:  5%
20180324-10:08:42.478 error: unparsable record: Invalid year (0)
20180324-10:08:42.478 error: data: f2 00 00 01 40 00 eb f2 17 80 23 42 18 00 4e 5e 01 04 51
20180324-10:08:42.614 error: skiped 7 input bytes
20180324-10:08:42.614 2023-08 02342    Power DC (VA)  2006.2 mVA       DCV  2.0978    0% f:  6%
--
20180324-10:09:19.199 2023-08 02342    Power DC (VA)  2006.2 mVA       DCV  2.0978    0% f:  5%
20180324-10:09:19.671 error: unparsable record: Invalid year (0)
20180324-10:09:19.671 error: data: f2 00 00 01 40 00 eb f2 17 80 23 42 18 00 4e 5e 01 04 51
20180324-10:09:19.806 error: skiped 7 input bytes
20180324-10:09:19.806 2023-08 02342    Power DC (VA)  2006.2 mVA       DCV  2.0978    0% f:  5%
--
20180324-10:11:36.427 2023-08 02342    Power DC (VA)  2006.8 mVA       DCV  2.0978    0% f:  6%
20180324-10:11:36.900 error: unparsable record: Invalid year (0)
20180324-10:11:36.900 error: data: f2 00 00 01 40 00 d1 f2 17 80 23 42 18 00 4e 64 01 04 51
20180324-10:11:37.035 error: skiped 7 input bytes
20180324-10:11:37.035 2023-08 02342    Power DC (VA)  2006.8 mVA       DCV  2.0978    0% f:  6%
--
20180324-10:12:51.825 2023-08 02342    Power DC (VA)  2006.6 mVA       DCV  2.0978    0% f:  6%
20180324-10:12:52.297 error: unparsable record: Invalid year (0)
20180324-10:12:52.297 error: data: f2 00 00 01 40 00 d7 f2 17 80 23 42 18 00 4e 62 01 04 51
20180324-10:12:52.433 error: skiped 7 input bytes
20180324-10:12:52.433 2023-08 02342    Power DC (VA)  2006.6 mVA       DCV  2.0978    0% f:  6%
--
20180324-10:16:26.477 2023-08 02342    Power DC (VA)  2006.6 mVA       DCV  2.0978    0% f:  6%
20180324-10:16:26.949 error: unparsable record: Invalid year (0)
20180324-10:16:26.949 error: data: f2 00 00 01 40 00 d7 f2 17 80 23 42 18 00 4e 62 01 04 51
20180324-10:16:27.084 error: skiped 7 input bytes
20180324-10:16:27.084 2023-08 02342    Power DC (VA)  2006.6 mVA       DCV  2.0978    0% f:  6%
--
20180324-10:18:36.415 2023-08 02342    Power DC (VA)  2006.8 mVA       DCV  2.0978    0% f:  5%
20180324-10:18:36.887 error: unparsable record: Invalid year (0)
20180324-10:18:36.887 error: data: f2 00 00 01 40 00 d1 f2 17 80 23 42 18 00 4e 64 01 04 51
20180324-10:18:37.022 error: skiped 7 input bytes
20180324-10:18:37.022 2023-08 02342    Power DC (VA)  2006.8 mVA       DCV  2.0978    0% f:  6%
--
20180324-10:37:36.767 2023-08 02342    Power DC (VA)  2006.8 mVA       DCV  2.0978    0% f:  5%
20180324-10:37:37.239 error: unparsable record: Invalid year (0)
20180324-10:37:37.239 error: data: f2 00 00 01 40 00 d1 f2 17 80 23 42 18 00 4e 64 01 04 51
20180324-10:37:37.374 error: skiped 7 input bytes
20180324-10:37:37.374 2023-08 02342    Power DC (VA)  2006.8 mVA       DCV  2.0978    0% f:  5%
--
20180324-10:40:23.155 2023-08 02342    Power DC (VA)  2006.8 mVA       DCV  2.0978    0% f:  6%
20180324-10:40:23.628 error: unparsable record: Invalid year (0)
20180324-10:40:23.628 error: data: f2 00 00 01 40 00 d1 f2 17 80 23 42 18 00 4e 64 01 04 51
20180324-10:40:23.763 error: skiped 7 input bytes
20180324-10:40:23.763 2023-08 02342    Power DC (VA)  2006.8 mVA       DCV  2.0979    0% f:  6%
--
20180324-10:41:13.916 2023-08 02342    Power DC (VA)  2006.8 mVA       DCV  2.0978    0% f:  6%
20180324-10:41:14.388 error: unparsable record: Invalid year (0)
20180324-10:41:14.388 error: data: f2 00 00 01 40 00 d1 f2 17 80 23 42 18 00 4e 64 01 04 51
20180324-10:41:14.523 error: skiped 7 input bytes
20180324-10:41:14.523 2023-08 02342    Power DC (VA)  2006.8 mVA       DCV  2.0978    0% f:  6%
--
20180324-10:45:01.864 2023-08 02342    Power DC (VA)  2007.1 mVA       DCV  2.0978    0% f:  6%
20180324-10:45:02.337 error: unparsable record: Invalid year (0)
20180324-10:45:02.337 error: data: f2 00 00 01 40 00 d2 f2 17 80 23 42 18 00 4e 67 01 04 51
