Hi everyone,

would like to ask whether anybody has experienced problems for finding a GPS fix with the MTK when it is cold outside. I have had issues with mine during last autumn and decided to remove the battery from the adapter board, since it was reported to cause more trouble than be useful. This summer I had a couple of flights with outside temperatures around 20 degrees Celsius (should be 68 degrees Fahrenheit) and could find a fix in around 1 minute when having a clear view of the sky. Now it is around 10 degrees Celsius outside (50 degrees Fahrenheit) and it could not find a fix at all within 30 minutes yesterday.

I'm thinking it's temperature related since some days before that I could even get a fix inside on the window board.

Now I'm thinking to order the Ublox and hope that works better in this climate. But given that some people seem to try high altitude applications with the MTK I was wondering whether others have experienced issues..

Cheers, Andre

Views: 488

Reply to This

Replies to This Discussion

When I put it in the freezer I cannot get any GPS readings at all. ;)

How about the fridge then? :)

I get an error 31:35:20:63:30:31:64 repeated over and over

but seriously, good question, I'll be following with interest (sorry everyone, my input is just a bit of fun, but I really am interested in any experiences with temperature and how it is affecting your systems)

I definitely had trouble last winter with mine but the issue was not time to acquire the first fix during a cold start. It was causing my Mtk to reset the binary protocol but still maintain a fix. Essentially my unit would stop providing data to the Apm since it switched back to nmea format which was disabled in the Arducopter gps processing. At power up the software sends an init string to binary format since you can't flash the startup mode of the Mtk. If something in the Mtk hardware glitches you will stop getting gps updates. I saw this when my loiter suddenly stopped working in flight but the Mtk's light showed a fix. Recording the gps output with a logic analyzer showed this happening in the data when I cooled down the gps. Not exactly a reduction in precision, but a sudden system fault. Just a word of warning and a link to my original post.

I've also flown the Ublox in fixed wing applications and find that it works better for me than the MTK - particularly since it does have the backup battery on the adapter. It is just hard to mount in a vertical configuration on the stock Arducopter frame and it is quite a bit more expensive. 

Thanks for the reply! This confirms my suspicion about the temperature sensitivity. My Mediatek is of the same age (in terms of point in time when I ordered it at least). In your blog you write about trying a newer one, has this helped?

I would also really like to try the Ublox and get back into the air as long as the weather here permits it but it's permanently sold out in the store..

Since I haven't seen any other posts pointing to this issue I wonder whether others have just better weather or if they do not blame their GPS issues to temperature, if they have any..

Cheers, Andre

I bought a new one and it worked ok in the cold weather earlier this year. I didn't see the strange behavior with it and I wonder if they had a bad batch. I also have had issues with the cabling and just ran into that again lately with it dropping back into NMEA mode after power up. I have 3 total MTK's now and (1 with the temperature issue) and it will be interesting to see how they hold up as it soon will be getting cold here in midwest USA. I did notice a couple of posts back in January that reported their GPS loiter fail in the middle of flight.  I responded back to look at it possibly dropping out of binary and did not hear back.




Your cabling issue- would you say it's an RF/sheilding or interference issue, or stictly a physical connection issue? I'm curious, as I've also seen some inconsistent performance. GPS is reported well in Planner, but quad seems unaware of its location at times- usually after a successful flight, land and take-off.

Is it possibly a voltage issue?  What is the voltage level of the APM board?   If the 5v supply voltage on the APM is high enough, maybe the serial communication high signals from the 3.3v (or lower if the GPS regulator is drooping) GPS may not register in the 5v APM as a high signal.   That's what I've been wondering this weekend.


I admit the thought had crossed my mind but I had not confirmed it (just decided to order a U-Blox asap). I'll try to make some fridge tests in the next few days. Since I don't have an oscilloscope I guess all I can do is to check the supply voltage.

I have been following your post btw. If I understood correctly, you believe your BEC related issues are due to the 3.3V/5V differences which may cause problems for the signal levels. Please keep us posted on any conclusions. If this turns out to be the case I may give it a try adding a level shifter myself.

In my case, I'm pretty sure that the higher supply voltage turns on the issue.  I have a level shifter that could tear out of something else, and hook up between the gps and apm, and confirm if that allows it to work. 

Reply to Discussion


© 2020   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service