UPDATE: See the manual.
Hey guys, this is mainly a question for our friends at 3DR, but anyone who has looked at the traces on the APM2 daughterboard, and knows the GPS module, could offer ideas.
I wonder what you'd think about (as a hack, as a revision to the APM2 daughterboard, etc) installing some jumpers or switches to "disconnect" the onboard GPS, as needed. We might also add a battery socket to maintain the almanac, for those who want to accept the extra weight of the small battery, but that is another topic.
I have a small but growing fleet of APMs, some APM1 and some APM2s. However, I never have enough, so I find myself swapping them around as needed. I leave the magnetometers, air speed sensors in place and connect them as needed. Some airframes are better served with an external GPS, but I would prefer to have all my APM2s be, physically, the same. That is, I'm willing to buy all my APM2 with the extra cost of the onboard GPS whether or not I need it, in order to ease my "fleet management" if you will. If the daughterboard were not soldered in place, I might instead get extras, but this approach might work best.
However, I would also like to buy or use uBox and other external GPS units, as needed. I think this represents opportunities for the 3DR and affiliate stores. Many users will purchase the APM2 w/GPS, and some will later decide they want to use an external. Some will buy a separate unit w/o the GPS, but most will not. Most of the population simply live with the existing GPS. However, if there were a switch or jumper(s) to kill the onboard GPS, then casually swapping to an external GPS (and, casually purchasing additional GPS units) is easier. It also allows for subsequent upgrades (opportunities for sales) to future GPS units.
As a hack, are there recommendations on which trace(s) are best to cut and/or install a jumper for an existing APM2? I imagine there are a couple of ways to do this, not all are ideal. Is there any merit to a revision of the APM2 where there is a jumper or switch for the GPS?
The next rev of APM 2 will have that.
Does that mean that there is no way to make the GPS antenna separable from the current version? I was considering buying another mediatek from the diyd store and cabling it back to the tail so it'd stay away from my Frsky 2 way receiver.
Is this not possible on the current board?
The current board comes in two versions: with built-in GPS and with external GPS. If you buy the first kind, there's no easy way to disable to built-in GPS. Maybe in a future version we may enable that, but right now if you want to try different GPS modules choose the version with the external GPS.
No doubt as Chris says, in a later hardware release, all will be good. As for those of us now with the current hardware we own, posed with the need (for whatever reason) to implement external GPS, where do we turn? Below refers to another of your posts on the same subject
If I don't get an answer there, I will eventually start cutting traces on one of my boards to figure it out myself, and I will post something there. Either way, we'll have an answer there eventually.
Let me know when you go for the chop Mike, I'll go in sympathy with you. I have just been through 15 odd pages of the hardware forum and found 10 threads started on the same topic, for nil posted results. One member has apparently done it successfully but could only manage to post a thumbnail size picture of his attempt. I have familiarized myself with the Tx pin on the gps unit and feel confident enough to give it a snip.
In my case I will just need to make sure I have earth continuity between the other gps unit I have installed in the aircraft (for antenna tracking) and then piggyback the Tx from that unit onto the daughter board of APM2. Should that not work, I have a U-blox to try on the external gps header.
5 minute job............ ... :)
Any comments on the chance of success with this mod Chris? I figure if it all goes pear-shaped I'll have an excuse to buy the external gps version of APM2 right? :)
I have been considering cutting the trace that is closest to the "reset" button, which should be power. This way, the onboard GPS does not even power up, saving a tiny bit of juice. I suspect this will kill the 3D fix LED onboard, so you can easily see that this APM2 has a non-functional onboard GPS. Then, just plugging in an external GPS should work without additional changes.
However, there are a few things I'm not certain about yet -
* Will the onboard GPS unit "leach" power from the RX/TX lines, or other connections, and power up partially, or create other problems?
* There is nearly enough space to put in a jumper, however, I have not sorted out how to make it mechanically sound. The risk of a hacked in switch or jumper disconnecting during flight is too high, in any way I've considered so far...so once cut, the onboard GPS can only be used again with a very careful solder job.
Are there any thoughts on a better way to do this? I think this is easier and better than cutting the TX line, which does not appear to be accessible, largely because cutting the power reduces the draw on the battery (a tiny amount) kills the GPS Fix LED onboard (as a visual indicator) and is accessible to be cut without removing the soldered-in daughterboard.
I want to do this too. Could we solder a resistor between the enable pin (2) and ground do you think? This is probably the least intrusive first test.
My main concern would be the TX and RX lines still being connected to the disabled module...
I haven't seen any differences between the schematics and the board myself, but being a 4 layer board it can be tricky to spot. Did you see anything?
As far as implementation: I would either put a SMD resistor between EN pin 2 and GND pin 3 or a through hole resistor between EN pin 2 and one of the GND mounting posts (e.g. the one by the TX and RX pins).
The through hole would be messier, but has more potential for the jumper idea you have.
I can't find any information on the behaviour of the TX and RX lines in the disabled state on the datasheet. So I guess we just have to try it and see what happens?
I'm pretty sure the enable pin has been left floating, meaning that there is no trace to get to, you just have to solder directly to the pad. It's going to require some precision, but I reckon it's do-able.
I think the cut the trace method will work. For me, that's going to be a last resort. I think I'd prefer to try and de-solder the GPS chip first.
I don't see any harm in plugging a second GPS in to see what happens. If it does work I can imagine some very undefined behaviour though!
Hopefully someone might jump on here with a more concrete answer. If not I will see if the enable pin approach works in a few days time.
Yes, you're right, of course. For some reason I was under the impression that #2 and #3 were both connected to the same trace, but on looking closer, I see that #2/enable is not actually connected to that trace. It should be easy enough to test, even without soldering.
And so I have... here is what I found:
Using a 4kOhm resister (it was the first resister I found) to bridge #2 pin enable to a ground pin (on the input rail because it was easy; later on pin#3 of the GPS unit) does disable the GPS, partially. The 3DFix LED remains lit (and the GPS continues to report #ofSats ... in my case, 7) but the APM2 blue LED goes out, and the GPS status changes from "2" to "0".
If you boot up with the resister bridging #2 and #3, and run a GPS test, you get:
G !G !G !G !. G !G !G !G !. G !G !G !G !. G !G !G !G !. G !G !G !G !. G !G !G !G !. G !G !G !G !.
However, if you let it boot, then bridge, the GPS test repeats the same location data it had reported the moment you bridged (may be a code error... I suspect we get new data for the #ofSats and sat status is updated to 0, but the old location data is not flushed out of the variables in the test only.)
What concerns me, however, is the 3Dfix blue LED on the daughter board remains lit, which could lead to confusion. Although you will lose the blue LED on the APM2, that is only true when you have the external GPS disconnected, so it is not clear when you are running with one, or with the other. Thoughts?
Nice one Mike. This is looking promising.
I think the 3D fix blue LED on the daughter board is going to stay lit because of the way it's wired up with a transistor:
While waiting for a fix, the GPS module toggles pin 5 high and low. When a fix is found the pin stays low. A transistor has been used (T1) such that when the pin is high the LED goes out because R4 is shorted through T1 to ground; when low the transistor turns off and the LED lights.
With the module disabled, the transistor is not shorting R4 to ground and the LED lights. There isn't actually a 3D fix. Although confusing, I think this is an OK solution.
The sat count and position must be stale variables as you say.
Have you got an external GPS to plug in to see if the TX and RX lines are free for use whilst the on-board is disabled with the 4k bridging resistor?