20180324-10:45:02.472 error: skiped 7 input bytes
20180324-10:45:02.472 2023-08 02342    Power DC (VA)  2007.1 mVA       DCV  2.0978    0% f:  6%
--
20180324-10:48:21.463 2023-08 02342    Power DC (VA)  2006.8 mVA       DCV  2.0978    0% f:  6%
20180324-10:48:21.935 error: unparsable record: Invalid year (0)
20180324-10:48:21.935 error: data: f2 00 00 01 40 00 d1 f2 17 80 23 42 18 00 4e 64 01 04 51
20180324-10:48:22.071 error: skiped 7 input bytes
20180324-10:48:22.071 2023-08 02342    Power DC (VA)  2006.8 mVA       DCV  2.0978    0% f:  6%
--
20180324-10:51:01.102 2023-08 02342    Power DC (VA)  2006.8 mVA       DCV  2.0978    0% f:  6%
20180324-10:51:01.574 error: unparsable record: Invalid year (0)
20180324-10:51:01.574 error: data: f2 00 00 01 40 00 d1 f2 17 80 23 42 18 00 4e 64 01 04 51
20180324-10:51:01.709 error: skiped 7 input bytes
20180324-10:51:01.709 2023-08 02342    Power DC (VA)  2006.8 mVA       DCV  2.0978    0% f:  6%
--
20180324-10:52:48.427 2023-08 02342    Power DC (VA)  2008.6 mVA       DCA  204.66    0% f:  6%
20180324-10:52:48.900 error: unparsable record: Invalid year (0)
20180324-10:52:48.900 error: data: f2 00 00 01 40 00 cf f2 17 80 23 42 18 00 4e 76 15 02 4f
20180324-10:52:49.035 error: skiped 7 input bytes
20180324-10:52:49.035 2023-08 02342    Power DC (VA)  2008.6 mVA       DCA  204.66    0% f:  6%
--
20180324-11:02:35.006 2023-08 02342    Power DC (VA)  2006.8 mVA       DCV  2.0980    0% f:  5%
20180324-11:02:35.478 error: unparsable record: Invalid year (0)
20180324-11:02:35.478 error: data: f2 00 00 01 40 00 d0 f2 17 80 23 42 18 00 4e 65 01 04 51
20180324-11:02:35.613 error: skiped 7 input bytes
20180324-11:02:35.613 2023-08 02342    Power DC (VA)  2006.9 mVA       DCV  2.0978    0% f:  5%
--
20180324-11:03:35.824 2023-08 02342    Power DC (VA)  2006.4 mVA       DCV  2.0978    0% f:  5%
20180324-11:03:36.296 error: unparsable record: Invalid year (0)
20180324-11:03:36.296 error: data: f2 00 00 01 40 00 d5 f2 17 80 23 42 18 00 4e 60 01 04 51
20180324-11:03:36.431 error: skiped 7 input bytes
20180324-11:03:36.431 2023-08 02342    Power DC (VA)  2006.4 mVA       DCV  2.0978    0% f:  6%
--
20180324-11:14:06.817 2023-08 02342    Power DC (VA)  2006.8 mVA       DCV  2.0978    0% f:  6%
20180324-11:14:07.290 error: unparsable record: Invalid year (0)
20180324-11:14:07.290 error: data: f2 00 00 01 40 00 d1 f2 17 80 23 42 18 00 4e 64 01 04 51
20180324-11:14:07.425 error: skiped 7 input bytes
20180324-11:14:07.425 2023-08 02342    Power DC (VA)  2006.8 mVA       DCV  2.0978    0% f:  6%
--
20180324-11:21:51.963 2023-08 02342    Power DC (VA)  2007.0 mVA       DCV  2.0978    0% f:  6%
20180324-11:21:52.435 error: unparsable record: Invalid year (0)
20180324-11:21:52.435 error: data: f2 00 00 01 40 00 d3 f2 17 80 23 42 18 00 4e 66 01 04 51
20180324-11:21:52.570 error: skiped 7 input bytes
20180324-11:21:52.570 2023-08 02342    Power DC (VA)  2007.0 mVA       DCV  2.0978    0% f:  6%
--
20180324-11:33:36.262 2023-08 02342    Power DC (VA)  2007.8 mVA       DCA  204.75    0% f:  6%
20180324-11:33:36.734 error: unparsable record: Invalid year (0)
20180324-11:33:36.734 error: data: f2 00 00 01 40 00 db f2 17 80 23 42 18 00 4e 6e 01 04 51
20180324-11:33:36.870 error: skiped 7 input bytes
20180324-11:33:36.870 2023-08 02342    Power DC (VA)  2007.8 mVA       DCV  2.0978    0% f:  6%
--
20180324-11:36:30.886 2023-08 02342    Power DC (VA)  2007.2 mVA       DCV  2.0978    0% f:  6%
20180324-11:36:31.358 error: unparsable record: Invalid year (0)
20180324-11:36:31.358 error: data: f2 00 00 01 40 00 dd f2 17 80 23 42 18 00 4e 68 01 04 51
20180324-11:36:31.493 error: skiped 7 input bytes
20180324-11:36:31.493 2023-08 02342    Power DC (VA)  2007.2 mVA       DCV  2.0978    0% f:  6%
 

