Micro tuning techniques - different than larger quads


#141

Bring me up to speed on build specs: weight with and without battery, what motors, and current pids.


#142

90mm
omnibus f3 brushed
8.5 chaoli
750 nanotech
clear hubsans


#143

What setpoint transition?


#144

i do not even know what it is.
Remember, I’m smart,
yet that doesn’t mean I’m sharp.
new things are difficult for my learning curve.
lol!

ive never messed with these sliders


#145

OK that looks like BF 3.0 defaults


#146

You said it’s a slow bounce back so transition is doing its job, but it’s your base tune that’s over shooting too much and transition is slowing it. You could try moving it down to .2, and that will help some - maybe even enough, but we should fix what is at the root of the overshoot.

It’s worth noting in newer versions they have lowered transition to 1 at defaults … that is your stick acceleration when you start to move the sticks … where you have that at 2 is super twitchy ( shows every tiny movement on your sticks). You can lower that to smooth things out but that won’t help your bounce back - it’s unrelated. That’s all personal preference - I’m just mentioning it since you said you haven’t used the sliders.

2 things can cause bounceback in your tune… (this is what you want to know). Either just D is too low, or both P and D are too low. Start by raising D to 5 points higher than P. Also drop transition to .2. Then report.


#147

this is a second attempt, in the first i had p and d in the 90-110 ranges yet I’m not sure if d was higher than p.
Thanks ill try this and report back.


#148

Move setpoint transition down to .2 also. What did it do at previous tune?

What’s happening is setpoint transition is slowing all stick returns to make bounceback less likely to occur from a reduced rotation rate. In your case you have an overshoot that is trying to oscillate and it’s slowing it to just one slow bounceback. This is an awesome feature but it makes tuning hell cause we are looking for little nuances in the very moment that it is trying to hide them. Alternatively you could crank it up to 1 to “turn it off” and see exactly what your base tune is doing. It won’t be a bounce back then - more likely an oscillation. May make it easier to adjust D before adding transition back in at .2 to smooth out the last bit. With me so far?


#149

i did what you mentioned and the wobbles are coming out in hard banks now.
It seems worse.
like a boat on the water.
ill turn up to 1 now and fly.

update:
no change in bounce back.
its a slow calm bounce.
I do a hard fast roll, stop my stick and it slowly lags back.

i can upload a vid to chat if you want to see what its doing…


#150

Put it back on your original pids, move transition to 1 and record that. This will show the characteristics of just your tune so we can analyze pids without any secondary influence. Don’t worry … we will get it.


#151

And just to be certain … you are moving your stick back to center yourself right? Not just releasing it and letting the spring in your tx do the work? Don’t be offended by that pls, I just want to make sure.


#152

i went back.
seems not to be holding its direction now.
add i?

could just have been too low i?


#153

I is very slow to respond … usually not active enough to do anything in a flip or roll. So not likely the problem with bounce back but yes add it if you aren’t holding direction well


#154

What is your flying weight?


#155

super heavy.
65.5
my other FC had no issues tho.

I’m trying from scratch.
added 8 points i and now upping d till tight oscillations?

maybe meet me in the chat?


#156

So after that session do you feel like we made any improvements?


#157

every thing is much better…
how do i change center stick response again.
The weight i can feel the difference. yet the confused on the transition and what it does.
I need more “grip” in center stick with the same mellow weight.

thanks


#158

To simplify weight and transition… think of it this way: imagine a roll… weight will accelerate or slow the initial response to your stick moving. Low is kind of like expo (but only just at the beginning of the move) and high gets twitchy (accelerates into the move). Don’t ever go lower than .75

Transition is active at the end of the move as your roll in this example stops. Low transition values causes a deceleration of the move as tour stick gets close to center again. This helps control bounce back. It will also serve to smooth the ride at low values. .2 to .3 is usually the sweet spot and a super perfect tune will tolerate .5 without bounce back.


#159

If you want more grip but like the mellow feel of where your setpoint weight is at … then your other options to get more response in mid stick is to increase rc rate or lower expo. Rc rate moved up will also make it roll faster at the end of the stick. Lowering expo will leave your end of stick roll rate basically the same but just bump up response in mid srick. If you change either one of these and feel it bites better but has gotten too twitchy, then you can lower setpoint weight more.

To summarize:
-higher Rcrate increases everything across the whole stick range
-lower Expo increases response in mid stick range
-lower setpoint weight softens feel very very close to center stick

Close to center, mid stick, whole stick.

Super rate changes more of the ends of the stick and will not help you with mid stick feel


#160

Just wanted to say you did good today bro,
Tuning for hours when you’d rather just be flying is an exercise in frustration. Especially when you are trusting someone else and communication is coming through a limiting digital interaction and your getting brief instructions with little communication of the intentions behind them. Glad we made progress for you. I admire your attitude.

To summarize our findings for anyone else interested … Low rcrate combined with high superrate - when commanded to roll under constant power - was actually moving the quad in such a way that it accumulated error during the maneuver that caused a bounce back. Changing up to a higher rcrate and lower superrate helped the actual roll to happen as a closer match to the stick commands… accumulated less error in the maneuver, and stopped the bounceback. After that condition was fixed … we then had to figure out how to recreate the pilots preferred feel with a whole new stick map. Tons of tuning fun. Haha

We also learned tuning when a motor is failing can drive you nuts chasing ghosts. :smile: