new SiK stable release - version 1.9

[Andrew Tridgell]

Kevin and I have done a new release of the SiK firmware for 3DR radios
(and RFD900) here:

http://firmware.diydrones.com/SiK/stable/

This release includes the low latency work by Kevin, plus a number of
smaller bug fixes.

Kevin, during the final testing before release it initially failed the
mavtester.py tests, showing quite a lot of packet loss. I spent some
time looking into it and found two bugs in pymavlink that caused the
issue. I've pushed out a new pymavlink release as well with the fixes.

As part of the investigation I added some new packet loss detection in
mavtester.py, and also added support for running it over UDP:

./tools/mavtester.py --port1 127.0.0.1:14550 --port2 127.0.0.1:14550 --rate 4

it also now supports the --rtscts flag, which allows you to test RTSCTS
functionality.

Cheers, Tridge

[Kevin Hester]

cool packet loss detection!

--
You received this message because you are subscribed to the Google Groups "drones-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to drones-discus...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Roberto Navoni]

Great Job
Tridge :)
Best
Roberto

[Andrew Tridgell]

Hi Kevin,

> cool packet loss detection!

I've now pushed some more improvements to mavtester.py, using threads
for receive to avoid overflows in the Linux serial stack.

The result shows that we get zero packet loss even at very high demanded
rates with the --rtscts option, both with and without MAVLINK=1 (that
means we get zero loss for straight binary data, as MAVLINK=0 means the
radios treat the packets as a byte stream).

For example:

tools/mavtester.py --port1 /dev/ttyUSB2 --port2 /dev/ttyUSB1 --rate 20 --override-rate=30 --rtscts

shows zero packet loss over extended tests.

Without --rtscts we get zero loss as long as we keep the rates low
enough not to overflow the radios buffers. That is expected.

Cheers, Tridge

[Marco Robustini]

Thanks Kevin and Tridge for your work, great news!

Marco

[britdavis]

Hi Tridge,

 

Just want to add that I got the 2.4 MP working on my 3DR 915s. Using node 0 for my Pixhawk, node 1 for my Nexus 7, and node 2 for my Laptop. Having MP allows me to not have to disconnect when I'm using both. Have you worked with it yet?

[MB]

Hi Tridge,

thanks for your work. It is working great with my 868S Radio.

Cheers
Marco

[Marco Robustini]

Doeas anyone have really improvement with 1.9 firmware?
I've a lot of lag/retry during the first connection, long time and many retry, i've only upgraded my radio, unchanged parameters.

Bests, Marco

On Tuesday, March 25, 2014 12:21:04 PM UTC+1, Andrew Tridgell wrote:

[Emile Castelnuovo]

There are some issues with the new firmware 1.9 and the 3DR radio config tool.

I have 433 module, and if I reset the parameters i get 

Min freq 433050

Max freq 434750

Although when I modify any parameter, and then save I see in the AT commands that the min and max freq gets written 414000 and 415000 respectively.

If I look in the freq drop down list I find no reference to the 433050 and  434750 frequencies, so I suppose the 3DR radio config (ver. 1.3) is writing wrong parameters.

But If I change by hand the two params, they get written correctly.

Best,

Emile

--

[Marco Robustini]

Hi Emile, have you tried with the 3DR Radio Tool inside the Planner?
One month a go i've opposite problem (LOL), this issue with the Planner and not with the config tool.

Marco

[Emile Castelnuovo]

Yes Marco, and the problem seems the same.

If I setup defaults for radios everything is fine, but as soon as I update one parameter (eg. ID), the parameters are scrambled.

Is there any place I can open an issue?

Can somebody check the 433MHz radio setup in planner and 3DR radio config tool?

Beware that if the min and max frequency are updated incorrectly (although they are shared between radios) the link is poor and can be only a few meters before it disconnects. I suppose this is due to the fact that the antennas are good for a definite range of frequencies.

Cheers,

Emile

[Michael Oborne]

What verison of the standalone config tool are you running? There was a new version release a while ago that fixed any frequency change issues.

 

Michael

[Emile Castelnuovo]

Version 1.3 downloaded from the ardupilot.com website 6 days ago.

http://ardupilot.com/downloads/?did=89

I think it is still the December version.

Thanks,

Emile

[Emile Castelnuovo]

Attached the log of the 3DR standalone client.

