Of firmware and "twitchy" birds

[PatR]

I've read post after post of people having landing issues and thinking their H performs "jerky" or is too sensitive on the controls, and how the latest b30 firmware release is supposed to have resolved all that. The reason for this post is to boldly state that I have yet to experience any issues of excessive control sensitivity or any landing issues with the b27 firmware, which I am still flying in two different H's. As of now I can see no reason to change it.

After relocating from Oregon to California it was necessary to re-calibrate both systems, each having the same firmware with v03.01.b27 for the -16, v00.24 for the Tx, v1.3 for the A/P, v3.2.26 (a) for the camera, and v1.25 for the gimbal. Both aircraft perform perfectly with control response matching the manner it was input, even with the speed slider at the high point for "Rabbit". Slow and smooth input provides slow and smooth aircraft response. There is a quick but slight lean in the commanded direction when departing a hover but nothing I would consider to be twitchy or excessive in any way with an aircraft of above average performance. The softer I provide input the softer the lean when departing a hover. Regardless of how rough I provide input the initial "twitch" as some call it never translates to the video image. They both fly the same now as they did before leaving Oregon.

Landings go as I tell the H to do. If my fingers are moving fast or too much on the sticks the bird will react accordingly. If the descent is slow and smooth, even if windy, they both settle onto the gear without a bounce. If I hold the throttle stick back for the few seconds the motor speeds drops to idle. If I don't move the right stick after landing there is no evidence of the aircraft leaning in any direction. When I press and hold the arming button down the motors stop.

So my questions are, where are all the problems with the b27 firmware I've read so much about? Having a pair of H's it would stand to reason that one of them would demonstrate something negative, but neither does after numerous flights, flown hard and soft. Are these problems truly problems or induced by users that haven't quite yet developed a feel for their controls? Is it possible some people are over controlling their aircraft and impatient or distracted during the landing and shut down process?

 

[Windluv]

I have essentially the same result; however, since I re-adjusted my flying style to using thumbs rather then pinch the prior month, on occasion I will now release the throttle stick just after touching down and it will 'bounce' up to the center position and just a bit more. When this happens the H will obviously want to throttle up. It is just as you said : hold the throttle down for a couple seconds then turn the motors off. Never a problem with those deliberate actions.

 

[PatR]

I think back to when I was teaching young fliers how to handle the controls with RC planes. Without fail all of them over controlled in the beginning. They either pushed the sticks too far or moved them too quickly. For most it took a long time for them to learn moving them softly, and in much smaller amounts would delver a smoother flying aircraft. Coordinating relative movements between the left and right sticks also took some time.

 

[ArnhemAnt]

Are these problems truly problems or induced by users that haven't quite yet developed a feel for their controls? Is it possible some people are over controlling their aircraft and impatient or distracted during the landing and shut down process?

In my opinion, the issues are more than likely due to the amount of experience (or lack of) from the user. A quick glance through this forum and the amount of members who have purchased the H as their first ever multi rotor tells a pretty true story.
Like anything, it takes time to learn. Not just how to fly, but also the idiosyncrasies of each type of multi rotor. There's obviously a lot more than this (battery care, flight safety/responsibility, etc, etc) for a new user to also learn. Note, by using the term 'new user' I'm not necessarily talking about someone who has never flown before, as we are all new users of the H when we first get them and I had to learn a lot about how best to fly this particular platform, even though I have been flying multi rotors for the past six years.

 

[nicholas hickman]

Let me chime in;

not long ago (two weeks?) I experienced some very "twitchy" movement from the H when transitioning from hover to movement. It didn't matter if it was in hare or turtle mode, the slightest input was as if it was in super sport mode. I noted it was drifting forward & right very slightly but had to pack up (it was lunch break at work).

The following day I did a fresh compass calibration after which the H performed as it should, smooth as butter. This could have been just coincidence but either way there was a definite and aggressive twitch to it.

However I am not suggesting there is or isn't an issue with any version of firmware, I am running the latest and the auto landing is flawless.

just my two penneth or two cents worth.

 

[NorWiscPilot]

I've read post after post of people having landing issues and thinking their H performs "jerky" or is too sensitive on the controls, and how the latest b30 firmware release is supposed to have resolved all that. The reason for this post is to boldly state that I have yet to experience any issues of excessive control sensitivity or any landing issues with the b27 firmware, which I am still flying in two different H's. As of now I can see no reason to change it.

