- This topic has 0 replies, 2 voices, and was last updated 6 years, 7 months ago by Anonymous.
- June 11, 2016 at 6:18 pm #10122AnonymousInactive
I purchased from Sparkfun about 10 days ago.
Display software is running on Nexus 7 (2013), Android 4.4.4, OS is Cyanogenmod v11-20141115 M12 flo.
When I originally received the Mooshimeter, I checked VDC accuracy against my Fluke DVM’s, and everything looked good. I was running older release firmware and software on the device as received from Sparkfun, as evidenced by the black background graphs and the fact that it automatically switched to graph mode on android device rotation.
I joined the beta test program and upgraded to the latest firmware & software beta a couple of days ago:
App version 1.0.22 (1820)
Firmware version 1462564395
Now the VDC readings are inaccurate- I’m getting higher readings from Mooshi than from my Flukes- I have two Fluke DVMs that agree with each other to 0.01v, but the Mooshi is reading approx 3.5% higher, IE I’m getting 13.945vdc on the Mooshi when both my Flukes indicate 13.42vdc, and 2.770vdc on the Mosshi when my Flukes show 2.68vdc (battery measurements)
I’d swear that the Mooshi & Flukes agreed to 0.01vdc when I first tested, before updating to the latest firmware & software beta, and I’d love to “back down” to the original software & firmware to verify whether the update is at fault, but although reverting the Android software should be fairly easy, it’s not clear that I can easily revert the firmware.
Has anyone else seen this, or have any advice?
Thanks & Aloha-
- June 15, 2016 at 4:02 am #10163AnonymousInactive
Looking through older forum posts, I found a note [“Problems with Android App 1.0.18 (993)”] about the meter calibration data being stored in flash alongside the firmware- now I’m wondering whether my calibration data may’ve been overwritten or zero’d out?
I did have some difficulty with the firmware update, it took at least two tries to succeed (Android 4.4.4 issues? or inexperience?)
Still wondering whether there is a way to backdown to my original firmware, to check whether that might fix the problem. Does the original release software have the original firmware imbedded, like the beta update did?
(IE then backdown/revert would be a similar process to the original beta update)
- June 15, 2016 at 9:58 am #10165AnonymousGuest
I found this in the forums here from James. I think it will take you all the way back to the latest non-beta firmware. Maybe it will help you figure out the accuracy issue. — Duane
“Roll back to older android app version:
Yes, you can leave the beta program and reinstall the app from the google play store. The link to enroll/leave the beta is here:
Note: if you upgraded your Mooshimeter firmware, you will need to downgrade it to work with the older app version.”
- June 15, 2016 at 3:03 pm #10190AnonymousInactive
Thanks Duane- yes I saw that post too, but have not tried reverting yet because James also sez:
“If you upgraded your Mooshimeter firmware, you will need to downgrade it to work with the older app version”-
As I mentioned above, I’m not sure how to roll back the FIRMWARE… if it’s bundled with the software, as it is with the beta, then it should be straightforward, but I don’t know whether that’s true or not..
- June 20, 2016 at 10:45 pm #10306adminKeymaster
Just wanted to confirm that for the Android apps, the firmware is bundled with the app. You can move to an older version of the app by leaving the beta program as Duane suggests, or if you’re installing via APK files directly you can see all available releases at https://moosh.im/s/
Hope this helps, sorry for the trouble
- June 22, 2016 at 4:05 pm #10331AnonymousInactive
Thanks James- yes I have now reverted software & firmware to my original:
Nexus 7 Android 4.4.4
Mooshimeter Version 1.0.17(444)
But my VDC readings are still high by 2.5%-3.5%
Fluke <=> Mooshimeter
39.44 <=> 40.75
26.25 <=> 26.92
13.17 <=> 13.50
The mooshimeter readings will slowly fall, for instance the 40.75v reading above had fallen to 40.45v after several minutes (Fluke was still showing 39.44v)
I also noticed that after taking a VDC measurement, if I shorted the mooshimeter V-C leads, the meter would initially show -0.15 vdc, slowly rising toward zero (-0.045v after 2 minutes), as if there was some capacitance bleeding off.
I’m still wondering whether I have somehow erased my calibration data, during the beta updates, or perhaps the meter is damaged? I’ve only used it for low level battery VDC readings so far (0-40vdc)
Aloha James- hope I can get this fixed, I really like this meter!
- June 23, 2016 at 9:01 pm #10352AnonymousInactive
James- Please advise: Should I pursue an RMA return/exchange from Sparkfun, or wait for suggestions? I am happy to do further testing or investigation before returning this unit, or to return it directly to you for analysis- my needs are not urgent.
I really like what you’ve done here- this is a great design, and something I’ve needed for a long time. There are other DVM/logging devices out there with less capability at a greater cost- I want Mooshimeter to succeed! I’m sure others feel the same- I don’t want to disrail your efforts.
Aloha & Mahalo!
- June 24, 2016 at 1:37 am #10359adminKeymaster
Sorry for delayed response Joel – I started drafting a reply thinking about what could go wrong and got distracted.
There are two slightly different hardware versions in the wild, and I think it sounds like your meter somehow got confused about what kind it is. You can see what kind you have in the silkscreen on the bottom of the PCB: RevH or RevI.
RevH has a 1.25V reference voltage. RevI has a 1.21V reference voltage. I’m guessing that maybe you have a RevI that thinks it’s a RevH – that would make all readings go off by a factor of 1.25/1.21 = 1.033 or 3.3%. The resistance measuring circuitry also changed quite a bit, if my diagnosis is correct your resistance measurements would be totally bunk. Is that the case?
I haven’t dug in to the app history but I see you’re using Android app version 1.0.17. Can you try reinstalling firmware from release 19 (app-release-19.apk in the directory I linked earlier) and measuring again? Sorry for the trouble. Release 19 is the present production release.
Regarding theory of calibration data wiped: I can assure you that didn’t happen because the meter would not function without it.
Sorry this got long – thanks again for the excellent feedback. User feedback drives product development. Let me know how it goes – if the firmware update does not fix the issue email me at email@example.com and I’ll handle the return.
- June 24, 2016 at 12:57 pm #10370AnonymousInactive
Thanks James for the suggestions- I know you must be juggling a *lot* of issues right now, and I very much appreciate your personal help & support.
You are correct that my PCB shows “RevI 04/2015”, and the 3.3% offset you mention sounds exactly like what I am seeing on the VDC scale (except for the drift over time?).
I just checked resistence measurements: in lower ranges (100’s of ohms) Mooshi matches my Fluke to 1% or better, at higher ranges (K-ohms to M-ohms) I’m seeing a 2%-4% high reading, similar to my VDC results. Not totally wonky, just a bit off…
I’ll get the release 19 APK today, and report my results after loading.
Thankyou so much for this device, and esp. for your stellar support-
- June 27, 2016 at 3:45 pm #10426AnonymousInactive
(oops- this never got sent last Friday? Retry:)
I uninstalled Mooshimeter from my Nexus 7 (Android 4.4.4), downloaded app-release-19.apk directly from https://moosh.im/s/, and installed that.
Surprisingly, install of app-release-19.apk showed as “Version 1.0.20 (1769)(20), new build 1460332895”. But install proceeded OK, and Nexus shows “Mooshimeter V.1, Build 1460332895, Android app 1.0.20 1769)”
But- I still getting innaccurate battery VDC measurements, and the indicated voltage falls slowly with time:
Fluke <=> Mooshimeter
39.82 <=> 41.67
39.80 <=> 40.53 (after 5 minutes)
39.61 <=> 40.35 (after 90 minutes)
(note that actual battery VDC was falling slowly after initial charging)
Thanks James- I’ll be emailing you separately about the possibility of a return/exchange.