Welcome to FPVDronePilots!
Join our free FPV drone community today!
Sign up

FPV Manual Mode -- Fifth Flight

Chaosrider@brilliantfrogs

Well-Known Member
Joined
Nov 23, 2021
Messages
142
Reaction score
29
Age
69
Getting better each time. This time I got in a solid 8 minutes of FPV Manual time. This time, I paid attention to how long it took me from engine start, and then climb up 300 ft to my switchover point, and throw the switch. 30 seconds. So I decided to come home and hit the Brake at 8:30, which was about as low as I wanted to get in terms of battery power anyhow. It’s a lot more conservative than I am about deciding what “low power” means! But once I tell it to ignore it, it does.

I expect that from this point forward, my flights will get shorter, rather than longer, because I’ll be going a little faster each time…

It told me again to tighten the throttle stick. Why would it do that? How would it have any clue how stiff they are? I’m pretty happy with the current setting, although a little stiffer wouldn’t hurt. I’m definitely going to tighten up the right stick a bit. The biggest “uncommanded annoyance” is when it starts to drift backward. Now, tilting the nose down a bit to arrest that is natural at this point, but I’d like to keep it from drifting that way by default. Clearly, the center position of the right stick lets it drift backward, when I don’t intend it. If I tighten it up a bit, it should stop doing that.

Setting the camera at a fixed orientation and leaving it there does indeed help. I’ve re-programmed my brain to think of the camera image as my attitude indicator, as long as I leave it fixed. I’ve also evolved my understanding of what the sticks do. Knowing that it’s an oversimplification, I treat the left stick as my vertical speed control, and the right stick as my forward/reverse control. Works good! The actual motions are a bit more nuanced than that, but what’s happening is that my fingers are just figuring out what to do, and then after the fact, I observe what I did. Not many more flights, and I’ll be able to describe that a little better.

For a long-time fixed wing pilot like me, one way to explain the value of putting the camera in one place and leaving it alone, could be to ask the simple question, “Do you want an attitude indicator, or not?” I think that makes the point pretty clearly. I’m sure I’ll diddle with the camera from time to time, but for the most part, I’m going to leave set it and leave it where it is while I’m still in my early learning. Good advice!

More than once, I’ve found myself wishing I could hear the sound of the engines in flight. Sometimes I can…if I’m not too far away!

I was able to get it to sort of hover, after a fashion, by cheating. This was the first time that I’ve flown in Manual mode when there was non-trivial wind. I was able to keep it roughly in the same place by flying into a 10-12 mph wind. My objective now is to fly a little more slowly, but also more steadily. The wind will only get worse as the day wears on, so that was my last FPV Manual flight for today. On my first flight tomorrow, I plan to just do laps up and down the canyon, from here down to the Flag, and improving my control each time, and as I improve my control, increase my speed.

I did see 70 mph briefly, but that just sort of happened during an…excursion…followed by a recovery. It seemed to have more of a tendency to roll to the left without me telling it to on this flight. Wind, perhaps? I’m paying more attention to coordinating the turns, and that has significantly enhanced my directional control!

The Beast definitely does a major shake, rattle and roll when I hit the Brake button! I’m used to it now, but it’s still a bit surprising.

My biggest on-going screw-up isn’t in Manual mode per se. When I shift back to Normal mode for landing, I don’t automatically account for the stiffness of the left stick, and I expect it to re-center on its own. I catch it and correct it quickly, but that’s a bad habit that I need to break.

It wouldn’t do at all for me to successfully terrorize the skies in Manual, just to plop it into the sagebrush in Normal when I’m on final for landing!

1654549729874.png
 
It told me again to tighten the throttle stick. Why would it do that? How would it have any clue how stiff they are?
There are 2 screws, one releases the spring, the other adjusts friction to your liking. You didn't fully release the spring.
Clearly, the center position of the right stick lets it drift backward, when I don’t intend it. If I tighten it up a bit, it should stop doing that.
You still haven't understood that the stick position has no direct connection to "drifting", only the attitude does.
 
Simply put, if the right stick is held off-center the drone will flip over. If you don't understand why, you are not understanding rate mode.

Try it.
 
........

You still haven't understood that the stick position has no direct connection to "drifting", only the attitude does.
I took a look at several flights to see what happens to pitch after Manual is selected and before it actually switches to Manual. In my flights as well as your one flight pitch slowly increases during the PreManual phase. E.g.,
1654612472677.png
The purple background is the PreManual phase followed by the beige background denoting the Manual phase. Note that ctrl pitch (red) is 0. It can be seen that during the Premanual phase pitch (green) dampens then is slowly increasing. After switching to Manual with no pitch input the increasing pitch continues causing the FPV to move backwards.

