Micro tuning techniques - different than larger quads


#181

Yeah, that’s exactly why I created a new method for brushed outdoor quads. I forget exactly what post number it was born on but here is the short version:

Step 1: set yaw P to about 90, and all I values to somewhere between 75 and 90.

Step 2: pick a really windy day and go fly a leisurely straight line under fpv while the wind is knocking your micro around. You will see big bumps. Raise P on both pitch and roll till the big bumps turn into very tight very high frequency oscillations. If its a really day you may still hear them, and that’s OK. You just want to see smooth video at the end of the whole process.

Step 3: go fly the same straight lines and raise D on pitch and roll till the tight oscillations when it’s pushed on by the wind or turbulent air smooth out visually.

All done. Quick and simple - no crazy flying needed

Ocassionally I will do a straight down SLOW decent while flat into my propwash and apply the same method - but it will never get totally perfect there. As good as you can get it on that is kind of fine tuning.


#182

After this it should fly about perfect. If you can’t find wind then use high altitude to produce turbulent air - it’s up there. The theory behind this is that bumpy air can push harder on a brushed micro than it can push on itself in any maneuver you do in the Stingy method. Therefore we tune to the weather and not to the sticks. The sticks fall into place if it can take the wind. I usually like a 20mph+ gutsy day.


#183

@PJC - I also used to use my own version of the “Stingy” method - which is basically just the old school tuning method that has been around since the days of multiwii. I started this thread because it just didn’t work that well for brushed micros. After some brainstorming here with all you guys - that’s where the new method above came from. For reference, this is a brushed only technique. I still old school my brushless quads cause they have enough power that they can create their own extreme conditions with stick inputs needed for analyzing and correcting a tune. But for brushed, you should be able to fly in any conditions after this and have it behave as if it’s a dead calm day!! I will add that some times I have seen extreme rates mess up the whole operation and make a quad untunable. In other words - it is critical your rates can be achieved by your craft or you will hit a point where your quad can’t keep up with stick inputs and you just accumulate error with every stick input causing erratic untunable behaviors. (That’s really important )


#184

A side note, not sure whether it was already mentioned here (it has been mentioned in the 1106 already but is tuning related):

If you are using current brushless motors (e.g. RCX 1103 / Lizard95 1104 / Realacc 1106 / Brotherhobby 1106 to mention a few) with a 3S+ setup, make sure that you don’t use betaflight stock pids (I’m refering to BF 3.2, build from the latest git revision here, but this is must probably also valid for 3.1.7). Roughly halving the pid values seems to be a good starting point. P and D are way too high for tiny high torque setups.


#185

@NotFastEnuf, ahhh, of course, now it’s dead calm outside. :stuck_out_tongue:

As I’m looking thru this thread I’m thinking my PID’s need to be more in the range of my Whoop, and not be scared to pump up the numbers…I was thinking since it’s a 100mm build, it might be more like my torrent, but like you’ve pointed out, these brushed builds are a totally different tuning animal.


#186

I only fly brushed 6, 7 and 8.5mm builds and in as much as I want to keep the numbers down, I always end up in 120 to 130 range for P, 40 to 70 for I and D between 120 and 160. My 6mm “Pugsley” always manages to get caught in some sort of prop wash that I can’t eliminate by tuning and I’m already at 130,60,160 and everything comes back warm, so I’ll keep cranking the numbers. :smiley:


#187

On a lark I put my “Qreacher Whoop” numbers in on my new Qreacher X (X100) build…those numbers seem CRAZY for some reason, but OH YES! MUCH better flight. Finally feeling micro-oscillations and needs some more D probably, but flips are much nicer now.


#188

Still in progress, but I’m amazed at how well it flies now compared to my old tune.

Now my Dark 17.6’s come down pretty hot (it’s also 90+F), but I’ve heard the Darks run a bit hot, am I right?


#189

If you want to just throw numbers at it try these:
105 90 125
110 90 130
105 90

That would be what I would fly my first flight on if it were mine.


#190

And it looks fantastic by the way!! I’m surprised you didn’t go for the diamond mount pattern in the frame that matched the canopy holes, but it’s probably a little tougher in your configuration with hot glue on 4 corners. One of these days I’ll have to motivate to set myself up another carbon fiber dip when time permits. Dont forget to get me a canopy list together!


#191

Funny, as I tuned this evening, my numbers were getting close to those…other than D…dayum those are high! :smiley: I’ll give it a shot though!


#192

Got to fly it this morning…and…YAAAAASSSSSS!! I think there is some room for me to play to do some fine tuning, but this is a great start for a 100mm 8.5 build. I was headed in this direction, but I was afraid of running D that high. But wow…it flies great!

I know all those pre-built Eachine brushed RTF’s don’t have PID’s like this and no wonder most people think they fly like crap and go brushless. If they had this kind of tune, I don’t think they’d be so quick to jump.

Thanks again, @NotFastEnuf


#193

Friends don’t let friends fly a bad tune. Lol

Glad it worked out!


#194

Post some flight video and identify any bad behaviors. … We’ll knock em out.

Also post your rc & superrates… just to make sure you don’t have any guano crazy numbers that brushed just can’t achieve. :wink:


#195

and yes, I do like my Pitch and Roll to be slightly different…I blame my thumb. :wink:


#196

Looks very good. Where you get into trouble with brushed rotational rates causing tuning problems is right around 900 to 1000 degrees/s. BUT - that is with superrates high and rc rate closer to 1. In that case the craft is sort of waiting till the last second to push hard and just gets too late of a start. In your case you have rcrate high and srate low which makes a more linear curve that the craft can keep up with. You are either wise, lucky, or just have good instincts!! Well done!


#197

Speaking of blaming thumbs - that’s the same reason I run setpoint weight below 1. I have the mind of a great pilot and the thumbs of a total hack! :joy: It smooths me out and makes me appear more technically precise.

gnōthi seauton


#198

Not one of my best runs, and I’m thinking my batteries (Lectron Pro 700maH 30c) might be getting old, but here’s a run with @NotFastEnuf’s tune. I am at a point now where I’m trying to figure out if it’s a tune issue or just a limitation of brushed…especially on some recovery things like some quick split-s maneuvers which usually end in hitting the ground.

Anyway, this is all Acro (with FPV_Angle_Mix) and still a hell of a lot of fun. :slight_smile:


#199

Looks awesome man!! No way your vtx/rx is on the right channel. I had the same situation the other day when I was actually on the wrong band but had a channel in that band that still looked good up close.


#200

It also didn’t help that I forgot to put up the antenna on my DVR. :stuck_out_tongue: