Replies to This Discussion

Permalink Reply by Norman Sanchez on June 15, 2012 at 8:21pm

There is an idea I like, I'm on the possibility of placing two sensors Pressure in the skates, and as fit only in the auto landing.
it may adapt something?
http://www.tekscan.com/flexible-force-sensors#specifications

• ▶ Reply

Permalink Reply by John Campen on June 15, 2012 at 9:12pm

wow yes!! this could also initiate a power kill for not only landing but also for flip overs to reduce prop and motor damage. How would you suggest to connect it to the APM?

• ▶ Reply

DeveloperPermalink Reply by Marco Robustini on June 15, 2012 at 11:12pm

Autolanding with hexa/y6/octo/x8 is not recommended (imho), instead use autoapproach.
If the quad is pretty big bounces are obvious, autoapproach allows you to bring the quad in front of you and to land safely manually.
With Mikrokopter and Wookong the same thing happens, the only advantage is that these systems having amazing position hold and the descent does not make any drift, so the quad is difficult to flip at the ground because don't drift when touch.

• ▶ Reply

Permalink Reply by Derick Delica on June 16, 2012 at 12:35am

Is the Optical Flow Sensor already supported in 2.6?

• ▶ Reply

Permalink Reply by Gordon on June 16, 2012 at 12:37am

Been away for a while and just had my first flight with 2.6 on an old 1280. Planner 1.1.92

Great work team stabilize was impressive.

Noticed a couple of things though.

1. Everything up to 2.5.5, I could make fit by defining MTK16. This is the first time I have had to disable CLI, so I couldn't find a way to do an erase and reset in the planner. Had to reload 2.5.5, erase, then load 2.6 and do setup. Did I miss something obvious somewhere?

2. Now here is the weird one. With the Define GPS MTK16 line in place, the planner would only report 3D lock over USB. Over xbee it just said NO GPS. Removing the MTK16 line fixed it all up. I would like to still use the MTK16 in the Config.h for when the code starts to expand again in the future. Like a lot of people, I'm on a tight budget and would like to milk the 1280 as long as possible.

• ▶ Reply

Permalink Reply by Gustav Kuhn on June 16, 2012 at 6:18am

Hi Gordon,

Can you describe what you did to fit 1.6 on an APM1 1280?

I bit the bullet, and bought a 2560 board from Sparkfun, flew the quad around in a small garden, seems very stable, again :-)

I don't see any of the yaw issues some are reporting either.

• ▶ Reply

ModeratorPermalink Reply by John Church on June 16, 2012 at 8:28am

Hi Gustav,
The most direct way of squeezing it on the 1280 is to add #define CLI_ENABLED DISABLED to your apm_config.h .
Have fun,
John

• ▶ Reply

Permalink Reply by Gustav Kuhn on June 16, 2012 at 8:55am

Thanks John !

Logging, cam stab, does that take a lot of space, can it be disabled safely ?

1.6 seems to achieve most of the original goals aimed for, when I started with this hardware more than a year ago.

Huge waste, to just dump the present APM1......

• ▶ Reply

Permalink Reply by Cliff-E on June 16, 2012 at 10:26am

I run with logging disabled vs cli and clears enough space up to ac2.5.4. Then run wireless and always run mission planner etc to get some sort of log (as a tlog).

• ▶ Reply

Permalink Reply by Gordon on June 16, 2012 at 4:37pm

Hi Gustav,

John beat me to it whilst I was sleeping. No need to disable logging as this is done automatically when you select 1280 as the board in Arduino IDE. Another line for the config.h is #define GPS_PROTOCOL GPS_PROTOCOL_MTK16, saves another 5.9k.

Unfortunately it gives the problems I had above even though it compiles fine. I don't know when it was last working as I haven't flown since about 2.2. When I used this line in 2.5.5 I only got as far as doing setup over USB, never went flying which is why I didn't see it earlier.

It has been listed as issue #392 back on 16/3/2012 and looks like it affects ublox as well.

It would be nice if someone smarter than me could take a look at this as there is only 3.17k left on the 1280

• ▶ Reply

Permalink Reply by Martin Poller on June 17, 2012 at 5:27am

Hi Gordon and others,

I managed to squeeze 2.6 on to my 1280 yesterday, disabled  CLI, logging and defined MTK16, but had immediate problems getting the motors to arm.

Mission planner showed that rc was setup properly, had all channels connected and responding correctly, 6 modes saved and changing  on flight data screen.

But the rc connection seemed to be intermittent, and once the motors were armed the throttle responded very eratically, with the motors surging to full power, then stopping entirely at random.

Also although I had full GPS lock, the flight data screen seemed to be in delay, but did eventually show 3D lock. The red LED on the APM was not lit solid as it normally does when GPS lock is achieved.

Reverted back to software 1.0.49 and everything works fine again.

Does anyone have any ideas or suggestions please?

Kindest regards

Martin

• ▶ Reply

Permalink Reply by Gustav Kuhn on June 17, 2012 at 5:47am

Hi Martin,

I also tried to get 1.6 into the 1280 board, prior to this thread, no luck.

Neither with 1.5, which at the time tipped my hand to get a 2560 board from Sparkfun.

Just on two weeks later, after tax here, I had the board, for about $65.

Oh well, the 1280 board is going to be a superduper Arduino now :-)

• ▶ Reply

• ‹ Previous
• 1
• …
• 8
• 9
• 10
• 11
• 12
• …
• 68
• Next ›
• Page