Electrohub Flip 1.5 Slow Yaw

[AdamBomb42x]

So this is my first "real" multirotor that I just finished on Sunday. It flies great except the yaw authority is very weak. I have done some tweaking; I setup auto-level, calibrated the ESCs, and set my sub-trims. I did try to bump up the yaw rate in the MultiWiiConf, but that just made it unflyable. It would flip as soon as it came off the ground, so I brought the yaw rate back to 0. I've seen videos of others with the same setup and their yaw looks way better than mine. All the rates in my DX6i are set to 100% with 50% expo.

 

[brettp2004]

What did you bump your yaw rate up to? I think I turned mine up only to like .15. Also, I don't have any expo on my rudder so I don't have to push the stick as hard.

 

[makattack]

Well, the you're using the Flip1.5 with default settings, there's already something like 65% expo configured on it already. With your TX based 50% expo, that's a lot of mushiness around the center! I have my TX set for 0% expo and just set the expo / rates on the flight board. Most multirotors with these types of flight controllers are setup similarly. Do all your tuning on the flight controller, not the TX.

 

[Snarls]

I've also heard around here on the forums that the slow yaw responsiveness on MW is usually because the yaw PIDs are too high. Try lowering the yaw P value and see how it is. Personally I use both a slightly lower yaw P value and a .30 rc yaw rate.

 

[Lickinlimes]

On the web page for the Flip 1.5 configuration he says that if you want a faster yaw then you can change the rate to .3 in MWC. If I wasn't on my phone I would find the link for you!
I'm speaking from experience here, I did the same thing

 

[Craftydan]

Adam,

Something isn't right there. Increasing your yaw rate shouldn't change the flyability of your airframe, just how touchy it is on the yaw stick, and even then it shouldn't need enough of touch from you on takeoff to insta-flip . . . and again this kinda thing doesn't cause insta-flips.

I think you need to reflash your board, clearing your EEPROM. Something isn't right with that load.

 

[AdamBomb42x]

Thanks for all the replies, this noob appreciates it.

Dan, I thought the same. My understanding was that upping the yaw rate should just make the yaw more responsive. BTW, I set it to .5 just as a test. After breaking a prop I connected it back to my PC and looked at what it was doing. I would bring the throttle up to half and I would see motors D10 and D11 spin up significantly faster than D3 and D9. Also, D11 would be faster than D10 but by a smaller margin. As I left the motors spinning at half throttle while monitoring the rpms, the difference in rpms between D10&D11 and D3&D9 would increase over time. Once I put the yaw rate back to 0 (or as close as the interface would let me), it flew fine with the exception of the slow yaw.

Thank you for all the suggestions, I will be trying them one at a time once I get a chance. I will most likely have to do this on Sunday as my wife will demand full attention on Saturday.

 

[AdamBomb42x]

I flew my quad some today with the expo turned off on the transmitter and it does feel better, but it has a tendency to throttle up a lot when I'm only giving it a small of throttle. It also will wobble when climbing fast and descending. Pitch and roll feel great, no complaints there, but yaw is still lacking. I'm going to start adjusting the P gains, any suggestions on what I should start with before I start?

 

[makattack]

That could be vibration affecting the flight controller. Dialing up the P gains will only make that worse, I suspect. How have you mounted the flip1.5? Are you using foam tape or some other method? Have you balanced the motors and props? If you can carefully dynamically balance them, after statically balancing them individually, that would be the best too... It's just a guess on my part as I'm fairly newbish as well!

 

[brettp2004]

Something I've noticed about the spider quads in particular is that if you just give a little bit of yaw it wants to roll and wobble a little bit, but if you commit to a turn and yaw a little harder than you normally would it seems to do a little better.

 

[AdamBomb42x]

That could be vibration affecting the flight controller. Dialing up the P gains will only make that worse, I suspect. How have you mounted the flip1.5? Are you using foam tape or some other method? Have you balanced the motors and props? If you can carefully dynamically balance them, after statically balancing them individually, that would be the best too... It's just a guess on my part as I'm fairly newbish as well!

I have not balanced the motors, but I have meticulously balanced the props. I mounted the flip with velcro and made sure that it is gripping tightly.

