Picloc 3x setting experience etc.

[achta]

Yes Yes, good copter, good camera and setup, and good team

 

[MombasaFlash]

CALLING GEORGE, CALLING GEORGE ....

When you come up for air (which of course you are not actually permitted to do under the terms of your isolation sentence ) please note the following and ... er ... fix it! Today :livid:!

I loaded v56.69 in place of .67 but the same issues persist with Pan and AUX1. Pan is wild and AUX1 is stuck in RTH mode. The RADIO tab shows that both of these channels have off-the-charts centre trim of around 1300us and show no yellow bar at all. They are far too off-centre to be brought back under the limited scope of Sub-Trim. The other three primary channels, Tilt, Roll and AUX2 are normal.

These two issues appeared with v56.67.
v56.43 did not have this.

Swap over to a 3X with HD55 and everything is normal.

 

[Remoteworks]

Reverted back to hd55 as well.
No Spektrum control with v56.67 and v56.69.


+1 MombasaFlash

Waiting for a usable combination of the relative stability of HD55 and the extra functionality of v56.


@George. Did you already sent a replacement for my broken one?

 

[EricH]

MSAflash: Just loaded 53.69 X3. I don't have the issue you experience w/ AUX 1 and PAN. I'm using composite PPM, not spectrum. Must be a bug when you use spectrum.
BTW, I gave some thought about the other problem you had w/ drifting roll and tilt and think it might have to do w/ the 10 turn pot fitted to the AV200. They just don't have enough resolution (I did experiment w/ 10turns at one point). I currently use one turn pots. That's pretty much the only difference left between my set up and your. Hope it helps.

E

 

[GeorgeM]

Argh !!!!>>>>>Finally...

I was for 1 complete day without an internet connection, and you can imagine how bad I felt that I couldn't upload the fixed file that Mombasa reported earlier on !!
Right now I have a 100mtr Cat5 to my friendly neighbor to be able to write this

Uploading the correction file v56.70 in 1 minute.

George

 

[krleas]

George i would be very happy if u could answer on my emails. Thank you.

 

[GeorgeM]

Frustration is somewhat diminished now.... Files v.56_70 are uploaded here.

Krleas... Emails ...Emails ...Emails.... I promise to try

Thankyou !!

 

[krleas]

Lanzar, that does look great...especially unprocessed. Too bad the cameraman wasn't looking at a monitor... ) Good job though!

Cameraman is Lanzar, pilot is Krleas.

 

[MombasaFlash]

cameraman is lanzar, pilot is krleas.

i knew it !! HA!

... Anyway, i am going to hazard a guess. You have bought the whole caboodle from kopterworx and krleas (in disguise wearing 'normal' coloured specs) has come over to give you a hand. I can even see the ginormous 25kg sj-8 flightcase in the garage...

 

[Remoteworks]

Hi George,

Is it correct that there's no hex file (v56.70=fixed RC5 inputs 3+4 and Spektrum DSM) attached on this link?

Kind regards,
RW

 

[GeorgeM]

Link is correct. Navigate to the firmware.

 

[Remoteworks]

Alright thnx.

RW

 

[graham_newton]

I have seemed to sort out problem of tilt and roll axis moving very slowly (5 seconds for the last 5 degrees, rough estimate witht a bubble level).
I set the accellerometer correction tab to low and there is no more problem. but what have I sacrified but putting this parameter to Low?
i.e. what exaclty does this parameter do?

Another thing with .70 FW... im not sure if I have not done something wrong but moving the ACC slider does not have any effect on the servo movement. It is at the same value that it was prior to the FW update.

We are getting there!

 

[GeorgeM]

The ACC slider gets more aggressive as you UP the Gyro filter...ie at 42 HZ(gyro), the Gyro is quite filtered, so quick changes are smoothed out... and the ACC has only small (change) values to work with,
the contrary to low Filter on the Gyro... at 188Hz, there is considerable to work with.

All of the above is only visible on the fastest servos possible. A standard servo will not show any difference.
To be able to show something on a normal-slow servo, the X3 firmware would need to be used.

I am working on some new parameters for near future. It's gonna be good !!

G.

 

[MombasaFlash]

... Is it correct that there's no hex file (v56.70=fixed RC5 inputs 3+4 and Spektrum DSM) attached on this link?...

George must have misunderstood your question.
The correct answer is that there is no hex file attached to that particular link. It is just information that the RC5 inputs 3+4 and Spektrum DSM have been fixed in v56.70.

I have seemed to sort out problem of tilt and roll axis moving very slowly (5 seconds for the last 5 degrees, rough estimate witht a bubble level).
I set the accellerometer correction tab to low and there is no more problem. but what have I sacrified but putting this parameter to Low?
i.e. what exaclty does this parameter do?

Another thing with .70 FW... im not sure if I have not done something wrong but moving the ACC slider does not have any effect on the servo movement. It is at the same value that it was prior to the FW update.