Offline Iagash

  • Regular Contributor
  • *
  • Posts: 69
  • Country: de
Re: EEVBlog 121GW Discussion thread
« Reply #459 on: March 25, 2018, 07:38:48 pm »
Here are some records with failing xor checksums:

Code: [Select]
20180325-21:31:47.828 error: invalid checksum: f2 17 80 23 42 18 00 25 45 01 04 70 1d 00 00 04 40 80 50
20180325-21:31:52.351 error: invalid checksum: f2 17 80 23 42 18 00 25 49 01 04 70 25 00 00 04 08 00 64
20180325-21:32:01.463 error: invalid checksum: f2 17 80 23 42 18 00 25 45 15 02 fb 1f 00 00 04 40 00 4e
20180325-21:32:06.053 error: invalid checksum: f2 17 80 23 42 06 00 25 46 01 04 70 20 00 00 04 40 00 6e
20180325-21:32:07.741 error: invalid checksum: f2 17 80 23 42 18 00 25 40 21 02 7e 22 00 00 04 40 00 76
20180325-21:32:11.521 error: invalid checksum: f2 17 80 23 42 18 00 25 47 01 04 70 01 00 00 04 40 00 6c
20180325-21:32:20.566 error: invalid checksum: f2 17 80 23 42 18 00 25 49 15 02 7e 07 00 00 04 40 00 43
20180325-21:32:25.089 error: invalid checksum: f2 17 80 23 42 18 40 25 43 01 04 70 16 00 00 04 40 00 5d
20180325-21:32:26.844 error: invalid checksum: f2 17 80 23 42 18 00 25 a9 15 02 7e 21 00 00 04 40 00 0a
20180325-21:32:29.611 error: invalid checksum: f2 17 80 88 42 18 00 25 42 01 04 70 0c 00 00 04 40 00 46
20180325-21:32:31.366 error: invalid checksum: f2 17 80 23 42 18 00 22 42 01 04 70 0c 00 00 04 40 00 46
20180325-21:32:35.889 error: invalid checksum: f2 17 80 23 42 0c 00 25 44 01 04 70 14 00 00 04 40 00 58
20180325-21:32:48.714 error: invalid checksum: f2 17 80 23 42 18 00 25 40 01 00 70 0b 00 00 04 40 00 43
20180325-21:32:53.236 error: invalid checksum: f2 17 80 23 21 18 00 25 3e 15 02 7e 1d 00 00 04 40 00 37
20180325-21:32:54.991 error: invalid checksum: f2 17 80 23 42 18 00 25 09 01 04 70 01 00 00 04 40 00 37
20180325-21:32:57.759 error: invalid checksum: f2 17 80 23 42 01 00 25 3d 15 02 7e 20 00 00 04 40 00 09
20180325-21:32:59.446 error: invalid checksum: f2 17 80 23 10 18 00 25 3d 15 02 7e 20 00 00 04 40 00 09
20180325-21:33:09.571 error: invalid checksum: f2 17 80 23 42 18 00 25 45 15 02 7e 07 00 00 04 40 00 4e
20180325-21:33:12.339 error: invalid checksum: f2 17 80 23 42 01 00 25 43 01 04 70 16 00 00 04 40 00 5d
20180325-21:33:14.161 error: invalid checksum: f2 17 80 23 42 18 00 25 43 20 04 70 16 00 00 04 40 00 5d