I was tempted to conclude that the FPV always has this very slow pitch increase during the PreManual phase. But, here is case where pitch is decreasing at a relatively fast rate. I'm supposing the reason is that, unlike the previous example, PreManual started while the FPV was moving.
1654613002691.png
 
Would need to check in various conditions but as far as I've noticed without paying attention pre-manual still maintains a hover, so logically the aircraft attitude has to adjust to the current conditions... so when you're let go it'll be "about right" but of course not exactly since as mentioned on the other thread "exactly" is impossible without constant computer control. There's also not much point being exact since when you're taking control you'll then be going somewhere anyway, it's just so you don't end up with a wildly crazy thing when switching.
Not to mention the throttle on which the green bar is quite wide so you can be further off.
 
Last edited:
  • Like
Reactions: Fpvbefokt
There are 2 screws, one releases the spring, the other adjusts friction to your liking. You didn't fully release the spring.

You still haven't understood that the stick position has no direct connection to "drifting", only the attitude does.
You seem to have a greatly exaggerated notion of how well you can know what I do or don't understand...

I know where the adjustment screws are. I just can't figure out what set of inputs would suggest to the system that I should tighten the throttle. Through what data path would it know/believe that? I went to adjust the right stick this morning, and it turned out the actual physical left stick had unscrewed a bit from the base, and I tightened that down. I'll see if that warning goes away.

Every time I do the switch from Normal to Manual mode, the drone drifts backwards and downwards. Every time. At about the same rate of divergence. Uncorrected, both of those motions accelerate.

I have the stick positions in the same place every time, because that's what the stick position screen wants me to do before getting into manual mode. So, if the stick positions were slightly different, with a bit more throttle and pitched a bit farther down, and they stay in that amended position, it will tend to wallow around longer, before diverging.

I'm not expecting static stability.

This seems obvious to me, but then, everybody always sucks at explaining something that they think is obvious...

The OP suggested that rather than "hover", a better concept might be something like "radius of confinement". Can I keep the drone within a radius of, say, 20 meters around a point? Or less?

I'll give that a shot at some point, possibly today.

Thx.
 
Simply put, if the right stick is held off-center the drone will flip over. If you don't understand why, you are not understanding rate mode.

Try it.
"If you can't explain something simply, you don't understand it well enough."

--Albert Einstein

So, what's a simple explanation of "rate mode"?
 
"If you can't explain something simply, you don't understand it well enough."

--Albert Einstein

So, what's a simple explanation of "rate mode"?
I may not understand it well enough, but here is something I posted about rate mode over on the Mavic forum earlier with regard to another quad.

Rate mode means that the pitch and roll stick movements change the rate at which the pitch or roll angle changes. Move the stick a little and the angle changes slowly. Move it a lot and the angle changes quickly. And when you return the stick to center, the rate of change will go to zero and the aircraft will maintain whatever pitch or roll angle it is currently at. To get back to horizontal, you have to apply opposite stick and release it when the aircraft is horizontal. If you hold the stick all the way to one side, the aircraft will continuously flip over and over, until it hits the ground.
 
I know where the adjustment screws are. I just can't figure out what set of inputs would suggest to the system that I should tighten the throttle.
Simple, the radio has a switch/sensor that triggers when the throttle spring is fully released, and if it's not it tells you. And again it's not about the friction which I guess is what you call "tightening".
 
Last edited:
Simple explanation: Yaw. What happens to yaw if you hold the stick slight off center to the right or left? I.e. trim the yaw? could you fly straight?

Of course not, because if the yaw input is not centered at zero it will rotate around the vertical axis, around and around, forever, and you couldn't control the aircraft. The bigger the deflection, the faster the RATE of rotation around the yaw axis.

In manual the roll and pitch controls behave like yaw. Hold a tiny bit of forward pitch, and the aircraft with pitch slowly forward, and not stop. It will continue until upside down, and continue until upright again. And then continue, pitching over forward until you release the stick.

Do you see why putting "trim" on pitch would be a disaster?

If it still isn't clear, go try it and see what happens. Be ready on the brake.
 
I may not understand it well enough, but here is something I posted about rate mode over on the Mavic forum earlier with regard to another quad.

