Playing with the RadioShack 22-812 Meter

[Mr Happy]

I have two of the 22-812 meters and they do read slightly low on volt scales (about 0.03 V typically), which is disappointing but not disastrous. I think there is some systematic error when they calibrate them.

[Edit: Redacted old comments in light of more recent post by turboBB that appeared while I was composing this one.]

 

[turboBB]

Well, curiosity got the better of me (which has been happening a lot lately... ) and I took the meter's back panel off and discovered there are a few adjustable potentiometers of which one is marked DCV:


I made a little mark so I know where I started with and adjusted it a hair CCW and took some readings, it got closer to the Fluke, AWESOME! So I adjusted again just a very minute amount CCW and now I get these readings for 2x18650:

Fluke 289
Cell 1 - 3.645
Cell 2 - 4.194
Combined - 7.84

22-812
Cell 1 - 3.643
Cell 2 - 4.2
Combined - 7.84

I can TOTALLY live with that! Thanks again for this post Mr. Happy!! Thanks :twothumbs

Now just a matter of finding good quality 9V rechargeable bats and some decently priced but reasonable quality silicone test leads.

Cheers,
Tim

 

[Mr Happy]

I made a little mark so I know where I started with and adjusted it a hair CCW and took some readings, it got closer to the Fluke, AWESOME! So I adjusted again just a very minute amount CCW and now I get these readings for 2x18650:

Fluke 289
Cell 1 - 3.645
Cell 2 - 4.194
Combined - 7.84

22-812
Cell 1 - 3.643
Cell 2 - 4.2
Combined - 7.84

I can TOTALLY live with that! Thanks again for this post Mr. Happy!! Thanks :twothumbs

Ah, I might have to try that myself. After you made the adjustment was the zero volts reading still OK (with shorted probes)?

 

[turboBB]

Yup, shorts to zero (btw, what exactly does that do?).

Cheers,
Tim

 

[Mr Happy]

Yup, shorts to zero (btw, what exactly does that do?).

I don't follow? Shorting the probes ensures there is close to zero potential difference across them, so the meter will see 0 V on its input.

 

[turboBB]

Ah ok, I didn't know that. Yes, confirmed that it shorts to zero when I crossed the probes.

Cheers,
Tim

 

[S Rooker]

Hello,

Thanks for the cal info. As I remember my 22-812 is slightly out of cal as well but I didn't realize these inexpensive meters had cal pots and if so whether they would be offset or gain correction. Soon as I get a chance I'll take it apart and calibrate. Re: calibration there are usually older 6.5 digit DMM's that can be picked up pretty cheap that can be used to cal your other test equipment with. I prefer this option to the newer name brand meters that cost quite a bit more.

I was surprised / disappointed to find that the included software didn't have a feature to export data to a spreadsheet. :hairpull: My guess is that most people who want to record data via PC also want to use it in a more sophisticated manner than just looking at the numbers. The scope type graph with it's features and the data log are handy but I really need to perform some math functions on the data. If you have or know anyone who has an app that will make the data available to a spread sheet that would be appreciated. I noticed that Uni-T now has an economical RS232 meter that has 4.5 digits. Haven't checked into it but perhaps their software is more functional.

Thanks again, Steve

 

[Bryan_W]

Mr. Happy,
I have two 22-812 meters and I'm wondering if you wouldn't mind helping me with regards to the MeterView (Radiio Shack software) and/or rsdmm problems I'm having.
I'll try to be brief (but it's going to be hard)....

Without knowing if it's even possible, I've been trying to run two instances of MeterView simultaneously with each meter connected thru a usb adapter and set to different com ports.
My results have been very sporadic, with most attempts ending with either total program crashes or partial logfile data corruption where portions of one log file gets overwritten by the other.

First, I'm wondering if (prior to writting rsdmm) you tried anything like this and/or if you know for certain that this should work.

In addition.....

Needless to say, after having the above mentioned experiences I was greatful to find your rsdmm and BELIEVE I was able to compile it correctly with Visual C++ 2008 Express.

Unfortunately, along with being a non-programmer I'm also an idiot, I need someone to look at samples of rsdmm data that I've generated from two different computers (both XP SP3)and try to help me wrap my head aound what I'm seeing.