I don't really understand what's happening here. It looks a bit like random bit errors. could it be a birate mismatch between CPU and BLE module?

Maybe someone with more experience with this kind of errors can comment on this.
 

Offline Seppy

  • Supporter
  • ****
  • Posts: 189
  • Country: au
  • Curious
Re: EEVBlog 121GW Discussion thread
« Reply #460 on: March 26, 2018, 12:40:36 am »

It does appear that you have the wrong app, try the other app avaliable on the app store. This is not the format that my app logs, so yeah it must be the wrong app.

https://play.google.com/store/apps/details?id=eevblog.x121gw&hl=en

I also had the problem with your android app, that I couldn't save the log when I left it running once over night. When pressing save the app just froze for some seconds and then continued. Maybe my phone has too little memory.

I'll look into this, not sure what caused it but i'll do my best to reproduce, it may take a while to reproduce these type of issues can sometimes take a while to figure out the exact mechanism as its likely something specific about the android API. I'll get back to you or just release a fix as an update if it is reproducible.
 

Offline Iagash

  • Regular Contributor
  • *
  • Posts: 69
  • Country: de
Re: EEVBlog 121GW Discussion thread
« Reply #461 on: March 26, 2018, 11:52:35 am »
I'll look into this, not sure what caused it but i'll do my best to reproduce, it may take a while to reproduce these type of issues can sometimes take a while to figure out the exact mechanism as its likely something specific about the android API. I'll get back to you or just release a fix as an update if it is reproducible.

Thanks, but don't spend much time on this. Short logs are working fine. The fault could also be my phone I just tried it once.

I'd be much more interested if you can see the same data errors that I see with my linux script.

Maybe you could add logging of these failed records raw data to your app, so that I can compare these?
 

Offline FrankBuss

  • Supporter
  • ****
  • Posts: 2365
  • Country: de
    • Frank Buss
Re: EEVBlog 121GW Discussion thread
« Reply #462 on: March 26, 2018, 06:16:13 pm »
Here are some records with failing xor checksums:

Code: [Select]
20180325-21:31:47.828 error: invalid checksum: f2 17 80 23 42 18 00 25 45 01 04 70 1d 00 00 04 40 80 50
20180325-21:31:52.351 error: invalid checksum: f2 17 80 23 42 18 00 25 49 01 04 70 25 00 00 04 08 00 64
20180325-21:32:01.463 error: invalid checksum: f2 17 80 23 42 18 00 25 45 15 02 fb 1f 00 00 04 40 00 4e
20180325-21:32:06.053 error: invalid checksum: f2 17 80 23 42 06 00 25 46 01 04 70 20 00 00 04 40 00 6e
20180325-21:32:07.741 error: invalid checksum: f2 17 80 23 42 18 00 25 40 21 02 7e 22 00 00 04 40 00 76
20180325-21:32:11.521 error: invalid checksum: f2 17 80 23 42 18 00 25 47 01 04 70 01 00 00 04 40 00 6c
20180325-21:32:20.566 error: invalid checksum: f2 17 80 23 42 18 00 25 49 15 02 7e 07 00 00 04 40 00 43
20180325-21:32:25.089 error: invalid checksum: f2 17 80 23 42 18 40 25 43 01 04 70 16 00 00 04 40 00 5d
20180325-21:32:26.844 error: invalid checksum: f2 17 80 23 42 18 00 25 a9 15 02 7e 21 00 00 04 40 00 0a
20180325-21:32:29.611 error: invalid checksum: f2 17 80 88 42 18 00 25 42 01 04 70 0c 00 00 04 40 00 46
20180325-21:32:31.366 error: invalid checksum: f2 17 80 23 42 18 00 22 42 01 04 70 0c 00 00 04 40 00 46
20180325-21:32:35.889 error: invalid checksum: f2 17 80 23 42 0c 00 25 44 01 04 70 14 00 00 04 40 00 58
20180325-21:32:48.714 error: invalid checksum: f2 17 80 23 42 18 00 25 40 01 00 70 0b 00 00 04 40 00 43
20180325-21:32:53.236 error: invalid checksum: f2 17 80 23 21 18 00 25 3e 15 02 7e 1d 00 00 04 40 00 37
20180325-21:32:54.991 error: invalid checksum: f2 17 80 23 42 18 00 25 09 01 04 70 01 00 00 04 40 00 37
20180325-21:32:57.759 error: invalid checksum: f2 17 80 23 42 01 00 25 3d 15 02 7e 20 00 00 04 40 00 09
20180325-21:32:59.446 error: invalid checksum: f2 17 80 23 10 18 00 25 3d 15 02 7e 20 00 00 04 40 00 09
20180325-21:33:09.571 error: invalid checksum: f2 17 80 23 42 18 00 25 45 15 02 7e 07 00 00 04 40 00 4e
20180325-21:33:12.339 error: invalid checksum: f2 17 80 23 42 01 00 25 43 01 04 70 16 00 00 04 40 00 5d
20180325-21:33:14.161 error: invalid checksum: f2 17 80 23 42 18 00 25 43 20 04 70 16 00 00 04 40 00 5d

I don't really understand what's happening here. It looks a bit like random bit errors. could it be a birate mismatch between CPU and BLE module?

Maybe someone with more experience with this kind of errors can comment on this.

This looks like a baudrate problem, or maybe wrong start bit detection or something like this. E.g. when 0x42 turns into 0x21, one bit shifted to the right. Same for 0x18 turning int 0x0c, one bit shifted.

Very unlikely that it has anything to do with the RF transfer from the Bluetooth module to the receiver, because as I noted earlier, BLE has a sophisticated error correction. Not just xor, but CRC, which is easy to calculate with tables, I've implemented this once myself for another product. You can verify this if you increase the distance: there shouldn't be more errors, but it gets slower, because packets gets resend, at least when I tested it with my Blue Gecko Silabs IC, which had no packet errors at all at the application level.

Maybe you can measure the RX/TX lines between the Bluetooth module and the CPU to check if the baudrate is exact in your multimeter, if David can't reproduce it?
So Long, and Thanks for All the Fish
Electronics, hiking, retro-computing, electronic music etc.: https://www.youtube.com/c/FrankBussProgrammer
 

Offline plexusTopic starter

  • Contributor
  • Posts: 41
Re: EEVBlog 121GW Discussion thread
« Reply #463 on: March 26, 2018, 07:37:03 pm »
I would love an iphone app for the 121GW that is better than the existing one in the app store. Better GUI and user experience!
 

