- This topic has 0 replies, 4 voices, and was last updated 4 years, 9 months ago by
.
- Posts
I picked up a Mooshimeter from SparkFun the other week and am having a few problems.
Android Nexus 5X using version 1.0.17 (444) and firmware 1445847981.
The first one is inconsistent logging. I tried to set up the meter to simply log internal temperature onto a 32GB SD card at 1 minute intervals over a 48 hour period.
While I did get some data, at some point in time the logging appeared to stop working and I ended up with well over a hundred csv files with one or two log entries in each. I’m still in the process of splicing all the files together to see how complete the data set is.
The second problem is that I am unable to update the firmware on my device – the Mooshimeter app crashes dialog says “Unfortuanately, Mooshimeter has stopped”. I’ve reported the crash a few times.
I haven’t tried the latest beta app, but I’ll try that shortly.
Update: I tried to reflash one more time and it worked (build 1447458797 uploaded), but now for some reason every time I try to connect to the device it brings me to the Firmware Update screen every time instead of connecting to the meter as normal. The LED is doing quick flashes as expected.
Have tried resetting, reprogramming (even in legacy mode).
Edit: Rebooting my phone for some reason got it working?
Hi Drees,
I’m happy to hear you got everything working. Regarding erroneously reconnecting in Firmware Update mode – this is due to an Android bug where it caches the services offered by a BLE device and doesn’t clear the cache even when those services change. So Android was remembering the Firmware Update services and not showing the Metering services to the app. This is why a phone reboot fixed that problem.
If you’re interested in reading more about the problem you can see it here:
http://stackoverflow.com/questions/22596951/how-to-programmatically-force-bluetooth-low-energy-service-discovery-on-androidLet me know if you have other questions, best
~JamesThanks for the followup.
The problem I’m currently having is getting SD card logging to work reliably. I was happy to see that 10 minute intervals have been introduced in the latest beta firmware!
The problem I’m having is that logging just either seems to stop working or work unreliably. Is it normal for the meter to create a bunch of files with only one line each?
So far, I’ve just been trying to log the temperature using the internal sensor with a measurement interval of 1 minute.
How can I troubleshoot this? Is there anything I can do to make sure logging is working properly? It’s kind of hard to tell just by watching the LED.
Hi Drees,
There were some updates to the logging in the latest firmware release. Please check this blog post, which also includes details on the new LED readout. It should be easier now to tell if logging is working.
https://moosh.im/2016/06/trifecta-beta-release-new-firmware-android-app-and-ios-app/
Best
~JamesI’m still having mixed results with the latest in the play store, Version 1.0.26 (1840) firmware 1466836533.
One issue is that the app gets stuck “LOADING” after changing logging settings, for example the timing. If I go out to the meter choosing screen and back that seems to get readings to work again most of the time.
Trying to log internal temperature at a 10 second interval worked for about the 10 minutes I tried it, but the values were like 2.947354E002 when the actual temperature is around 21C.
The latest improvements are looking good, but would be nice to get logging working properly, it’s one of the Mooshimeter’s best features.
I will continue to play around some more.
drees,
The first time I logged internal temperature I was shocked at the large values until I realized it was in Kelvin. So, 2.947354E002 —-> 294.7354 degK – 273.15 conversion constant = 21.585 degC. Similar to the timestamp conversions, I usually make another column in Excel spreadsheet and let Excel do the conversion of all the temperature data (to degF since I am in the USA).
— DuaneThat makes sense. But I think it’s also a new bug/regression in the code as the previous firmware used to log data (well, when I could get it to log, anyway) in either C or F depending on the setting.
I have been doing more logging, and one trick that seems to help in making sure that I have good log files is to be sure to connect to the meter and turn off logging before I pull the SD card.
Previously, I would just pull the SD card out.
What’s the proper procedure? Is disabling logging required? How often does the meter actually write to the SD card?
One other issue is that the internal temperature appears to be off by quite a few degrees on my meter. It is reading about 3-4 degrees C/K low compared to a few other sensors I have.