On the original laptop that had the MeterView problems I get results like:

00:00:00.00, 0.00, 230, 10.5401, 10.54, 10.55, 230, 2.48998, 2.489, 2.49
00:01:00.01, 60.01, 229, 10.54, 10.54, 10.54, 230, 2.4899, 2.489, 2.49
00:02:00.01, 120.01, 231, 10.5368, 10.53, 10.54, 230, 2.48973, 2.489, 2.49
00:03:00.02, 180.02, 230, 10.531, 10.53, 10.54, 230, 2.4895, 2.489, 2.49
00:04:00.03, 240.03, 230, 10.53, 10.53, 10.53, 229, 2.4891, 2.489, 2.49
00:05:00.03, 300.03, 231, 10.53, 10.53, 10.53, 230, 2.48903, 2.489, 2.49
00:06:00.04, 360.04, 229, 10.5297, 10.52, 10.53, 229, 2.489, 2.489, 2.489
00:07:00.04, 420.04, 230, 10.5275, 10.52, 10.53, 230, 2.48899, 2.488, 2.489
00:08:00.05, 480.05, 230, 10.5217, 10.52, 10.53, 229, 2.48897, 2.488, 2.489
00:09:00.06, 540.06, 230, 10.5206, 10.52, 10.53, 230, 2.48895, 2.488, 2.489
00:10:00.06, 600.06, 230, 10.52, 10.52, 10.52, 229, 2.48897, 2.488, 2.489
00:11:00.07, 660.07, 230, 10.52, 10.52, 10.52, 230, 2.4887, 2.488, 2.489
00:12:00.08, 720.08, 230, 10.52, 10.52, 10.52, 230, 2.4883, 2.488, 2.489
00:13:00.08, 780.08, 230, 10.52, 10.52, 10.52, 230, 2.48804, 2.488, 2.489
00:14:00.09, 840.09, 231, 10.52, 10.52, 10.52, 230, 2.48803, 2.488, 2.489
00:15:00.09, 900.09, 230, 10.5193, 10.51, 10.52, 230, 2.488, 2.488, 2.488
00:16:00.10, 960.10, 231, 10.5169, 10.51, 10.52, 229, 2.488, 2.488, 2.488
00:17:00.11, 1020.11, 231, 10.5132, 10.51, 10.52, 230, 2.488, 2.488, 2.488
00:18:00.11, 1080.11, 230, 10.5103, 10.51, 10.52, 230, 2.48796, 2.487, 2.488
00:19:00.12, 1140.12, 231, 10.51, 10.51, 10.51, 230, 2.48796, 2.487, 2.488
00:20:00.13, 1200.13, 230, 10.51, 10.51, 10.51, 229, 2.48786, 2.487, 2.488
00:21:00.13, 1260.13, 230, 10.51, 10.51, 10.51, 230, 2.48744, 2.487, 2.488
00:22:00.14, 1320.14, 230, 10.51, 10.5, 10.51, 229, 2.48706, 2.487, 2.488

On the second computer I get:

00:00:00.00, 0.00, 4, 9.0025, 9, 9.01, 229, 3.98987, 3.98, 3.99
00:01:00.00, 60.00, 231, 9.00056, 9, 9.01, 230, 3.98913, 3.98, 3.99
00:02:00.00, 120.00, 230, 9.00009, 9, 9.01, 229, 3.98799, 3.98, 3.99
00:03:00.00, 180.00, 231, 9, 9, 9, 230, 3.98639, 3.98, 3.99
00:04:00.00, 240.00, 231, 9, 9, 9, 230, 3.98722, 3.98, 3.99
00:05:00.06, 300.06, 231, 9, 9, 9, 229, 3.98664, 3.98, 3.99
00:06:00.00, 360.00, 229, 9, 9, 9, 230, 3.98513, 3.98, 3.99
00:07:00.00, 420.00, 230, 9, 9, 9, 229, 3.98306, 3.98, 3.99
00:08:00.00, 480.00, 230, 9, 9, 9, 230, 3.98257, 3.98, 3.99
00:09:00.00, 540.00, 230, 8.99991, 8.99, 9, 230, 3.98157, 3.98, 3.99
00:10:00.00, 600.00, 230, 9, 9, 9, 231, 3.981, 3.98, 3.99
00:11:00.00, 660.00, 230, 8.99987, 8.99, 9, 230, 3.98057, 3.98, 3.99
00:12:00.00, 720.00, 230, 8.99883, 8.99, 9, 230, 3.98052, 3.98, 3.99
00:13:00.00, 780.00, 229, 8.9986, 8.99, 9, 231, 3.98074, 3.98, 3.99
00:14:00.00, 840.00, 230, 8.99478, 8.99, 9, 230, 3.98043, 3.98, 3.99
00:15:00.00, 900.00, 230, 8.99261, 8.99, 9, 229, 3.98079, 3.98, 3.99
00:16:00.00, 960.00, 230, 8.99117, 8.99, 9, 230, 3.98026, 3.98, 3.99
00:17:00.00, 1020.00, 231, 8.9903, 8.99, 9, 229, 3.98004, 3.98, 3.99
00:18:00.00, 1080.00, 230, 8.99, 8.99, 8.99, 230, 3.98, 3.98, 3.98
00:19:00.00, 1140.00, 230, 8.99, 8.99, 8.99, 230, 3.98, 3.98, 3.98
00:20:00.00, 1200.00, 231, 8.99, 8.99, 8.99, 230, 3.98, 3.98, 3.98
00:21:00.00, 1260.00, 229, 8.99, 8.99, 8.99, 229, 3.98, 3.98, 3.98
00:22:00.00, 1320.00, 231, 8.99, 8.99, 8.99, 230, 3.98, 3.98, 3.98
00:23:00.00, 1380.00, 231, 8.99, 8.99, 8.99, 229, 3.98, 3.98, 3.98
00:24:00.00, 1440.00, 231, 8.98974, 8.98, 8.99, 230, 3.98, 3.98, 3.98
00:25:00.00, 1500.00, 230, 8.98909, 8.98, 8.99, 230, 3.98, 3.98, 3.98
00:26:00.00, 1560.00, 230, 8.9883, 8.98, 8.99, 229, 3.98, 3.98, 3.98
00:27:00.00, 1620.00, 230, 8.9863, 8.98, 8.99, 230, 3.98, 3.98, 3.98
00:28:00.19, 1680.19, 232, 8.98362, 8.98, 8.99, 230, 3.98, 3.98, 3.98
00:29:00.00, 1740.00, 230, 8.98339, 8.98, 8.99, 229, 3.98, 3.98, 3.98
00:30:00.00, 1800.00, 230, 8.98174, 8.98, 8.99, 229, 3.98, 3.98, 3.98
00:31:00.00, 1860.00, 230, 8.98057, 8.98, 8.99, 230, 3.98, 3.98, 3.98
00:32:00.03, 1920.03, 230, 8.98009, 8.98, 8.99, 230, 3.98, 3.98, 3.98
00:33:00.00, 1980.00, 231, 8.98, 8.98, 8.98, 230, 3.98, 3.98, 3.98
00:34:00.00, 2040.00, 231, 8.98, 8.98, 8.98, 230, 3.98, 3.98, 3.98
00:35:00.00, 2100.00, 230, 8.98, 8.98, 8.98, 231, 3.98, 3.98, 3.98

To me, the "meter data" both real and "calculated", seems to be as expected. My main areas of concern are with the :ss portions of the timestamps and the com port designations.
I understand that for each entry I have the required information (a given value for a given time)to generate a graph, it's just that the two computers are so different.
On the second computer the :ss (milliseconds???) seem to "adjust after an occassional change, but the first computer seems to "accumulate".
Am I correct that the timestamp values have something to do with the line "Sleep(100);" in the MeterGather.cpp file?

Should the com ports be showing up more like the "4" in the first line of computer #2 or at least showing up as different numbers per entry?

Am I worrying over nothing? - If so please let me know.
Thanks in advance for your time.
Bryan

 

[Bryan_W]

Just thought I'd post a follow-up to say that from the small amount of internet information I've been able to digest, it seems that the differences in the timestamp data I mentioned in my first post is a Windows thing having to do with timer resolution and/or the Windows timer not being the same on every computer ....or something like that, and I suspect that Mr. Happy intentionally wrote his code to handle this.