Offline Seppy

  • Supporter
  • ****
  • Posts: 189
  • Country: au
  • Curious
Re: EEVBlog 121GW Discussion thread
« Reply #464 on: March 27, 2018, 05:40:02 am »
Here are some records with failing xor checksums:

Code: [Select]
20180325-21:31:47.828 error: invalid checksum: f2 17 80 23 42 18 00 25 45 01 04 70 1d 00 00 04 40 80 50
20180325-21:31:52.351 error: invalid checksum: f2 17 80 23 42 18 00 25 49 01 04 70 25 00 00 04 08 00 64
20180325-21:32:01.463 error: invalid checksum: f2 17 80 23 42 18 00 25 45 15 02 fb 1f 00 00 04 40 00 4e
20180325-21:32:06.053 error: invalid checksum: f2 17 80 23 42 06 00 25 46 01 04 70 20 00 00 04 40 00 6e
20180325-21:32:07.741 error: invalid checksum: f2 17 80 23 42 18 00 25 40 21 02 7e 22 00 00 04 40 00 76
20180325-21:32:11.521 error: invalid checksum: f2 17 80 23 42 18 00 25 47 01 04 70 01 00 00 04 40 00 6c
20180325-21:32:20.566 error: invalid checksum: f2 17 80 23 42 18 00 25 49 15 02 7e 07 00 00 04 40 00 43
20180325-21:32:25.089 error: invalid checksum: f2 17 80 23 42 18 40 25 43 01 04 70 16 00 00 04 40 00 5d
20180325-21:32:26.844 error: invalid checksum: f2 17 80 23 42 18 00 25 a9 15 02 7e 21 00 00 04 40 00 0a
20180325-21:32:29.611 error: invalid checksum: f2 17 80 88 42 18 00 25 42 01 04 70 0c 00 00 04 40 00 46
20180325-21:32:31.366 error: invalid checksum: f2 17 80 23 42 18 00 22 42 01 04 70 0c 00 00 04 40 00 46
20180325-21:32:35.889 error: invalid checksum: f2 17 80 23 42 0c 00 25 44 01 04 70 14 00 00 04 40 00 58
20180325-21:32:48.714 error: invalid checksum: f2 17 80 23 42 18 00 25 40 01 00 70 0b 00 00 04 40 00 43
20180325-21:32:53.236 error: invalid checksum: f2 17 80 23 21 18 00 25 3e 15 02 7e 1d 00 00 04 40 00 37
20180325-21:32:54.991 error: invalid checksum: f2 17 80 23 42 18 00 25 09 01 04 70 01 00 00 04 40 00 37
20180325-21:32:57.759 error: invalid checksum: f2 17 80 23 42 01 00 25 3d 15 02 7e 20 00 00 04 40 00 09
20180325-21:32:59.446 error: invalid checksum: f2 17 80 23 10 18 00 25 3d 15 02 7e 20 00 00 04 40 00 09
20180325-21:33:09.571 error: invalid checksum: f2 17 80 23 42 18 00 25 45 15 02 7e 07 00 00 04 40 00 4e
20180325-21:33:12.339 error: invalid checksum: f2 17 80 23 42 01 00 25 43 01 04 70 16 00 00 04 40 00 5d
20180325-21:33:14.161 error: invalid checksum: f2 17 80 23 42 18 00 25 43 20 04 70 16 00 00 04 40 00 5d

I don't really understand what's happening here. It looks a bit like random bit errors. could it be a birate mismatch between CPU and BLE module?

Maybe someone with more experience with this kind of errors can comment on this.

This looks like a baudrate problem, or maybe wrong start bit detection or something like this. E.g. when 0x42 turns into 0x21, one bit shifted to the right. Same for 0x18 turning int 0x0c, one bit shifted.

Very unlikely that it has anything to do with the RF transfer from the Bluetooth module to the receiver, because as I noted earlier, BLE has a sophisticated error correction. Not just xor, but CRC, which is easy to calculate with tables, I've implemented this once myself for another product. You can verify this if you increase the distance: there shouldn't be more errors, but it gets slower, because packets gets resend, at least when I tested it with my Blue Gecko Silabs IC, which had no packet errors at all at the application level.

Maybe you can measure the RX/TX lines between the Bluetooth module and the CPU to check if the baudrate is exact in your multimeter, if David can't reproduce it?

