Page 1 of 1

DragonOSD+ GPS modules

Posted: Thu 28 Jan 28 2010 6:32 pm
by Daniel Wee
At the point of writing, there are only three viable choices of GPS families on the market that offer update rates of greater than 1Hz and still reasonably affordable. The three families are MediaTek (MTK), Antaris (u-blox) and SkyTrax (Venus). The most common of these are the MTK based modules, which include models such as LS20033, LS20031, EB-85A, and other. Many companies offer GPS'es based on the MTK chipset. They were the first to offer economical 5Hz modules and have recently updated their units to 10Hz capability. ... ht=locosys ... ts_id=8266 ... ts_id=8975 ... 250&page=1

The next chipset, the Antaris, comes from Atmel, and offered up to 4Hz capability although it has been mentioned that some of the newer ones might even handle up to 10Hz. The u-blox modules uses this chipset and has been noted for its good accuracy as compared to the MTK chipset. GPS modules such as the GS407 would be an example of such a product. ... ts_id=9436 ... ts_id=9566 ... -GPS-GS407

The SkyTrax offering uses the Venus chipset and is capable of 10Hz update rates. This is a rather new module and it is uncertain at this point if there will be many modules based on this chipset. Sparkfun does, however, offer two boards that use the chipset and they appear to work quite well when paired with an active antenna. ... ts_id=9758 ... ts_id=9133 ... ts_id=9171

Unfortunately, the previously very popular SiRF III chipset still does not support more than 1Hz update rates and so they would not be particularly suitable for use with GPS based autopilots. This would include modules such as the EM-406.
GPS attachments 005a.JPG
GPS attachments 005a.JPG (73.47 KiB) Viewed 8575 times

Re: DragonOSD+ GPS modules

Posted: Thu 28 Jan 28 2010 8:08 pm
by Daniel Wee
The GPS modules will need to be connected to the DragonOSD+ using suitable cable connections. Do NOT assume that the default connector pin-outs will match the pin-out required by the OSD board. There are 4-primary connections (+3.3V, Gnd, RX, and TX). Note that the DragonOSD only supplies 3.3V so the GPS modules need to be able to operate on 3.3V (almost all the newer ones and the listed ones do.)
GPS attachments 007a.JPG
GPS attachments 007a.JPG (70.15 KiB) Viewed 8575 times
GPS attachments 008a.JPG
GPS attachments 008a.JPG (76.29 KiB) Viewed 8575 times
The extra pin (last on the right in the photo) is a diode blocked line to charge a supercap (if there is one) and is sometimes unneeded.

The DragonOSD+ will typically attempt to auto-detect the module baud rate, and then the module type. If this succeeds, normal operation will continue. However, if the auto-detect phase fails, it will fall back on the GPS choice selected in the configuration menu. Under normal circumstances though, this should not fail and if it does, it probably indicates a more complex problem at hand. At the point of writing, the DragonOSD+ will not recognize the SiRF chipset to discourage its' use.


Re: DragonOSD+ GPS modules

Posted: Sun 08 Aug 08 2010 9:46 pm
by Daniel Wee
  • Locosys LS20031 - recommended
    Flytron FGPMMOPA6B - not recommended

GPS takes a long time to get a position fix or lock

Posted: Tue 21 Dec 21 2010 9:32 pm
by Daniel Wee
There are two things you need to be aware of here:-

1. The first time you run the GPS, it will take a much longer time to get the initial lock because it doesn't have your local information. This information is usually downloaded from the satellites at about 50bps (50 bits a second, or about 5-bytes a second if you factor in some overheads). It needs to download the entire epheremis data set and that can take a long time. Some people have left the unit out there for over 30-minutes on the first lock. Thereafter, it should get a position fix much quicker if the GPS has a battery backup in it (which the LS20031 does.)

*for more info:

2. GPS can be jammed by interference as well. This could be from the camera or any number of devices. I assume that when you leave the GPS out there, your camera is active or you would not be getting a screen to see at all. Be advised that testing has shown cameras to be a possible source of GPS interference. If you disconnected the camera for your test, then you must have had the PC or laptop attached, which could be generating even more GPS interference. GPS operates on 1575MHz (1.575GHz) so anything near that frequency (such as VTX or the laptops CPU running at 1.2GHz or 1.5GHz) could prevent normal operation of the GPS.

The GPS runs off 3.3V. One thing you can try, if you are good with wires and soldering, is that you can simply connect 2 AA batteries in series and power just the GPS with those batteries and leave it out in the open. The GPS will flash once a second if it has attained a lock. The LED is normally dark until it gets a lock.

Putting a ferrite choke on the GPS wire is probably one of the things you can do to keep interference out of your GPS. If you camera is on - chances are it's causing the problem. Many people under-estimate the problems coming from cameras simply because it is not often talked about. I have done extensive testing and you can read the report here:-


Assuming that your GPS is flashing, that's a good sign. In this case, another parameter can affect HOME setting in the DOSD+, the HDOP parameter. Once a lock is attained, the accuracy takes a while to settle. Initial position locks can be off by tens of meters. The DOSD+ can wait until the accuracy is within the desired range before accepting the positions from the GPS. This is done using the HDOP (Horizontal Dilution of Precision) setting in the menu. Setting a higher value gets you a home fix more quickly but it will be less accurate. Setting too low a value may mean that you could never get a lock if your GPS reading cannot reach that level of precision. The default values generally work quite well in this regard.