Replies to This Discussion

DeveloperPermalink Reply by Marco Robustini on July 26, 2012 at 11:52pm

Yeah Peter, i've found same issue during beta testing and I immediately reported to the development team with accurate logs but being an isolated problem of a few was not taken into account (for now).
But how often it has happened many problems reported then became public knowledge when he released a new release, i'ts normal.
I still have these "randomly surging"......... :-(

• ▶ Reply

Permalink Reply by Peter Chaffe on July 27, 2012 at 3:49am

Hi Guys,
Will try to answer as many questions as possible.

Rob,
It happens here immediately(ie. in Stabilize) reacts like too high gyro gain, but observable while firmly on the ground well below take-off rpm which I though would have ruled out gyros and their associated settings. It is perfect on V2.6


Marco,
I've Uploaded this clip which explains the problem better, Listen to the sound track carefully. it is bad enough to be immediatly noticible, especially after the smooth response with V2.6.
http://youtu.be/xPmZ3clVzcc


Randy,
My dataFlash has not behaved well since V2.6, I erased the dataFlash before the video test above(using cli), and enabled MOTORS in the cli(using 'enable MOTORS'). But can't see the results here, CTUN shows Throttle_in. Should I be looking at Angle_boost!

Now I see that although I erased the dataFlash before the test flight. Have just downloaded card now and it shows a KMZ file from sometime last week, (about 4 sessions ago...  Obviously I need to get to grips with this dataFlash card...

• ▶ Reply

Permalink Reply by Peter Chaffe on July 27, 2012 at 9:09am

Just been out for a more lengthy test, quite windy, Transition from Stabilize to alt_hold seamless, In alt_hold  good height stability both on baro and sonar,(possibly better than V2.6).

Switch to Loiter, again seamless,  was drifting down wind about 15 ft, then creeping back in very slow cycles of about 1 minute duration (for windy conditions this is the best I've managed)

Switch back to Stabilize again seamless.

Not a suitable place to perform a mission but overall view is that V2.7 is a definite improvement over 2.6. For my frame I needed to adjust throttle end points, but consider that just frame/motor tuning.

Alas the motor surging was evident throughout Stab, alt_hold, and loiter. Spoiling an otherwise good result...

• ▶ Reply

DeveloperPermalink Reply by Marco Robustini on July 27, 2012 at 11:39pm

If you arm the motors in Acro and takeoff there's this problem?
Try to reduce a lot RATE_P and give me your report.
My problem is another, not this one, i don't have this "throttle issue".

• ▶ Reply

DeveloperPermalink Reply by Marco Robustini on July 28, 2012 at 4:53am

Peter, backup your parameters, install version 2.55 with Arduino, then reflash 2.7, reset the board form CLI, reload your parameters and keep me informed about the result.

• ▶ Reply

Permalink Reply by Peter Chaffe on July 28, 2012 at 10:07am

Hi Marco,

Quote:< If you arm the motors in Acro and takeoff there's this problem? >:unquote

Armed motors in acro mode, Motors are still surging, in fact the surge is slightly more violent than in Stab mode.
---------------------

Quote:> Try to reduce a lot RATE_P and give me your report.>:unquote

OK starting default 0.18 changed to 0.1 = much the same.

changed to 0.05 = becoming more difficult to handle, motors are still surging, on ground and while hovering, and causing the frame to jump more during surges.

Changed to 0.00 now impossible to fly, motor surge greatly reduced, but still detectable.

above settings tested in both stab and acro, each mode being set before arming motors.(same results in both modes)


A peculiar thing exists with rate_p = 0.0 In STAB mode I just increase throttle enough to start motors turning minimum rpm, then leave throttle alone. motor rpms increase almost to take-off rpm, this took about 1 minute, then rpm started reducing all the way down to minimum rpm, this again took about 1 minute, it then continued cycling in this fashion!

I then tried the same in acro mode. Just increase throttle enough to start motors turning minimum rpm, then leave throttle alone. motor rpms increase to full throttle, this took about 1 minute, I didn't wait to see if this one cycles!

I then tried to repeat the problem, no good. Then found this:

Switch to Stab, enable motors, take rpm up to minimum rpm = stable rpm(might increase slightly)
Now pulse the motors a few times as if trying to take off, then throttle 0 to minimum rpm, then the cycling will start

Switch to acro, enable motors, take rpm up to minimum rpm = stable rpm(might increase slightly)
Now pulse the motors a few times as if trying to take off, then throttle 0 to minimum rpm, then the increase to full rpm will start.

Hope this sheds some light.

Will do the reflash next...

Permalink Reply by Peter Chaffe on July 28, 2012 at 11:00am

OK flashed V2.55 erased eeprom, and factory reset.

Flashed V2.7 done the full configuration.

All settings V2.7 default.

Motors are still surging, perhaps slightly less on the ground, but could be my imagination, in the air the motor pulsing is still observable.

The difference this time is that I've left the throttle endpoints default and throttle (so far) is fine.(I had lots of problems with this in 2.6)

Needs a good test flight to be sure though...

Will probably play with the config next, see if I can reduce the pulsing!

Need to get this sorted before the 3dr uBlox arrives...

Permalink Reply by Artur on July 26, 2012 at 1:12pm

Hi,

I have Hexa U. Will anyone know how to write a mixer to the frame?

B.R.

Artur 

• ▶ Reply

Permalink Reply by Dani Saez on July 28, 2012 at 10:31am

Hi Arthur, some time ago, I wrote this "how to", for motor mixing. You can start here: http://code.google.com/p/ardupirates/wiki/motormixing

Dani

• ▶ Reply

Permalink Reply by Denis on July 26, 2012 at 3:38pm

I have tested Arductopter 2.7 with my old tuning.

In mode Alt Hold, the copter was stable for some time and then begin to oscillate. I must lower Alt Rate (THR_RATE_P).

In loiter mode, it circle with a diameter the is bigger. Which value I must lower?

Thanks.

• ▶ Reply

Permalink Reply by TheSlanderer on July 26, 2012 at 4:05pm

I tested this yesterday, and noticed that sometimes that on powerup, my ESCs were beeping, and none of the status LEDs on the APM 2.0 are on. Multiple attempts at power cycling required to fix this--no idea why. Anyone else see this?

Also, although this is the first version I'm testing it on, optical loiter isn't working at all--regular loiter is fine, but optical flow just causes it to begin to circle at a pretty fast rate.

Lastly, like with 2.6, the default PID values are still kind of twitchy..when I was last using it back in May, the defaults were a lot more stable. Haven't gotten time to retune.

There were a few other wonky things I noticed, but nothing worth reporting. Good just overall, though.

• ▶ Reply

Permalink Reply by Glenn M on July 26, 2012 at 7:06pm

I've just noticed that I'm not getting logs for mag heading now. It's just coming through as 0, and also my base altitude seems to be off by about +55 meters. However on my minimOSD altitude displays correctly while flying.  I'm on APM2.

Suggestions?

Attachments:

• 2012-07-26 17-22 120.log, 211 KB

• ▶ Reply

• ‹ Previous
• 1
• …
• 4
• 5
• 6
• 7
• 8
• …
• 18
• Next ›
• Page