Rate mode means that the pitch and roll stick movements change the rate at which the pitch or roll angle changes. Move the stick a little and the angle changes slowly. Move it a lot and the angle changes quickly. And when you return the stick to center, the rate of change will go to zero and the aircraft will maintain whatever pitch or roll angle it is currently at. To get back to horizontal, you have to apply opposite stick and release it when the aircraft is horizontal. If you hold the stick all the way to one side, the aircraft will continuously flip over and over, until it hits the ground.
This was helpful. I've read this before, perhaps in your own previous post about it that you mention. Then as now, I understand the description, the underlying physics, and what it would suggest for aircraft behavior.

My issue hasn't been understanding it, my issue has been believing it, subconsciously at least.

My mind sort of wrinkled it's virtual nose, and said, "Who would want that? It's got to be some negative side effect of some kind."

Now, I believe it. As I mentioned in another post, there's an unavoidable trade-off between stability, and maneuverability. I also mentioned that some fighter aircraft are intentionally designed with negative stability, for increased maneuverability.

The FPV in Manual mode has negative stability. And this is done on purpose, because some people like the loss of stability to get the maneuverability. *I* don't, but some people do. I just want to go fast. I knew that Manual mode wouldn't have positive stability, and it seemed initially like it had neutral stability. But now I believe that it has negative stability, and that explains a lot of the behavior I've seen. I just couldn't believe that initially.

To some extent, it was just the term "rate mode" that threw me off. That term meant nothing to me, but it is in fact describing a condition of negative stability, where the system diverges at an accelerating rate in response to small changes.

I just completed my first laps up and down the canyon at full throttle, and I've seen speeds in the 80 - 82 mph range, which is all I can reasonably expect at this elevation. I'm not graceful with it yet, but that gets better with every flight. Mostly, my fingers just need to tune in to what the aircraft wants.

Today my full throttle CanyonBall run was at 250 - 350 ft ATL. My objective with Manual mode has always been to do the CanyonBall at full throttle, at 100 ft ATL. My current path will get me to that point quite reliably. Once I do that, I'll have to decide if the coolness of the speed, is worth the "tax" of the loss of automatic stability controls.

Maybe it will be, maybe it won't. But I won't know 'til I get there!

What I really want is the "Extreme Sport" mode -- full power, full speed, full stability. Maybe eventually it will come to pass...

It's very loud at that speed. I think one of my neighbors is taking shotgun pot shots at it, which I never heard before today. But he'll have a hard time nailing it. No duck or goose cruises at 80 mph!

1654713191761.png

Thx!!
 
........

What I really want is the "Extreme Sport" mode -- full power, full speed, full stability. Maybe eventually it will come to pass...
......

Thx!!
There probably is a way to achieve this. It would be a bit of a project though. It would entail hacking the FPV to allow a higher tilt maximum which is the same as a higher speed. This could be done for Sport mode to make it the "Extreme Sport" mode you described. The FC would still be flying the FPV and providing the S mode stabilization.

Hacking the FPV can be done with
Drone Hacks PC software: Drone-Hacks, the best way to hack your DJI Drone - Download

I've used this on my Mavics to increase max speed.

One caveat though. There is a reason for max tilt. I suspect it's about preventing the FC from becoming unstable in relatively low power situations. Abrupt exogenous inputs can sometimes cause unstable calculations (e.g. pitch or roll oscillations). I think that's what happened here.
Mavic Pro Platinum strange crash
That Mavic had been modified.

It's a FW instability not a HW instability. Of course there is an absolute tilt max past which the FPV couldn't maintain altitude.

And, have you tried the Motion Controller?
 
There probably is a way to achieve this. It would be a bit of a project though. It would entail hacking the FPV to allow a higher tilt maximum which is the same as a higher speed. This could be done for Sport mode to make it the "Extreme Sport" mode you described. The FC would still be flying the FPV and providing the S mode stabilization.

Hacking the FPV can be done with
Drone Hacks PC software: Drone-Hacks, the best way to hack your DJI Drone - Download

I've used this on my Mavics to increase max speed.

One caveat though. There is a reason for max tilt. I suspect it's about preventing the FC from becoming unstable in relatively low power situations. Abrupt exogenous inputs can sometimes cause unstable calculations (e.g. pitch or roll oscillations). I think that's what happened here.
Mavic Pro Platinum strange crash
That Mavic had been modified.

It's a FW instability not a HW instability. Of course there is an absolute tilt max past which the FPV couldn't maintain altitude.

And, have you tried the Motion Controller?
I'm glad I hadn't had breakfast yet when I watched that vid...

I'm not into hacking my drones (or much of anything else), in no small part to avoid results like that. I'm amazed that thing survived that many tree hits, and still flew...after a fashion.

I'll just hope DJI makes that capability available. It shouldn't be that hard, just a different additional "Button Customization" option.

