Tuning soft frame

Info and discussion about the autonomous Micro FC, suitable for nanos, micros and mini multirotors

Re: Tuning soft frame

Postby Patnet » Mon Oct 26, 2015 5:19 pm

LPR wrote:
# Onboard parameters for system MAV 147

CTRL_TLT_ANG_F 0.125

CTRL_TLT_RTE_D 11850

Larry


Thanks Larry, what does the red and blue text mean? Too high, go higher, go lower? Which direction?
Patnet
 
Posts: 138
Joined: Thu Jan 29, 2015 10:24 pm

Re: Tuning soft frame

Postby LPR » Mon Oct 26, 2015 7:05 pm

Pat

The fact that your getting motor clipping means you need to change more than just parameters. The higher voltage of a 4s lipo would get you to fly much better or you could reduce the weight of your quad.

How much does your quad weigh now?

You may have motor clipping because of how your ESCs a calibrated or the settings of the M4. Clipping means your motor are run at near full throttle just to get off the ground, so not much is left for control of the quad. It's best to not go over 50% of full thrust in order to hover.

The blue number is probably a good number to start with. The red number will probably need to be reduced. None if this will help your major problem, motor clipping.

One of the developers on this forum works for Airbus and he told me that big planes are easy to build but quadcopters are hard.

Larry
LPR
 
Posts: 1323
Joined: Tue Jun 26, 2012 1:29 pm
Location: MN, USA

Re: Tuning soft frame

Postby Patnet » Mon Oct 26, 2015 10:24 pm

Hi Larry, thanks. It's 1.67 kilos now. If I add a 578 gram 4S battery and go to 5 amps in lieu of the 3.3 amp 272g 3S, will the all-in-weight not take a hit? Sure that's worth it?

http://hobbyking.com/hobbyking/store/__ ... _Pack.html
Patnet
 
Posts: 138
Joined: Thu Jan 29, 2015 10:24 pm

Re: Tuning soft frame

Postby LPR » Tue Oct 27, 2015 4:02 am

Pat

Looking at your motor log again I think the 3s lipo will be fine. The clipping is likely caused by vibrations and lowering the
CTRL_TLT_ANG_F to 0.125 and reducing the CTRL_TLT_RTE_D to around 1500 instead of 11850 should reduce vibration.

Joerg will know more about those parameters.

Larry
LPR
 
Posts: 1323
Joined: Tue Jun 26, 2012 1:29 pm
Location: MN, USA

Re: Tuning soft frame

Postby aBUGSworstnightmare » Tue Oct 27, 2015 7:43 am

Patnet wrote:Hi Larry, thanks. It's 1.67 kilos now. If I add a 578 gram 4S battery and go to 5 amps in lieu of the 3.3 amp 272g 3S, will the all-in-weight not take a hit? Sure that's worth it?

http://hobbyking.com/hobbyking/store/__ ... _Pack.html


So, it's 1.67kg incl. the 3S Lipo, right? So new AUW with 4S would be 1.67kg - 272g + 578g = 1976g

Don't know the performance of your motor+prop so let me assume they are equal to DJI:
ecalc_dji.jpg


Better opt for a lighter 4S battery - like i.e. http://hobbyking.com/hobbyking/store/__ ... _Pack.html - and buy some larger props - i.e. http://www.rctigermotor.com/html/2014/P ... 0/282.html - not much larger, but should give you a little more thrust on theses small motors.

Your resulting AUW will be around 1.8kg with this pack which should give you enough control authority and a good performing craft.

By the way: tested my setup (AUW 1.4kg for 3S!) with T9545 --> performance was very poor; you don't want to fly such a craft in windy conditions. With 4S (AUW with 4S6000mAh Nanotech pack) it shows an awesome performance.
I'm pretty sure you know the video...
aBUGSworstnightmare
 
Posts: 1460
Joined: Fri Jun 22, 2012 5:24 pm

Re: Tuning soft frame

Postby aBUGSworstnightmare » Tue Oct 27, 2015 1:38 pm

aBUGSworstnightmare wrote:
Patnet wrote:Hi Larry, thanks. It's 1.67 kilos now. If I add a 578 gram 4S battery and go to 5 amps in lieu of the 3.3 amp 272g 3S, will the all-in-weight not take a hit? Sure that's worth it?

http://hobbyking.com/hobbyking/store/__ ... _Pack.html


So, it's 1.67kg incl. the 3S Lipo, right? So new AUW with 4S would be 1.67kg - 272g + 578g = 1976g

Don't know the performance of your motor+prop so let me assume they are equal to DJI:
ecalc_dji.jpg


Better opt for a lighter 4S battery - like i.e. http://hobbyking.com/hobbyking/store/__ ... _Pack.html - and buy some larger props - i.e. http://www.rctigermotor.com/html/2014/P ... 0/282.html - not much larger, but should give you a little more thrust on theses small motors.

Your resulting AUW will be around 1.8kg with this pack which should give you enough control authority and a good performing craft.

By the way: tested my setup (AUW 1.4kg for 3S!) with T9545 --> performance was very poor; you don't want to fly such a craft in windy conditions. With 4S (AUW with 4S6000mAh Nanotech pack is roughly 1650g) it shows an awesome performance.
I'm pretty sure you know the video...
aBUGSworstnightmare
 
Posts: 1460
Joined: Fri Jun 22, 2012 5:24 pm

Re: Tuning soft frame

Postby Patnet » Tue Oct 27, 2015 5:20 pm

Thanks. This is all great stuff. So, this is my plan:

From earlier other thread:

"I would start by reducing
CTRL_TLT_ANG_F 0.125
CTRL_TLT_RTE_F 0.10000000

Your starting quite aggressive with your YAW_RT; I would suggest to start tuning with the defaults and increase them later on (o.k. let's compromise on 90 instead of 160 (which let the craft spin very fast!))."

