[datapool]

First thanks to this great "next" release! I am very happy with it.

I did my first autotune flight today and I am quite happy with the results. Please find attached my SystemIdentState results after the tune. I know there should be some information in it on how heavy vibrations on my quad are or not. Not sure how to read the data. Maybe someome can explain a little the data and what it means, just for learning purposes; I just like to understand the background and details to get comfortable with all parameters.

I am very happy with the PIDs in bank 3 now;-)

[liftbag]

Hi. Can you also report the autotune outer loop results?


Inviato dal mio iPhone utilizzando Tapatalk

[chromvis]

Your Noise level seems to be very good. Although, it would be interesting to get actual description of all parameters from developers who created them.
It looks like Nose is a variance of rotational velocity for each axis, and is not necessarily representing vibrations. I got once nose on the level of 400 in Pitch, and quad even can handle PositionHold with some jerking though. On the other hand just yesterday my NH200 quad (small) start shooting up when I switch to PositionHold - this behavior is typical for high vibrations. I did autotune right away and, to my surprise, got all Nose values below 5. The solution was to switch from OneShot125 to PWMSynch in ESC control - I suspected some sort of resonating vibration while hovering (But it may be just my imagination).
You system show hovering at just 25% throttle, seems very overpowered, could have weird behavior when swithcing from one mode to another.

[datapool]

Please find attached the outer loop results and the full PIDs which have been produced. Flying is very stable. So this is just for interpretation.

In regards to the hovering at 25% throttle "overpowered" topic: yes, I can feel it on the stick, very sensitive and fast on throttle! You have to be very carefull with the throttle stick or the quad will be "away fast";-) Any idea on reducing the overpowered state? Using smaller props?

Specs:
250 quad
~500g weight
revo FC
3S 1800mah lipo, flighttime ~10min
EMAX NANO ESCs 20A OneShot125
EMAX MT2204-2300KV motors
dalprop 5045

[liftbag]

Something is wrong. I cannot understand how it is possible the 25% hover throttle with your setup.
I have similar values only with 2206-2300kv motors running at 4S.
The Outer Loop PIDs seems reasonable, although a bit high.
I asked of outer loop PIDs because seem that very low tau values produces unusable outer loop PIDs.

This is SIState of my Carbon Hornet 2B. Specs are:
ZMX V3 2206-2300kv
DAL prop 5x4.5.3 V2
KISS 24A RE
genuine OP revolution
BOLT 1300mAh 4S Li-HV





Here the AutoTune PIDs.





Here the OPTune PIDs, much better.

[datapool]

Thanks liftbag. I am not sure if something went wrong. I am still happy with the PIDs. However, I am not an expert in regards to PIDs and I have not enough experience to judge if they are really good or not. My feeling is just they are fine to fly the quad. So I am here to get feedback and learn. I was also wondering about the high outerloop values. Before I had a higher P values for inner loop and lower P values for outerloop; found via EasyTune.

How did you get the OPtune PIDs? Did you look for the UOV manually and then used the calculator here?
http://cncmodel.altervista.org/blog/optune-calculator.html

I was thinking that EasyTune would do something similar. Any ideas/recommendations on what to check or change?

[datapool]

Here are my PIDs

a) used before AutoTune - discovered while playing with EasyTune function and
b) after AutoTune

I think from a "feeling" perspective - b) AutoTune gives me better stability and flight characteristics.

[liftbag]

Thanks liftbag. I am not sure if something went wrong. I am still happy with the PIDs. However, I am not an expert in regards to PIDs and I have not enough experience to judge if they are really good or not. My feeling is just they are fine to fly the quad. So I am here to get feedback and learn. I was also wondering about the high outerloop values. Before I had a higher P values for inner loop and lower P values for outerloop; found via EasyTune.

How did you get the OPtune PIDs? Did you look for the UOV manually and then used the calculator here?
http://cncmodel.altervista.org/blog/optune-calculator.html

I was thinking that EasyTune would do something similar. Any ideas/recommendations on what to check or change?

Yes, UOV and calculator give me the better starting point to refine my PIDs. Unfortunately, none of my quad has had good AutoTune results as OPTune method did.

[datapool]

Final results from my end. Tested OPTune, but did not give me better results. AutoTune PIDs have been better on my end.
Due to "overpowered" status, I changed the props from 5045 to tri-4045. Much more smooth on the throttle now and hovering at 40%. Did a re-autotune. Completely happy with what I have, now.

Just to report: test, try, re-tune, change, google something, test, try, re-tune ... until you found your own stable settings; this is what I learned during the last weeks. There is no "golden" way for settings. But it is fun and has a good learning curve. Thanks to all who helped me here so far!

[TheOtherCliff]

A closing note:

There is a form of oscillation that is very fast, with small motion.  Often it can't even be seen.  We sometimes call it "invisible oscillation".  It happens if some PID is too high.  The result is that any throttle stick position above zero acts like it is closer to 50% than it really is.  In bad cases, all throttle stick positions above zero act very close to 50%.

If your quad climbs at 50% power then this makes your quad climb at lower throttle settings than it should.  You can tell that you have this if all your throttle stick positions feel like they are closer to 50% than they are (when flying).  Check 100% throttle too and you will find that it doesn't have as much power as it should since it is also being forced to be closer to 50%.

If your quad usually needs more than 50% power to hover, then it may not get off the ground at all if these invisible oscillations are happening, because even high throttle stick positions are closer to 50%.

For a test, reduce all PIDs by cutting them in half.  If the throttle stick now acts normally then you have these oscillations with your normal PIDs.

These oscillations must be removed from your initial PIDs before running AutoTune.

See the AutoTune wiki article for details.