After relocating from Oregon to California it was necessary to re-calibrate both systems, each having the same firmware with v03.01.b27 for the -16, v00.24 for the Tx, v1.3 for the A/P, v3.2.26 (a) for the camera, and v1.25 for the gimbal. Both aircraft perform perfectly with control response matching the manner it was input, even with the speed slider at the high point for "Rabbit". Slow and smooth input provides slow and smooth aircraft response. There is a quick but slight lean in the commanded direction when departing a hover but nothing I would consider to be twitchy or excessive in any way with an aircraft of above average performance. The softer I provide input the softer the lean when departing a hover. Regardless of how rough I provide input the initial "twitch" as some call it never translates to the video image. They both fly the same now as they did before leaving Oregon.

Landings go as I tell the H to do. If my fingers are moving fast or too much on the sticks the bird will react accordingly. If the descent is slow and smooth, even if windy, they both settle onto the gear without a bounce. If I hold the throttle stick back for the few seconds the motor speeds drops to idle. If I don't move the right stick after landing there is no evidence of the aircraft leaning in any direction. When I press and hold the arming button down the motors stop.

So my questions are, where are all the problems with the b27 firmware I've read so much about? Having a pair of H's it would stand to reason that one of them would demonstrate something negative, but neither does after numerous flights, flown hard and soft. Are these problems truly problems or induced by users that haven't quite yet developed a feel for their controls? Is it possible some people are over controlling their aircraft and impatient or distracted during the landing and shut down process?

PatR,

I cannot (or should not) comment on anything relating to firmware as I have not been comparing, tweaking, or testing for a firmware-based analysis. However, I can attest to alleviating "sensitive" controls.

Previous to reading up on controller rate adjustments (TyPilot, I believe posted about this), I had noticed it seemed both of my H's "twitched" if I moved them from hover to any direction other than hover. What I mean is, the transition wasn't as smooth as perhaps I wanted it to be. Regardless, I figured it was either a characteristic, my "style" or a combination. I hadn't even thought of any possibilities to change the behavior at the equipment level.

Then along came Ty's video and the light bulb illuminated a little brighter. My recollection of rate adjustments in the controller settings, back in my RC days, was kick-started. After applying my own adjustments and, to your point, concentrating on my own movements, I can create a very smooth H behavior.

This is not to say a rate change is the end all, be all. It is, however, another item in the tool box to help make flying a little more enjoyable. Isn't that what it is all about?

Hardware and firmware should not be the only "improvements" to be made, let alone taking all the blame. I agree with your [PatR] assessments that pilot skill and experience play a big part. With that said, if tweaking can be made on both sides, why not?

Jeff

 

[PatR]

Totally agree that we should have the ability to set up the controls to our preferred "feel". I have done the same for decades with my RC planes using exponential and dual/triple rates to suit my flying style. It's been pretty well established such can be done with the H but something also of importance is the user's competence with the control functions. That needs to be developed to some reasonable level before they start messing with rates or expo. Time needs to be spent developing basic flying skills, close in where the reaction to an action can be clearly observed until an understanding of control function has been generated. The time for image recording comes after that

 

[NorWiscPilot]

Totally agree that we should have the ability to set up the controls to our preferred "feel". I have done the same for decades with my RC planes using exponential and dual/triple rates to suit my flying style. It's been pretty well established such can be done with the H but something also of importance is the user's competence with the control functions. That needs to be developed to some reasonable level before they start messing with rates or expo. Time needs to be spent developing basic flying skills, close in where the reaction to an action can be clearly observed until an understanding of control function has been generated. The time for image recording comes after that

We're in total agreement!

 

[glider]

Totally agree that we should have the ability to set up the controls to our preferred "feel". I have done the same for decades with my RC planes using exponential and dual/triple rates to suit my flying style. It's been pretty well established such can be done with the H but something also of importance is the user's competence with the control functions. That needs to be developed to some reasonable level before they start messing with rates or expo. Time needs to be spent developing basic flying skills, close in where the reaction to an action can be clearly observed until an understanding of control function has been generated. The time for image recording comes after that

Not preaching because I know you have extensive experience with Ardupilot, just throwing in my 2c.

I've posting the following video before, but it demonstrates why I think the firmware in the H is lacking when in GPS mode. I base this on my experience using Arducopter. Once moving the H is nice and smooth, but is twitchy when first moving the stick. This has nothing to do with the controller. It is built in the firmware.

This copter in Position Hold mode (GPS mode for H) is smooth as silk with movements transitioning from dead stop to barely moving no matter how hard the sticks are moved. This is how the H should respond to stick movements in or out of GPS mode.

That said, I have no problems landing or flying the H once it is moving. However, there is still a definite "twitch" when first commanding a movement in GPS mode.