I haven't tried the Motion Controller yet, because I have a poverty-induced cap on buying any new drone stuff for a while...

Thx.
 
I'm glad I hadn't had breakfast yet when I watched that vid...

I'm not into hacking my drones (or much of anything else), in no small part to avoid results like that. I'm amazed that thing survived that many tree hits, and still flew...after a fashion.

I'll just hope DJI makes that capability available. It shouldn't be that hard, just a different additional "Button Customization" option.

I haven't tried the Motion Controller yet, because I have a poverty-induced cap on buying any new drone stuff for a while...

Thx.
I should have been explicit. DJI isn't going make higher speed available in either N or S mode because of the (rare) instability problems. Higher speeds without possible FC instabilities can only happen if the FC isn't flying the FPV; i.e. M mode. But, then pilot skill and knowledge are required to perform tasks done by the FC in N or S mode. It's a trade-off.
 
I should have been explicit. DJI isn't going make higher speed available in either N or S mode because of the (rare) instability problems. Higher speeds without possible FC instabilities can only happen if the FC isn't flying the FPV; i.e. M mode. But, then pilot skill and knowledge are required to perform tasks done by the FC in N or S mode. It's a trade-off.
Ah, OK, that makes more sense.

Every time I fly in Manual, more of the flight control tasks have been integrated into my brain, and occur in the background. As long as that trend continues, perhaps I won't need Extreme Sport.

Are these instabilities known and documented, or theoretical? Either way, what would be the cause?

It used to be common that human reflexes could outperform automation on some complex tasks, but that's becoming increasingly rare.

Thx.
 
On reflection, I think I could have been more clear as well.

I'm not talking about a change to what the default Sport mode does. I'm talking about an additional "Button Customization" option.

To activate Manual mode for the first time, you have to go through an elaborate menu dance to get to where you can tell it to set the "M" position on the RC to be Manual mode. I'm talking about adding an additional configuration option that you can reach by going through the same process.

You could select "Manual", or "Extreme Sport", or conceivably a range of other options.

The Extreme Sport mode that I describe would be less likely to produce a crash than the Manual option that currently exists...even granting the existence of the instability you describe, which you also describe as rare.

So since Extreme Sport would be safer than Manual, there's no obvious reason not to do Extreme Sport due to safety concerns.

Thx.
 
I took a look at several flights to see what happens to pitch after Manual is selected and before it actually switches to Manual. In my flights as well as your one flight pitch slowly increases during the PreManual phase. E.g.,
View attachment 5835
The purple background is the PreManual phase followed by the beige background denoting the Manual phase. Note that ctrl pitch (red) is 0. It can be seen that during the Premanual phase pitch (green) dampens then is slowly increasing. After switching to Manual with no pitch input the increasing pitch continues causing the FPV to move backwards.

I was tempted to conclude that the FPV always has this very slow pitch increase during the PreManual phase. But, here is case where pitch is decreasing at a relatively fast rate. I'm supposing the reason is that, unlike the previous example, PreManual started while the FPV was moving.
View attachment 5836
Interesting indeed, and completely consistent with my observations.

During the transition from Normal to Manual, the aircraft pitches up every time. Why would it do that? It's odd, but very consistent.

I've never done the switch while in motion.

Thx!
 
It doesn't, only the camera changes angle.
Really?

So the camera angle changes, of it's own volition, relative to the aircraft, independent of the aircraft, with no command from the pilot?

I don't think that's right.

The whole point of the advice that I got to set the camera angle somewhere and leave it there...good advice, I should add...was so that I could use the view on the screen as an attitude indicator. Works good!

IIRC, an OP showed some very cool graphs of flight data that explicitly showed the increase in the upward pitch of the aircraft during the transition from Normal to Manual.

Are you saying that's wrong?
 
So the camera angle changes, of it's own volition, relative to the aircraft, independent of the aircraft, with no command from the pilot?
There is command from the pilot, you told it to switch to M.

The way the camera is controlled is fundamentally different in M from N/S (the former locks the camera to the aircraft while the latter lock it to the horizon). When you put the switch in M the camera leaves the horizon-based lock and goes to +20° which is pretty much the ideal angle for M mode on the FPV.

IIRC, an OP showed some very cool graphs of flight data that explicitly showed the increase in the upward pitch of the aircraft during the transition from Normal to Manual.
If it's the same video I've seen it shows just what I described and that pitch doesn't change.
 
Last edited:

Members online

Forum statistics

Threads
6,135
Messages
44,721
Members
5,451
Latest member
Zacharyzertuche