I made these aforementioned changes at the weekend along with a new MAG calibration and it created a loss of control--bird acted like it had drunk beer and whiskey.

So, I've gone back to my earlier PIDs and new MAG. I have tested and MR flies fine--no longer wobbling like it's just left a nightclub.

Next will test. CTRL_TLT_ANG_F 0.125 whichh both Joerg and Larry suggest; then reducing the CTRL_TLT_RTE_D to around 1500 instead of 11850 to reduce vibration suggested by Larry. Will log. Then will look at battery issues.

I have couple questions:

Joerg, you say: "throttle range should be 0 - 1400, yours is - something." How do I change that. I can see in the Radio tab that mine starts at -54 and goes to 1400. How do I get the -54 to show as zero?

In calibrating SIMONKs I just do the usual, right? A post on the forum mentions MOT_ESC_CALI. I don't see that. Is that redundant?:

viewtopic.php?f=33&t=3003&start=130#p26079
Patnet
 
Posts: 138
Joined: Thu Jan 29, 2015 10:24 pm

Re: Tuning soft frame

Postby aBUGSworstnightmare » Tue Oct 27, 2015 6:09 pm

Patnet wrote:Joerg, you say: "throttle range should be 0 - 1400, yours is - something." How do I change that. I can see in the Radio tab that mine starts at -54 and goes to 1400. How do I get the -54 to show as zero?


You need to adjust the settings of your TX. You should be able to set range and trim (sub-trim, dependent on your TX).

Patnet wrote:Next will test. CTRL_TLT_ANG_F 0.125 whichh both Joerg and Larry suggest; then reducing the CTRL_TLT_RTE_D to around 1500 instead of 11850 to reduce vibration suggested by Larry. Will log. Then will look at battery issues.

Start with one parameter at a time. Reducing the CTRL_TLT_ANG_F should help you in dealing with vibrations. Test the result and then start reducing CTRD. Be careful when doing this! You can come to a point where it starts to wobble again --> experienced this when tuning the spider. Increased CRTD by 100 and kept this setting.

Patnet wrote:In calibrating SIMONKs I just do the usual, right? A post on the forum mentions MOT_ESC_CALI. I don't see that. Is that redundant?:
viewtopic.php?f=33&t=3003&start=130#p26079

Requires a custom (HOTT in this case) firmware, not available with official release.
aBUGSworstnightmare
 
Posts: 1460
Joined: Fri Jun 22, 2012 5:24 pm

Re: Tuning soft frame

Postby Patnet » Wed Oct 28, 2015 3:27 am

Just a quick update:

1. CTRL_TLT_ANG_F 0.125: caused failure in yaw response. Yaws very quickly one direction and slow the next. Possible cause: crash damage. Will look for crooked arms with a level.

2. "reducing the CTRL_TLT_RTE_D to around 1500 instead of 11850" caused major wobbles and big crash.

I'm unsure as to how to read the IMU_ACC logs still.

On page

http://autoquad.org/wiki/wiki/configuri ... -problems/

the two bottom images look completely different to me. Yet the text says they are examples of the same low vibration levels. Can this text be correct?
Patnet
 
Posts: 138
Joined: Thu Jan 29, 2015 10:24 pm

Re: Tuning soft frame

Postby aBUGSworstnightmare » Wed Oct 28, 2015 7:09 am

Patnet wrote:2. "reducing the CTRL_TLT_RTE_D to around 1500 instead of 11850" caused major wobbles and big crash.

Tuning such a frame requires patience --> never reduce a parameter such a big number at once! Take small increments (i.e. 500 or 250), spool up motors and look how the craft reacts.
If it starts wobbling don't attempt to fly! Land --> change parameter once more and try to fly.
As I already told you: you may come to a point were increasing / decreasing a parameter makes things worse. So, dial it back and start tuning another one (if needed).

Patnet wrote:I'm unsure as to how to read the IMU_ACC logs still.

Your frame is soft --> you're using PID. That's a combination which is very hard to tune since PID amplifies vibrations. That means if you have a lot of vibrations (high numbers in IMU_ACC) you need to reduce them. Ideal value here is (far) below +/- 2.
If you can't reduce them (i.e. because the frame is soft) you need to tune your PIDs to values were the vibrations won't cause any harm. This may not result in the 'best performing craft ever' but will allow you to fly and enjoy your copter.

Patnet wrote:1. CTRL_TLT_ANG_F 0.125: caused failure in yaw response. Yaws very quickly one direction and slow the next. Possible cause: crash damage. Will look for crooked arms with a level.

This statement makes no sense to me! I don't understand why you can loose YAW control by smoothing D-term for angle control.
From the wiki:
F term

The F term is used to smooth out the D term. Since we have relatively large D value, it’s not practical to apply all that output to the motors in a small time step (1/400Hz = 2.5 milliseconds). The F term regulates how much of the D term is applied at each time step. In the above example, a value of 0.25 means only 25% of the D term is applied to the motors at each time step.


Questions:
- did you start from scratch of with some old parameters?
- what are your YAW params?
-is your MAG calibrated correctly?
- just to make sure: LiPo is connected to ADC pad of your M4V2 (refer to http://autoquad.org/wiki/wiki/m4-microc ... onitoring/) for voltage monitoring?
aBUGSworstnightmare
 
Posts: 1460
Joined: Fri Jun 22, 2012 5:24 pm

PreviousNext

Return to AutoQuad M4 Micro Controller

Who is online

Users browsing this forum: No registered users and 2 guests

cron