 

[PatR]

Glider,

We completely agree the H autopilot code has considerable room for refinement. I avoid comparing Yuneec systems to anything running APM code because there's just no comparison. Ardu/APM, as you know, reigns supreme in code sophistication and smooth operations, removing any and all limits the operator desires. If Yuneec truly desired to be a power player they would dump their controller for Pixhawk/PX4 or later. They would also get away from WiFi.

 

[Air Pauly]

well Pat R you are welcome to throw the dice of inexperienced flying if you like, and i will send you my H and if you can fly it without it twitching everywhere then you can keep it because the only way this thing flies smoothly is with GPS off then its lovely, but at the moment a butterfly could land on one of the sticks and the thing would dip its nose heavily

even Yuneec said it should not be doing that and to send it back to them

 

[J J Antonetti]

I've read post after post of people having landing issues and thinking their H performs "jerky" or is too sensitive on the controls, and how the latest b30 firmware release is supposed to have resolved all that. The reason for this post is to boldly state that I have yet to experience any issues of excessive control sensitivity or any landing issues with the b27 firmware, which I am still flying in two different H's. As of now I can see no reason to change it.

After relocating from Oregon to California it was necessary to re-calibrate both systems, each having the same firmware with v03.01.b27 for the -16, v00.24 for the Tx, v1.3 for the A/P, v3.2.26 (a) for the camera, and v1.25 for the gimbal. Both aircraft perform perfectly with control response matching the manner it was input, even with the speed slider at the high point for "Rabbit". Slow and smooth input provides slow and smooth aircraft response. There is a quick but slight lean in the commanded direction when departing a hover but nothing I would consider to be twitchy or excessive in any way with an aircraft of above average performance. The softer I provide input the softer the lean when departing a hover. Regardless of how rough I provide input the initial "twitch" as some call it never translates to the video image. They both fly the same now as they did before leaving Oregon.

Landings go as I tell the H to do. If my fingers are moving fast or too much on the sticks the bird will react accordingly. If the descent is slow and smooth, even if windy, they both settle onto the gear without a bounce. If I hold the throttle stick back for the few seconds the motor speeds drops to idle. If I don't move the right stick after landing there is no evidence of the aircraft leaning in any direction. When I press and hold the arming button down the motors stop.

So my questions are, where are all the problems with the b27 firmware I've read so much about? Having a pair of H's it would stand to reason that one of them would demonstrate something negative, but neither does after numerous flights, flown hard and soft. Are these problems truly problems or induced by users that haven't quite yet developed a feel for their controls? Is it possible some people are over controlling their aircraft and impatient or distracted during the landing and shut down process?

I think the tendency to call it pilot error or "newb" is a little on the course side. I myself have had the landing issues with my H and I know it's related to the firmware and some bug in realsense. Yes I know I've not had this bird for real long ( late last year) but through the transitions of firmwares I've seen some buggy code rear its ugly head. When I first got my H the firmware was awesome ( I think it was 3 revisions ago). I could land on a dime no problems without bounce or whether I had OBS on or off. Now I just know I need to turn OBS off while landing ( yes I know what the manual says) other than that I think some firmware issues arise from how you fly your bird, different patterns might emerge from different styles. I'm just glad I got my H when it had the firmware it did because otherwise I might have junked the H and gone over to the darkside.

 

[PatR]

I'll have to admit, I've never bothered to turn OBS on to see what happens when using it. In fact, I've never turned OBS on at all.

 

[ArnhemAnt]

i will send you my H and if you can fly it without it twitching everywhere then you can keep it

I wish you and I lived closer. I'd love to have a look at the issues you are describing, particularly if I can get to keep it.

 

[CraigCam]

Not preaching because I know you have extensive experience with Ardupilot, just throwing in my 2c.

I've posting the following video before, but it demonstrates why I think the firmware in the H is lacking when in GPS mode. I base this on my experience using Arducopter. Once moving the H is nice and smooth, but is twitchy when first moving the stick. This has nothing to do with the controller. It is built in the firmware.

This copter in Position Hold mode (GPS mode for H) is smooth as silk with movements transitioning from dead stop to barely moving no matter how hard the sticks are moved. This is how the H should respond to stick movements in or out of GPS mode.

That said, I have no problems landing or flying the H once it is moving. However, there is still a definite "twitch" when first commanding a movement in GPS mode.

Honestly I think it's because it's a hex and it doesn't behave like a quad due to the different power and thrust ratios.

 

[Texy]

The first time I ever landed mine it flipped over, but I know now that it was my fault. I was landing it in a different place from where I launched it, and it wasn't aware that it was actually on the ground and it fought me for control a little bit. Now I know to hold the left stick down after it touches down and until the props go into idle speed.