Alright so I've logged data for 2 minutes on BLE and UART then validated the packets, it doesn't appear i'm getting any of the bad packets on android.
Similarly on windows there were no bad packets.

Do you have any specific conditions which might help reproduce this issue.

NOTE: If you compile my main C++ file be sure to set struct alignment to be for bytes not like 32 bit integers or something, often the default is not byte.
« Last Edit: March 27, 2018, 06:51:59 am by Seppy »
 
The following users thanked this post: Iagash

Offline Iagash

  • Regular Contributor
  • *
  • Posts: 69
  • Country: de
Re: EEVBlog 121GW Discussion thread
« Reply #465 on: March 27, 2018, 08:50:33 am »

Alright so I've logged data for 2 minutes on BLE and UART then validated the packets, it doesn't appear i'm getting any of the bad packets on android.
Similarly on windows there were no bad packets.

Do you have any specific conditions which might help reproduce this issue.

NOTE: If you compile my main C++ file be sure to set struct alignment to be for bytes not like 32 bit integers or something, often the default is not byte.

Thanks a lot. Yes, your data looks good.

Can you explain how to log the raw data on android?

Then I'd like to try this with my meter to see if the problem is still there.

I tried with 3 different linux systems, my desktop, a thinkpad X230 and a thinkpad X220. They all show the same invalid records.

Maybe I have a bad meter?
« Last Edit: March 27, 2018, 09:20:12 am by Iagash »
 

Offline Brumby

  • Supporter
  • ****
  • Posts: 12288
  • Country: au
Re: EEVBlog 121GW Discussion thread
« Reply #466 on: March 27, 2018, 10:04:15 am »
Just a question to clear it from the discussion....

Since this communication is still an RF link - have you tried alternate orientations between Tx and Rx, different distances, different environmental elements, such as RF reflective structures, RFI, etc?
 

Offline Seppy

  • Supporter
  • ****
  • Posts: 189
  • Country: au
  • Curious
Re: EEVBlog 121GW Discussion thread
« Reply #467 on: March 28, 2018, 12:41:14 am »

Alright so I've logged data for 2 minutes on BLE and UART then validated the packets, it doesn't appear i'm getting any of the bad packets on android.
Similarly on windows there were no bad packets.

Do you have any specific conditions which might help reproduce this issue.

NOTE: If you compile my main C++ file be sure to set struct alignment to be for bytes not like 32 bit integers or something, often the default is not byte.

Thanks a lot. Yes, your data looks good.

Can you explain how to log the raw data on android?

Then I'd like to try this with my meter to see if the problem is still there.

I tried with 3 different linux systems, my desktop, a thinkpad X230 and a thinkpad X220. They all show the same invalid records.

Maybe I have a bad meter?

At the moment raw packet logging is not a feature of the app, I use debug outputs (in visual studio, with debug mode) to get the data.

There are a fair few apps on the play store which seem suitable, I haven't used these but they appear to do what you want:
https://play.google.com/store/apps/details?id=com.zicasoftware.ziblemonitor
https://play.google.com/store/apps/details?id=com.edodm85.bluetoothbleterminal.free


« Last Edit: March 28, 2018, 04:53:45 am by Seppy »
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 37661
  • Country: au
    • EEVblog
Re: EEVBlog 121GW Discussion thread
« Reply #468 on: March 28, 2018, 01:44:27 am »
I tried with 3 different linux systems, my desktop, a thinkpad X230 and a thinkpad X220. They all show the same invalid records.
Maybe I have a bad meter?

Do you have any issues at all with the range switch?
There is a possibility that could be an issue.
 

Offline FrankBuss

  • Supporter
  • ****
  • Posts: 2365
  • Country: de
    • Frank Buss
Re: EEVBlog 121GW Discussion thread
« Reply #469 on: March 28, 2018, 07:38:15 am »
I tried with 3 different linux systems, my desktop, a thinkpad X230 and a thinkpad X220. They all show the same invalid records.
Maybe I have a bad meter?

Do you have any issues at all with the range switch?
There is a possibility that could be an issue.