 

[AdamBomb42x]

Update: I lowered the pitch and roll P rates from 3.3 to 3.0 and that small change eliminated the wobble when climbing and almost eliminated it when descending. I also increased the yaw rate to .1 and yaw authority is a lot better, but I'd like a little more so I'll bump it to around .15. All in all it feels very good, but I do still get some acceleration when hovering and not commanding it to do so. It has been windy so I don't know if the wind would have anything to do with that. I'll post a video if anyone would like to see it.

 

[joshuabardwell]

I would bring the throttle up to half and I would see motors D10 and D11 spin up significantly faster than D3 and D9. Also, D11 would be faster than D10 but by a smaller margin. As I left the motors spinning at half throttle while monitoring the rpms, the difference in rpms between D10&D11 and D3&D9 would increase over time.

This is normal behavior when the quad is constrained (e.g. tied down on the ground hooked to your computer). The vibrations cause the quad to think that it needs to correct the yaw a little bit, but it can't do that because it's tied down, so the correction slowly increases.

Yaw drift while hovering is a bit tricky to resolve. Fortunately, yaw is the ONE axis that it's okay to trim out drift on with your transmitter trims. This is because yaw is handled exactly the same in self-level (Angle / Horizon) and non-self-level (Acro) modes. But before you do that, you should make sure that you have set your endpoints and channel centers correctly. Have you confirmed in MW GUI that your channel endpoints are 1000/2000 and your center is 1500?

 

[AdamBomb42x]

I did set my centers in WM GUI, but the end points don't reach all the way to 1000/2000. I have my travel adjustments at 125%, which is the maximum that my DX6i will allow me to set them.

 

[joshuabardwell]

I did set my centers in WM GUI, but the end points don't reach all the way to 1000/2000. I have my travel adjustments at 125%, which is the maximum that my DX6i will allow me to set them.

If your centers are at 1500 in MW GUI, then that's not the cause of the yaw drift. If you've taken reasonable steps to remove vibration from your FC, then I'd say just use the trim on your transmitter to fix it.

 

[AdamBomb42x]

Yaw drift isn't my problem, it's yaw authority. Although, it's better with the yaw rate at .1 and I will be bumping that higher yet, just didn't get to it today.

 

[joshuabardwell]

I couldn't get decent yaw rate out of my MW Flip, unless I jacked the yaw rate parameter up so high that the quad started kind of flipping out whenever I did heavy yaw. I switched to an A-tail design and I couldn't be happier.

 

[AdamBomb42x]

I couldn't get decent yaw rate out of my MW Flip, unless I jacked the yaw rate parameter up so high that the quad started kind of flipping out whenever I did heavy yaw. I switched to an A-tail design and I couldn't be happier.

Are there any advantages to having an A-tail over a V-tail?

 

[joshuabardwell]

Are there any advantages to having an A-tail over a V-tail?

With an A-tail, there is no change needed in the motor logic in order to effect yaw. What I mean is, if you look at the default motor directions for MW, to yaw right, the back-left motor and the front-right motor spin up, and the other two spin down. With an A-tail, the back motors point in the right direction that this logic is preserved. With a V-tail, the yaw-logic of the rear motors must be reversed from the default MW directions, which complicates the transition from flat quad to angled motors. Either you reverse the directions of just the rear motors, in which case you're yawing with two motors on the same side of the quad, or you reverse the directions of all the motors. It's not a big deal, but given that V-tail and A-tail are otherwise identical (as far as I can tell), may as well go with the one that requires minimal fiddling. In practice, you can improve the performance of a V- or A-tail, by tweaking the motor mixes, but if you were to convert a working flat quad to a V tail, it would basically fly just fine without modification.

 

[joshuabardwell]

... incidentally, I have heard some people conjecture that there is a difference between an A-tail and a V-tail, and I think it's an interesting question. In terms of pure vector sums, there is no difference between tilting the motors to the left or the right. The only question is whether the air streams are mixing on the bottom of the props (output) or the top (intake). So perhaps there is some difference in how the air streams flow in a V-tail vs. an A-tail, but then again, the model that a prop works by blowing air backwards is only partially correct, as I understand it. In other words, a prop works by pushing air backwards, and therefore by pushing itself forwards by conservation of momentum, but I'm not sure how much it matters what happens to the air after it is affected by the prop--at least, once you're out of ground effect. Perhaps one could argue that, with a V-tail, the incoming air for both of the rear props is being pulled from the same location, and maybe it is more turbulent or something. I dunno... these are all interesting questions, but I don't know of anybody who has reliable answers to them. It'd be cool to see somebody test this all with a smoke streamer wind tunnel or something.