I have just been testing my GPS as it takes me ten minutes to get a 3d fix? so I decided to see what is going on, and yes it takes ten minutes still, but it indicates a 3d fix and sat count 3? this I thought is impossible? you need at least 4 sats for a 3dfix? after 18 minutes 4 sats and 20 minutes 5 sats 21 minutes 6 sats this is being tested outside I have tested with and without receiver connections and taking the board out of the Tricopter and getting the same results. Surely this cant be right for a gps that claims a lock in under a minute?
You need 4 sats for a 3D fix. The first time out of the box should take perhaps 12-15 minutes to download almanac data from the sats. After that it should take only a minute or two for a fix.
The DIY Drones "custom" firmware cripples some of the MediaTek GPS chipset functions, like saving your settings. Maybe it is also failing to save the almanac data. I'm also not sure if it's maintaining the empheris data either, so it probably requires a cold start each time you power it up.
I don't think it's losing the almanac data though, since my unit locks after a few minutes. Something must be wrong with your unit.
Hence my surprise when the sat count was 3 and I had a 3d fix? I have used it this week and it took maybe 5 minutes, but now it takes at least 10 and thats only 4 sats...
I have the same chipset GPS from a different module maker and it's performance is greatly reduced by a dirty electrical supply. So my thoughts are that either you're not getting good signal because you don't have a clear view of the sky or maybe you have a lot of electrical noise in your power system.
Lots of things could cause noise in your power, so maybe experiment with the wiring. Otherwise maybe the antenna got damaged or you have a bad unit. It can't have a 3d fix with 3 sats, so either the sat count is wrong or the fix is just 2d.
Very true. The (DIYD) MTK3329 may have many channels, may output data at a high rate, may have that spiffy keen binary interface, but it is one of the most "jammable" units I have. Power line noise, nearby RF from laptops, laser printers, etc. reduces signal levels and increases acquisition times as much or more than just going indoors.
I have tested it indoors and outdoors with similar results, apart from when its outdoors there is nothing that could cause interference because its in the middle of a field..
And the 3329 also has an anti jam capability according to the datasheet at ladyada..
RTC, almanac, and ephemeris all seem to be lost in under a second without power but I haven't done any quantitative testing to come up with a specific number. The caps on the adapter board may account for a couple hundred mS of retention.
As an aside, have you been able to install the non-DIYD 3329 firmware on one of these to see if the physical capability to store settings is intact? I wondered if the mfr was shipping with a smaller flash to save cost (and there simply wasn't room) or other "trick."
Do you have any info on the format the firmware is in? What processor it runs on? Could we disassemble and edit?
It would be nice to be able to upload via mavlink the satelite data as in agps that lasts 14 days..
I'm pretty sure it's keeping the almanac data, otherwise it would take 12+ min to lock at each startup. The DIY unit is using the same chipset as other modules that allow user settings to be saved in the firmware, so there's no reason the DIY unit shouldn't do the same, it's simply disabled in the crippled "custom" firmware.
I've got a fastrax up501 GPS module which uses the same MediaTek chipset. I don't really have much in the way of spare units to go trying to flash firmware on. There's always a good chance of bricking them.
What we really need is someone with some buying power (hint hint DIYDrones) to get the firmware source and/or the details and dev kit so that an open source firmware can be written. Then all sorts of restrictions can be removed and the firmware optimized for our specific application.
What they actually did is pay some shady Chinese company (GlobalTop Tech) to write some hack of a custom firmware. Some poor guy who can barely speak broken english spent a day writing it for a bowl of rice.
I have a similar problem. I have been testing my APM2 with the MediaTek GPS and the first time out the box the GPS took in the order of 45 minutes to download the almanac and get a fix. I thought maybe it's because I am in the Southern hemisphere that it takes longer. I left it running the whole day to make sure it catured the almanac properly. But then I powered it down and the next day it took about 15 minutes to get a fix. The blue LED was solid on, but the GPS status on the Mission Planner still said status 1 (no-fix). I tried again with only a short power down of a few seconds and it took another 10-15 minutes or so get a fix again. It seems like the almanac is not being stored.
The GPS unit you have has no battery backup. You need a battery to sustain the almanac and perform warm/hot starts.
However, that is not why it is taking so long for you to get a fix. You are experiencing another problem.
I located this post by stumbling on http://code.google.com/p/ardupilot-mega/issues/detail?id=660&st...
If I help solve your problem, would you consider closing or requesting that the ticket be closed?