throttle adjustment?

[13brv3]

Greetings,

I'm still tinkering, and got a chance to go to the beach to try the GPS hold today. It was also the first chance I had to even fly the board outside. Of course it was cold (by FL standards), and windy (15-20 kts). You know it's always more windy at the beach, but it was worse than I had hoped for.

Overall, the board was well behaved. I had the main D term down to 5500, which was free from oscillation inside, but it was clearly too low for any wind.

Hold testing was a mixed bag. The position hold did a pretty good job, except for one time when the wind was just too much for it to hold against. I think in any reasonable amount of wind it would have held pretty well. Altitude hold basically didn't It just did the high freq pulsing and continuously dropped toward the ground.

Before I start messing with the parameters, Steve had mentioned making sure the throttle is at 700 in hover. I found a link to this thread where it's explained as well- viewtopic.php?f=27&t=539&start=10 Unfortunately, mine doesn't seem to work that way, or I'm doing something wrong.

My test quad is light, and in hover, LOG_MOT_THROTTLE is 300. The only factor throttle term I can find is CTRL_FACT_THRO, which is 0.7. Using the formula in the thread above, I calculated the correct throttle factor to be 0.3, and I entered, saved, and wrote to ROM. After that, I tried another logged hover flight. Takeoff took quite a bit more throttle stick, and I managed to get a little hover before the yaw went somewhat nuts. It was a bit off, so I added some yaw stick, and it started spinning in the direction I moved the stick. It took lots of stick to stop it, then it was spinning the other way. Fortunately, I was only a few feet off the floor, so I just eased the throttle and let it land.

After that attempt, I looked at the log, but what I saw was the same 300 value for LOG_MOT_THROTTLE. The fact that the value didn't change, coupled with the super yaw behavior makes me think I must have changed the wrong parameter. Did I do it wrong, or is this different on R76 now?

BTW, I noticed a parameter names MOT_HOV_THROT, which sounds like something that would set the throttle level for hover, but I wasn't brave enough to mess with it.

Thanks,
Rusty

[LPR]

Rusty

The throttle factor works well for me when it's set so you can switch between manual mode and position hold and back and not have the quad change altitude very much.
On the three multicopters I have flying now I adjust or add an arm and spring to the throttle stick so it springs to center position (700) just like the pitch and roll stick.

With the throttle springing to center it's very easy to get in position hold and the autonomous mode, just let go of the stick.

One of the quads I built is just about the same as yours and it flies great with r 76 default PIDs except for Tilt Rate D is 5000 and to get a more aggressive altitude hold I set Navigation Altitude Position P = 0.35 and Navigation Altitude Position I = 0.02 .

Here is how it flies on a very windy day..

https://www.youtube.com/watch?v=8JWLDgvwul4

Remember to have some foam over the barometer.

Larry

[phynix]

Hello,

as far as I have understood it, the factor throttle acts between the throttle channel of the RC (normally channel 0 in the log) and the motor_throttle. So the channel0 should be adjusted to be 700 if you are hoovering. The motor throttle is the value going to the esc, so this is a fixed value for hoover. Please correct me if im wrong...

I also read in another thread that the parameter "hoover throttle" normally isn't used at all, so you can put in there whatever you like and it would not matter.

Sören

[JussiH]

Its really very simple

The CRTL_FACT_THRO controls the scaling of the throttle signal.

So in very basic terms, you first set up your control channels to a 0 to 1400 range and that when thr throttle is at midstick, the value is 700 in the gui.

Then you hover manually, and subsequently adjust the throttle factor so that the craft hovers at midstick in manual mode. Then when you switch to position hold, you will be at midstick and the copter should hold its altitude.

You can also increase the throttle deadband to 60, that makes it easier to hit the midstick position.

AQ has a DVH (Direct velocity hold) that is active in position hold. So if the stick is under midstick, you will decent at a rate proportional to the stick input, and same thing goes for above midstick. Only when you are within the throttle midstick deadband the copter will hold altitude.

Rusty, do you have a log you can post of the issue?

If you are experiencing drifting downwards in position hold, and you have verified there is a 700 (+/- deadband) signal coming to the controller at midstick, then you could need to increase the NAV_ALT_SPEED IM and OM parameters to allow more altitude control to the motors.

But check the other stuff first and post a log file.

[sandmen]

[phynix]

[Kisssys]

The link that the link phynix referred to should explain it better. I did a search in the wiki for throttle factor setting procedures but I couldn't find it. It's the first thing you need to do and very important and searching for bits and pieces doesn't seem too good. We'll look into that.

I would try these Yaw settings also. It's important that the motors in a hover have the same average power without yaw being applied. The AQ needs a neutrally stable machine so adjustment of the motor angles so that at a hover power on all the motor's are the same.

In a quad you can look at MOT_PITCH and MOT_ROLL, plots should overlap. Usually you'll see a motor is off but if not then you can figure out which way to rotate a motor to get the needed yaw by which props are working harder.


// YAW rate PID
#define DEFAULT_CTRL_YAW_RTE_P 300.0f
#define DEFAULT_CTRL_YAW_RTE_I 0.15f
#define DEFAULT_CTRL_YAW_RTE_D 50.0f
#define DEFAULT_CTRL_YAW_RTE_F 0.25f
#define DEFAULT_CTRL_YAW_RTE_PM 150.0f
#define DEFAULT_CTRL_YAW_RTE_IM 150.0f
#define DEFAULT_CTRL_YAW_RTE_DM 150.0f
#define DEFAULT_CTRL_YAW_RTE_OM 300.0f

// YAW angle PID
#define DEFAULT_CTRL_YAW_ANG_P 0.05f
#define DEFAULT_CTRL_YAW_ANG_I 0.0004f
#define DEFAULT_CTRL_YAW_ANG_D 0.0f
#define DEFAULT_CTRL_YAW_ANG_F 0.0f
#define DEFAULT_CTRL_YAW_ANG_PM 2.0f
#define DEFAULT_CTRL_YAW_ANG_IM 0.04f
#define DEFAULT_CTRL_YAW_ANG_DM 0.0f
#define DEFAULT_CTRL_YAW_ANG_OM 2.0f

[13brv3]

Thanks for the comments. Maybe I'm starting to understand a little better how this works. The previous thread said that LOG_MOT_THROTTLE would change, and should be adjusted to 700, but I see that it was really LOG_RADIO_CHANNEL0 that was changing. For me, throttle is channel 2 (3 by normal numbering <g>), and that is in fact changing with the CTRL_FACT_THRO changes. It was around 450 to start, and ended up being around 1000 with the change, because I calculated based on 310 as a starting point.

I also see that my radio input may need some adjustment since it looks like it's starting off at -100. I'll fix that first, then adjust the factor throttle for 700.

If I understand correctly, this is all just to make sure the throttle is in the deadband range when switching to hold, correct? If so, it shouldn't really be a big issue, other than to know that you might need to adjust the throttle some to get in the deadband. The larger deadband should help stay in range when flying with and without payloads.

I'll attach a couple screen captures of the stock 0.7 throttle factor, and 0.3 that I tried.

Thanks,
Rusty

[Kisssys]

[13brv3]

OK, I got the throttle range set to 0-1400, and changed the hover throttle channel setting from 475 to 700. I also bumped the NAV_ALT_SPEED IM and OM from 700 to 800. I'm not sure how much of a change that is, but I'll make sure I don't try it inside

Thanks,
Rusty (can't test for a few days)