How could the range switch cause invalid xor checksums?
So Long, and Thanks for All the Fish
Electronics, hiking, retro-computing, electronic music etc.: https://www.youtube.com/c/FrankBussProgrammer
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 37661
  • Country: au
    • EEVblog
Re: EEVBlog 121GW Discussion thread
« Reply #470 on: March 28, 2018, 07:41:13 am »
I tried with 3 different linux systems, my desktop, a thinkpad X230 and a thinkpad X220. They all show the same invalid records.
Maybe I have a bad meter?

Do you have any issues at all with the range switch?
There is a possibility that could be an issue.

I was thinking of the spurious VA values. We no longer get these as it's been (as far as we can tell) fixed in software. So if spurious values are being seen, maybe that's a possible cause?
But if it's a checksum thing, then forget I said it.
But as I believe David (seppy) has said, we are no longer getting corrupted data, and hence no more checksum errors.

How could the range switch cause invalid xor checksums?
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 37661
  • Country: au
    • EEVblog
Re: EEVBlog 121GW Discussion thread
« Reply #471 on: March 28, 2018, 07:45:55 am »
I'll look into this, not sure what caused it but i'll do my best to reproduce, it may take a while to reproduce these type of issues can sometimes take a while to figure out the exact mechanism as its likely something specific about the android API. I'll get back to you or just release a fix as an update if it is reproducible.

Thanks, but don't spend much time on this. Short logs are working fine. The fault could also be my phone I just tried it once.
I'd be much more interested if you can see the same data errors that I see with my linux script.
Maybe you could add logging of these failed records raw data to your app, so that I can compare these?

Is anyone else having issues with the bluetooth data? If not then we probably don't have a choice at this point but to consider the bluetooth issues fixed, as always subject the future any future identified and repeatable issues.
UEi fixed a bug in the micro send routine sending spurious data, and we have shortened (more than halved) the packet size to reduce known packet loss were were seeing. We no see no issues at all at our end on our app.
 

Offline Iagash

  • Regular Contributor
  • *
  • Posts: 69
  • Country: de
Re: EEVBlog 121GW Discussion thread
« Reply #472 on: March 28, 2018, 08:29:04 am »

At the moment raw packet logging is not a feature of the app, I use debug outputs (in visual studio, with debug mode) to get the data.

There are a fair few apps on the play store which seem suitable, I haven't used these but they appear to do what you want:
https://play.google.com/store/apps/details?id=com.zicasoftware.ziblemonitor
https://play.google.com/store/apps/details?id=com.edodm85.bluetoothbleterminal.free

Thanks.

I tried these and they can talk to the meter, but seem unable to receive and log the data packets.

 

Offline Iagash

  • Regular Contributor
  • *
  • Posts: 69
  • Country: de
Re: EEVBlog 121GW Discussion thread
« Reply #473 on: March 28, 2018, 08:37:37 am »

I was thinking of the spurious VA values. We no longer get these as it's been (as far as we can tell) fixed in software. So if spurious values are being seen, maybe that's a possible cause?
But if it's a checksum thing, then forget I said it.
But as I believe David (seppy) has said, we are no longer getting corrupted data, and hence no more checksum errors.

How could the range switch cause invalid xor checksums?

I don't see spurious VA values as in the early versions of the firmware.

With V1.10 I still see wrong voltage and current displays in certain parts of the measuring range, I see a spurious wrong decimal point in current display and I see wrong range display (A instead of mA) in the current display. All of this in DC VA-mode.

Apart from that the displayed values are consistently wrong.

I guess the invalid checksums must then be some electrical problem in my meter if no one else can see them.

As there seems no way to debug it further, I can't help you there.
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 37661
  • Country: au
    • EEVblog
Re: EEVBlog 121GW Discussion thread
« Reply #474 on: March 28, 2018, 09:37:44 am »
With V1.10 I still see wrong voltage and current displays in certain parts of the measuring range, I see a spurious wrong decimal point in current display and I see wrong range display (A instead of mA) in the current display. All of this in DC VA-mode.
Apart from that the displayed values are consistently wrong.

On the meter display or the app display? (or both)
Does the SD card logged data show correct or incorrect readings?

Quote
I guess the invalid checksums must then be some electrical problem in my meter if no one else can see them.

Again, how are you validating your checksums? with your own software?
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf