Replies to This Discussion

Permalink Reply by Michael Pursifull on October 10, 2011 at 2:45am

Yes; I have no personal experience with this myself, I have had some non-standard ArduPilot Digi gear on order for a month, but it is not here yet. But I have read about this problem with the extended range and also with series 2 modules.

 

What I understand from the reading is that maybe these other modules do not handle streaming data well. The use case that they were built for is more like brief bursts. And the APM knows nothing about the radios and how best to send data.

 

I have a lot of questions about this, and there are a lot of things I do not understand about it, but that is all I know. It might help you to find out more information, and maybe that will either help you find someone that knows the truth, or learn enough to investigate (and solve?) it yourself, maybe.

 

Digi tells me the modules I bought will ship on 10/10/2011 ;) So I can feel your pain (different modules) soon, I hope...

• ▶ Reply

Permalink Reply by jim C. on October 10, 2011 at 6:24am

Hi tjhowse , I also I have a pair of XT09, and supporting evidence; been doing and work well at close range, but when I go more than 10 meters is no loss of signal (the XT09 are configured with power 1watts and 57000 bps) .. and could not find the solution.

which pins are you using for the connection of the XT09?

• ▶ Reply

Permalink Reply by tjhowse on October 10, 2011 at 6:28am

DI, DO, GND and VCC with SHDN tied to VCC, plugged into the telemetry header on the front-left of the oilpan.

I have found no correlation between range and performance. It's equally bad at all distances.

• ▶ Reply

Permalink Reply by jim C. on October 10, 2011 at 12:23pm

I just used this configuration in the file APM_Config.h
# define SERIAL3_BAUD 9600
/ / # Define SERIAL3_BAUD 19200
/ / # Define SERIAL3_BAUD 38400
/ / # Define SERIAL3_BAUD 57600
/ / # Define SERIAL3_BAUD 115200

seems to work to put all the options

to operate the XT09 is an Obsession turned to me   :) ......

 

regards.....

 

• ▶ Reply

Permalink Reply by tjhowse on October 12, 2011 at 2:14am

Which firmware does everyone use on their XT09-SIs? I've got the choice of:

9XTEND 115K HOPPING (DIGIMESH - AUSTRALIA)

9XTEND 115K HOPPING (DIGIMESH)

9XTEND 9600/115K HOPPING (DIGIMESH)

9XTEND AUSTRALIA

9XTEND-PKG 9600/115K HOPPING

 

I've tried them all, with varying levels of failure. Which do people generally use?

 

• ▶ Reply

Permalink Reply by tjhowse on October 12, 2011 at 5:39am

It should also be noted that the radios range test fine. 2000 successes, 16 fails. 99.2%, using the range test utility in X-CTU.

• ▶ Reply

Permalink Reply by Scott Plunkett on October 12, 2011 at 6:52am

Have you tried any speed higher than the 57600? - It could be a buffer problem.  The problem people have with the XSC (Extended Range) XBees is that they top out at 9600.  I ran my 9Xtend radios at 115200, using the 9XTEND-PKG 9600/115K HOPPING radio firmware - but I must say that I have not tried it with any of the current generation GCS, I was only using it for telemetry to a data logger.  I have since then switched to a pair of microhard ehternet radios that I am attempting to get working.  You were likely never getting the 1W of power out of the radios- they require 730mA at 5V to push out the 1W.  I ended up creating my own power board based on Sparkfun's breakout board to give the radio a clean and separate power supply (running off of the same battery but with its own 5V1A LDO Power Regulator.)  Since we are not using flow control, you need to make sure the DI (Data In) and DO (Data Out) buffers don't overflow.  Make absolutely certain that the BD and BR (baud and rf data rates) are the SAME (BD set to 7 and BR set to 1)  If this is not the case you will get dropouts because when the buffer gets to a certain size a request (stop sending) is passed between the radios.  These radios really prefer to be run with flow control turned on, and that is why most people are having an issue.  The default setting is BD / BR set to 3 and 1 which works fine for small bursts of data, but matching them to 7 and 1 works MUCH better for the longer streams we want to exchange.  Give that a shot and let me know how it goes.  Feel free to PM me if you need more help.

 

 

• ▶ Reply

Permalink Reply by tjhowse on October 13, 2011 at 12:30am

Thanks for the tip. I changed to that firmware and pegged everything to 115200 and things have improved slightly. I think it's because the interruptions to the signal occur every X seconds, and the faster the baud rate the more data it can squeeze through in that window. It's still not good enough to read out all of the parameters without having to start over three or four times.

Is the telemetry you're sending down mavlink?

• ▶ Reply

Permalink Reply by Scott Plunkett on October 13, 2011 at 2:57pm

I was only sending a custom serial string made up of most of the same variables, but I had also incorporated some other information.  Keep in mind this was on the old ArduPilot (pre Mega.)  Do you have any idea how long the string is?  You are probably correct that you are just able to cram more down the pipe at the higher rate, but what puzzles me is the regularity of the interruptions.  This may seem overly simplistic, but did you check the data and rf rates on both radios?  Starting over three or four times just won't work.

• ▶ Reply

Permalink Reply by tjhowse on October 13, 2011 at 11:24pm

Yep, I have exactly the same settings on both modems.

I'm beginning to suspect one of the modems is faulty.

• ▶ Reply

Permalink Reply by jim C. on October 13, 2011 at 3:25pm

I tested this configuration and seems to work 9XTEND 9600/115K HOPPING (DIGIMESH) 206E version, to 115200, 1 watt.

note that 1 watt external power is required 5v 750 mah

• ▶ Reply

Permalink Reply by tjhowse on October 13, 2011 at 11:24pm

I'm only running at 10mW.

• ▶ Reply

• ‹ Previous
• 1
• 2
• Next ›
• Page