We are getting there!

Wrong. We are nowhere near yet.

But you seem to have fared better than me. Although the Pan and AUX1 channels have been restored to use with 56.70, nothing on Earth could stop all three axes from intermittently jerking. Just like it was with the first version of 56.43 that I tried. Every attempt to find the centre point for Tilt and Roll (forget about Pan, it still drifts at the rate of .25° per second) was abruptly ruined by one of the axes jerking off (no pun intended) its centre position by 2-4 degrees. Hitting RTH would invariably bring everything back to centre - except for Pan which had in the meantime shifted its centre to around the side somewhere. Do nothing and they just carry on jerking around tilting and rolling further and further away.

I tried setting Accelerometer sensitivity to Low, Default and High and none made any difference. I too experienced the slow down at the end of Tilt and Roll travel PLUS, they would both backtrack a few degrees before stopping moving. The magnetometer is as cursed as it ever was and regardless of starting position always chooses South as Home before continuing on its merry spin. This despite being recalibrated specifically after selecting Low Sensitivity.

I tried upping the Gyro Bandwidth every step eventually all the way to 256Hz. No changes. Changing Servo Refresh Rates made no difference. 3X stabilisation ATV's continue to do absolutely nothing. Tx ATV's still work.

Acceleration slider under GAINS continues to have no effect on the servo speed when compensating. However, as before, the reaction times are adequate anyway, even with ZERO Acceleration.

Additional fun was that, after making some parameter changes and hitting Save to EEprom, the poxy Tilt axis would occasionally catch me off-guard and suddenly spin round prompting a mad panic to yank out the USB cable and switch off the power before even MORE servo cables got torn up.

So, it is back to HD55. 3X ATV's, compass and Pan don't work on that either but at least the camera plate stays relatively still.

I am getting right tired of wasting days on this.

 

[MombasaFlash]

The ACC slider gets more aggressive as you UP the Gyro filter...ie at 42 HZ(gyro), the Gyro is quite filtered, so quick changes are smoothed out... and the ACC has only small (change) values to work with,
the contrary to low Filter on the Gyro... at 188Hz, there is considerable to work with.

All of the above is only visible on the fastest servos possible. A standard servo will not show any difference.
To be able to show something on a normal-slow servo, the X3 firmware would need to be used.

I am working on some new parameters for near future. It's gonna be good !!

G.

Please fix the basics first.

 

[graham_newton]

AHA! I have figured out the drifting issue on the roll and tilt. I do have too much vibes getting to the picloc.
I by increasing the gyro deadband the gyro does not pick up on the vibes coming from the lifting mechanism and the mount does not drift.
But if it is increased too much then the picloc does not pic up slow movement of the actuall gimbal and does not return to center when the gimbal is level.(if it is brought slowly back to level)

Mombasa I am getting the same effects :
slow down at the end of Tilt and Roll travel PLUS, they would both backtrack a few degrees before stopping moving.
with the high settings. Again, what exacly does the parameter do?

SO im guessing for systems witht more vibes one wants to UP the gyro filtering and have it at 42Hz?

Time for some gyro gel!

 

[Remoteworks]

George must have misunderstood your question.
The correct answer is that there is no hex file attached to that particular link. It is just information that the RC5 inputs 3+4 and Spektrum DSM have been fixed in v56.70.

Thnx. i figured by his respond!
Also there's no other way to put up some info regarding the new firmware.

RW

 

[MombasaFlash]

... with the high settings. Again, what exacly does the parameter do?
Time for some gyro gel!

There is, theoretically at least, an optimum balance between the gyro and accelerometer sensor input. A gyro on its own will compensate for axis movement but forgets where it came from and centers in a new place. This is why tail gyros are not enough. The accelerometers are there to keep tabs on that central starting point. But with too much accelerometer and not enough gyro input the axis will be over-reactive and too sensitive to other forces - like vibration for example. Too little accelerometer input and the system as a whole is not sensitive or precise enough.

But as George has said, the accelerometers are very sensitive items so the percentage required in the mix does not need to be too high.

Am I the only one with this ... erm ... jerking problem? I have not progressed beyond table settings with v.56 and I still have all of this un-commanded movement that certainly has nothing to do with vibrations.

Last point. Don't get too enthusiastic with your gyro gel too early. Picloc will eat your supply with alarming efficiency. Every time PL has to be updated - currently once every four hours - and be reset, magnetometer calibration re-checked, gyro and accelerometer calibration check etc, etc, another strip of expensive and sometimes very difficult to obtain gyro gel goes in the bin.

I shall stay with good ole velcro until the bugs are ironed out. Which probably means I shall stay with good ole velcro.

 

[yeehaanow]

Just put the gel on a plate that the PL can be removed from easily with a strap or zip-tie. Mine is hard mounted to the gimbal and doing just fine though.