Hi Dave,
How often does the meter actually write to SD card:
Data is written to the SD card every 10 seconds. If you are logging with aggressive settings (like no time delay between samples) it might be written more often.Proper procedure for removing the SD card:
Turning logging off before removing the SD card is technically the best way to do it, but the chances of corrupting a file are so low that I never do it myself (I just pull out the SD card). If you’ve been consistently seeing corrupt files, can I ask what kind of SD card you’re using? I’ve seen that behavior a few times from no-name brands, but generally quite rarely.Internal temperature:
This is the on-chip temperature for the ADC (an ADS1292 if you are interested), and you’re right that it’s not extremely accurate for absolute temperature. So the readings you’re seeing are not abnormal.Let me know how else I can help, best
~JamesI wasn’t seeing corrupt log files, just really small log files. It was like logging was turning off/on frequently.
I haven’t seen this behavior since I started running beta (1.0.26) firmware, though, will continue to play around with it.
I’m using a 32 GB Silicon Power SD card, says it’s UHS-1 Class10.
https://www.amazon.com/dp/B00T9QONN6That said, the meter did kill a set of AA batteries after sitting on my desk a week without being connected. Logging wasn’t turned on. I’ve used the meter less than 3 weeks and haven’t had it connected very long, and when logging, typically used 1-minute logging intervals for a day or two here or there.
I expected to get at least a couple months out of a set of AA batteries without having to go into standby mode as long as a phone isn’t actively connected? What’s the minimum voltage that it will run at (say you have a power supply that doesn’t sag, what’s the minimum voltage)?
Has anyone used recharging NiMH AA batteries? Do they supply enough voltage?
Hi Dave,
Sorry for slow reply. I’ve been hunting some of these hard-to-reproduce bugs and have made some progress. I did more testing with 32GB SD cards (though mine were SanDisk) and also saw my logs being split sometimes unnecessarily. This happens when the firmware encounters an error writing to the SD card. I made the firmware smarter so that when it encounters a write error, it will resume where it left off instead of splitting in to a new log file (among other small improvements).
If you aren’t already in the beta program, I recommend joining it and updating meter firmware.
Regarding NiMH batteries: They work fine, but won’t last as long as AAs. The meter will refuse to start up on less than 2.4V. You should definitely be seeing battery life longer than 3 weeks, please let me know if you encounter the same issue with new batteries.
Best
~JamesHello,
Just received my Mooshimeter from Robotshop and I have a couple of issues with it.
My main concern is Resistance. I short the Ohm and Com sockets with a short lead and the meter reads 1.487R, which is unlikely but…. I then tap ‘ZERO’ assuming that will, well ‘Zero’ the reading, which it does. It then says 1.487 in the box where it previously said ‘ZERO’, and the resistance reading is 0.000R, as you’d expect. Now if I remove one end of the cable shorting the connection on the Mooshimeter, the reading displays ‘OUT OF RANGE’, obviously, but the box which previously read 1.487 now reads 0.00000MR, and if I short the sockets again, the resistance reading is 1.487R again, and the same reading is also in the ‘ZERO’ box????.
I would expect tapping ‘ZERO’ to just zero the reading, and when I removed the lead and replaced it, the meter should display a resistance of 0R again.Am I missing something?
Hi PureZX!
Yes, app does strange things with resistance+zeroing+autorange.
Zeroing works just fine with manual ranges.
Like you said, it is broken now. If you or automation change ranges, zeroed value is no longer calculated to your results. Value is still shown in zero-box. That is wrong.
I’m not sure what would be the right way to handle it though. Or what other users are doing with it. Someones twisted workflow might get broken, but i’m still going to write couple thoughts of how i think it should work:
Maybe zeroing should be limited to 1 kohm range only. Not sure if that shift in results should be applied on bigger ranges or not, but when back at 1 kohm range, it should go active again. For at least as long as the app is connected to mooshimeter (as that zeroing is probably only made in phone and mooshimeter does not know about it).
In real life, that zero-shift should not be that many ohms (like <5 ohms, I personally get new test leads when old ones go over 2 ohms), so at 1 kohm and over it would disappear in noise and other errors anyway. It should not matter if it is always calculated in results or only when at 1 kohm range.
For a comparison of test leads, if I remember right, Fluke 50 m (~160 feet) protective earth continuity test lead that I use on job sites is only 1.3 ohms.
Actual use case for mooshimeters zeroing feature is to enable us to measure smaller than 20 ohm things, as that was the lower limit on original specs. There are no promises about precision etc at that level.. Check your results, like you should do with any measuring device, if those values really matter to you. I have 4 multimeters at my hands reach now and I most likely use more than one every time i have to measure something ;)