If you look at it you will see that frequencies gets set to 414000 as min and max.

After reset of the parameters I get these values:

S8:MIN_FREQ=433050

S9:MAX_FREQ=434790

But as soon as I change any value i get this in the log:

2014-04-12 21:51:29,646  INFO MissionPlanner._3DRradio - Connect btr 4 baud 57600 (:0) [1]

2014-04-12 21:51:29,646  INFO MissionPlanner._3DRradio - Connect first responce OK

...

2014-04-12 21:51:31,517  INFO MissionPlanner._3DRradio - responce 0 SiK 1.9 on HM-TRP

 (:0) [1]

2014-04-12 21:51:31,517  INFO MissionPlanner._3DRradio - Connect Version: SiK 1.9 on HM-TRP

 (:0) [1]

...

2014-04-12 21:51:34,141  INFO MissionPlanner._3DRradio - responce 0 S0:FORMAT=25

S1:SERIAL_SPEED=57

S2:AIR_SPEED=64

S3:NETID=25

S4:TXPOWER=20

S5:ECC=0

S6:MAVLINK=1

S7:OPPRESEND=1

S8:MIN_FREQ=433050

S9:MAX_FREQ=434750

S10:NUM_CHANNELS=50

S11:DUTY_CYCLE=100

S12:LBT_RSSI=0

S13:MANCHESTER=0

S14:RTSCTS=0

S15:MAX_WINDOW=131

 (:0) [1]

2014-04-12 21:51:34,242  INFO MissionPlanner._3DRradio - Doing Command ATS8=414000

 (:0) [1]

2014-04-12 21:51:34,251  INFO MissionPlanner._3DRradio - cmd ATS8=414000

 echo ATS8=414000

 (:0) [1]

2014-04-12 21:51:34,801  INFO MissionPlanner._3DRradio - responce 0 OK

 (:0) [1]

2014-04-12 21:51:34,901  INFO MissionPlanner._3DRradio - Doing Command ATS9=414000

 (:0) [1]

2014-04-12 21:51:34,908  INFO MissionPlanner._3DRradio - cmd ATS9=414000

 echo ATS9=414000

 (:0) [1]

2014-04-12 21:51:35,458  INFO MissionPlanner._3DRradio - responce 0 OK

 (:0) [1]

2014-04-12 21:51:35,558  INFO MissionPlanner._3DRradio - Doing Command ATS10=10

 (:0) [1]

2014-04-12 21:51:35,564  INFO MissionPlanner._3DRradio - cmd ATS10=10

 echo ATS10=10

 (:0) [1]

2014-04-12 21:51:36,164  INFO MissionPlanner._3DRradio - responce 0 OK

...

2014-04-12 21:51:47,414  INFO MissionPlanner._3DRradio - responce 0 SiK 1.9 on HM-TRP

 (:0) [1]

2014-04-12 21:51:47,414  INFO MissionPlanner._3DRradio - Connect Version: SiK 1.9 on HM-TRP

 (:0) [1]

...

2014-04-12 21:51:55,225  INFO MissionPlanner._3DRradio - responce 0 S0:FORMAT=25

S1:SERIAL_SPEED=57

S2:AIR_SPEED=64

S3:NETID=25

S4:TXPOWER=20

S5:ECC=0

S6:MAVLINK=1

S7:OPPRESEND=1

S8:MIN_FREQ=414000

S9:MAX_FREQ=415000

S10:NUM_CHANNELS=10

S11:DUTY_CYCLE=100

S12:LBT_RSSI=0

S13:MANCHESTER=0

S14:RTSCTS=0

S15:MAX_WINDOW=131

Note that min and max get set to 414000 both, but then are read 414000 and 415000 repectively.

Cheers,
Emile

[Emile Castelnuovo]

And the attachment.

:)

Emile

[Monroe King]

Great work Kevin & Tridge! Thank you, anything that increases data flow at range is much appreciated.  Now to play with it! :)

[Michael Oborne]

Please use the version in MP. As it will be more uptodate. The logging should show very similar.

[Maurice Barnes]

Michael I am having the same issue using MP. When I make a change it gives errors in updating the parameters to the remote radio it appears and then the remote gets written with incorrect data including Netid at times.  I then have to connect to the "remote" locally to fix the problem. I will get some logs later.

Regards,
Maurice