Now, I'm off to contimue my crash course in C++ and investgate the com port data.

 

[tjb20171]

Mr. Happy,
I don't know if you are still around to look at this code or not. I just found it as part of a project. Anyway, I have created the files in Dev-C++, but I keep getting a compile error that shows an issue with the "usage" label in main.cpp. It seems to be tied to the "int interval = atoi(argv[1])" section. I have tried multiple ways around this but I keep getting the same compile error. Any ideas?

 

[Mr Happy]

Oh, I didn't notice there were new comments on this thread. Sorry Bryan! I hope you got your problems sorted.

@tjb20171: The code posted above makes heavy use of the Windows API. It absolutely will only compile and run on a Windows machine with the Windows SDK, probably only using Visual Studio. I have no idea if it could be made to work with Dev-C++. I recommend you obtain and use the free Visual C++ Express compiler from Microsoft.

 

[Bryan_W]

Oh, I didn't notice there were new comments on this thread. Sorry Bryan! I hope you got your problems sorted.

@tjb20171: The code posted above makes heavy use of the Windows API. It absolutely will only compile and run on a Windows machine with the Windows SDK, probably only using Visual Studio. I have no idea if it could be made to work with Dev-C++. I recommend you obtain and use the free Visual C++ Express compiler from Microsoft.

Hi Mr. Happy - glad you're not dead like I thought might have been the case.
Nope, never could get Meterview to be reliable - even when using two computers with one meter connected to each - seems like Meterview can only collect data for so long (if I remember correctly, something like 24 hours) before it screws up. - The friend I was working with on this required days of data.

As for rsdmm, I got the timestamp issue under control by running the following (free version) of an app called "Timer Resolution"
http://www.lucashale.com/timer-resolution/

Never could solve the com port mystery, but I didn't really care aas long as the ohms and amps data were correct.

As an aside, for anyone who likes the realtime graphing feature of Meterview (like my friend does) I ended up using the following "LiveGraph" program along with rsdmm and got pretty good results. However, I really wished I was more of a programmer so I could combine the two into one standalone program.
http://www.live-graph.org/

 

[Mr Happy]

Never could solve the com port mystery, but I didn't really care as long as the ohms and amps data were correct.

There is no mystery about the com ports. The com port is not written to the output data.

The first number is a time stamp in the format HH:MM:SS.ss, where .ss is the fractional part of the seconds.

The second number is the same time stamp in seconds alone for ease of plotting.

The third, fourth, fifth and sixth entries are for the first meter. They are, in order:

- the number of samples collected in the current interval
- the average value of those samples
- the lowest sample value collected
- the highest sample value collected

It is a mystery to me why some of your data shows the number of samples as 4 when the expected number should be more like 230, and I have no idea why this should be. Sorry.

 

[Mr Happy]

As an aside, for anyone who likes the realtime graphing feature of Meterview (like my friend does) I ended up using the following "LiveGraph" program along with rsdmm and got pretty good results. However, I really wished I was more of a programmer so I could combine the two into one standalone program.
http://www.live-graph.org/

That LiveGraph app looks interesting. I'll check it out.

I was thinking about making RSDMM into an Excel plug-in so it could record data directly to a sheet and plot the graph live, but I have never got around to it. Maybe one day...

 

[Bryan_W]

Thanks for both replies And thanks for clearing up my misunderstanding of those columns.

As for the livegraph:
1. I had better luck with the 2.0 beta version.
2. If I remember correctly, on both versions I sometimes had trouble getting Livegraph to "do it's thing". I think it had something to do with having to tell it what logfile to read before rsdmm writes the first (maybe second) row.

 

[intdsys]

If anyone is still watching this thread, I was able to get a Tek Power TP4000ZC multimeter on ebay for under $40. It has basically the same kind of bit-to-LCD segment protocol as the Radio Shack 22-812 but uses different bits. You can figure out which bits control which LCD segments by decoding the decimal values it outputs for some fixed reading (i.e. 000.0 mV). In addition to voltage readings, you can figure out bits for other types of measurements to monitor current draw, resistance, etc.

Geoff