Replies to This Discussion

Permalink Reply by Michael Müller on July 5, 2012 at 9:18am

NOT SOLVED

I still have that GPS-Lock problem like many others. The only reliable method to get GPS-lock is to connect USB.

My GFK-shield-solution worked for a few days. Many other solutions I tried worked for today but not tomorrow. All GPS related is not reproducable over more than one day.

Even when powering APM over external battery with UBEC on the out rail, nothing else connect, laying on a table in the garden with blue sky, I sometimes get GPS-Lock and sometimes not.

I (and I think many others) really need a reliable solution with this issue. An autopilot without GPS is absolutly useless.

Any suggestions?

• ▶ Reply

Permalink Reply by Remi on July 5, 2012 at 9:27am

Hi,

I added a printed circuit board connected to the ground below the APM and replaced every cable with shielded ones and it seems to be a little better. Yesterday I easily got GPS lock where I didn't get any before my mod. Today I added the 433 MHz telemetry kit I just received and it seems to not lock anymore althrough the 3DRadio is 12-13 cm away from the APM and I used shielded cables to connect the 3DRadio to the APM. However I didn't make a lot of test today because I couldn't get more than a few minutes tries before it starts to rain :( I will confirm as soon as the weather is better...

Remi

• ▶ Reply

Permalink Reply by Vernon Barry on July 6, 2012 at 4:19am

I (and I think many others) really need a reliable solution with this issue. An autopilot without GPS is absolutly useless.

I counter that the APM is not useless, it's your choise of external electronics and wiring choices. The APM is made from off the shelf components. The GPS module is completely standalone. What ever you are doing is preventing it from working. If it only works from USB and arguably clean power source, then your choice of ESC or BEC is the problem and should be added to the list of junk parts that do not work. You are generating RF and EMI noise and that's the problem. Solve the problem by replacing the offending device. You should be able to isolate to either the BEC or the ESC. Add a ferrite torrid to the connection wires from the BEC and the ESC or just replace them.

http://www.radicalrc.com/category/Ignition-Plugs-224

 

• ▶ Reply

Permalink Reply by Remi on July 10, 2012 at 4:11am

Hi,

I made some further tests today. I finally ended up to put the ESC at the end of the arms, near to the motors and this + shielded cables + PCB connected to GND below the APM makes perfect GPS locks in 1-2 minutes with 8 to 11 sats. I have several potential noise sources (receiver, 433 3DRadio plus the ESCs) but all of them are now 6-7 centimeters away from the APM at least. As it was working correctly like this I also removed the ubec, the APM beiing powered one ESC's integrated bec.

Hope this will help.

Best Regards,

Remi

• ▶ Reply

Permalink Reply by Remi on July 6, 2012 at 4:36am

Would you please explain how to ground the board? I think connecting a cable to some ground, but which one? 12V from LiPo or 5V from ESC/BEC or from the UBEC which is powering the APM?Thank you.

I connected it to the APM GND which is the same as the 5V one from BEC.

• ▶ Reply

Permalink Reply by Michael Müller on July 10, 2012 at 4:35am

Thank you Remi. I am glad that it worked for you, and it gives me hope

I have installed a backup battery to the GPS, but it is still not perfect. More here in this thread.

• ▶ Reply

Permalink Reply by R. D. Starwalt on July 10, 2012 at 4:49am

Is it possible to provide some photos of your set up and installation?

Is the system DIY (like mine) or did you purchase a 3DR kit?

Was the APM2 the soldered version or the kit version?

Is your GPS on the daughter board or separate (connected by cable)?

Do you have a voltmeter you can connect and measure the voltage applied to the APM2?

Are you using the 3DR power distribution board (PDB)?

Your problem certainly sounds like a power/cabling issue but we can only speculate without pictures or more information.

-=Doug

• ▶ Reply

Permalink Reply by Michael Müller on July 10, 2012 at 5:34am

Doug, thank you very much for you offered help.

Here is a photo of my PDB and one with the APM over it. Space between them is about 1".

- It´s a DIY-frame.

- APM2 is the soldered version with GPS on the daughter board

- yes I have a voltmeter. The voltage applied to the APM2 is 5,22V from a Turnigy UBEC

- the PDB is not the one from 3DR

Once again many thanks in advance.

• ▶ Reply

Permalink Reply by Remi on July 10, 2012 at 10:31am

I think your setup is like my first setup when I encountered the same problems as yours. Putting the rx and the ESCs far away from the APM greatly improved things for me, I think you should try. I think you should put the ubec far away as well (or remove it like I did) because it's a switching ubec that may cause interferences. I've got the best results as soon as *every* electronic component was at least 5cm from the APM.

• ▶ Reply

Permalink Reply by R. D. Starwalt on July 10, 2012 at 11:07am

Ah, a Hexcopter! Thank you for the photos!

I suggest another experiment: Simplify the test - move the APM, UBEC, battery off the copter, if possible.

Battery ---> UBEC ----> APM   on a separate board/card...if possible. This should work and also allow you to connect to the PC to the APM. This will also eliminate interference from the ESCs. Try to take the test outside the house/building/apartment (open sky) to give the GPS the best chance to sync/lock.

If success (repeat many times) add parts, one at a time, to observe interactions (if any).

If a problem still, try this:  MainsAC---[5VDC supply]----> APM. Observe the result. If good, add parts as above.

The copter is a complex system with many connections. As Remi suggests, interaction is possible. We need to simplify the system to reduce variables. When simple, you will know if you can trust the APM/UBEC combination.

The output of the UBEC seems good but a little high for the APM. Is it possible to reduce the UBEC output to 5.0VDC(some have USB connectivity)? Anything more is wasted energy.

Stay confident! You will resolve this problem and we will all learn from it. ;)

-=Doug

• ▶ Reply

• ‹ Previous
• 1
• 2
• Next ›
• Page