I developed an “Initial Parameters” spreadsheet and formulas based on Leonard Halls tuning advice and data. This spreadsheet is referenced many times around here, and in particular in the tuning guide discussion. I was thinking of making it more interactive and somehow integrated into the doco, when it was kindly turned into a plug-in by Andras Schaffer.
So apart from that uninteresting history, all its data is available or can be derived from the tuning guide and wiki/doco.
Use the spreadsheet if you want to see how values are affected by different prop sizes, for example. The plug-in is the quick/easy way of getting the correct values into a new aircraft.
EDIT: actually any of us can update the Arducopter and MissionPlanner docs to include the plug-in, and even on the fact that plug-ins exist and can be added to MP
Alt-A should be fully removed, and replaced with a button in MP`s airframe setup screen.
That should make discovery a lot simpler and make it explicit that IT IS NOT OPTIONAL it is a part of the frame stuff.
Thank you Shawn for sending me set of initial parameter for my EDU-450.
I printed it out and compared each with the parameters presently set on my copter. Interesting.
I’m curious about the source of your set of parameters. Are they the product of the Leonard Halls tuning advice? Or are they from a well tuned copter - a product of auto-tune and/or maybe some manual tuning?
My initial parameters were set by loading the parameter set available from Mission Planner for the EDU-450. I then performed an auto-tune, calibrated my battery sensor using the Mission Planner calibration routine, and I’ve calibrated my compasses by running MavExplorer’s Magfit.
So I don’t think I’d be wise to use the battery (volts per amps, etc.) parameters on your list as I seem to have an accurate battery monitor now.
As I understand, some of your parameters such as MOT_SPIN_ARM and similar parameters aren’t really parameters affecting how the copter flies. (I admit my inexperience - I may be wrong about this.)
Given all this - I’m wondering if I might be better off keeping my existing parameters and performing the “ALT-A” plug-in tuning, followed by another auto-tune.
I have a fair amount of equipment on my EDU-450 that might make it heavier than some. I have a RDF900x telemetry radio, full size Crossfire diversity receiver, a FPV camera and VTX, a Naked GoPro, and carry a 5,200 mAh battery. That probably accounts for a couple hundred grams heavier than just a basic EDU-450.
I’ll continue my investigation and education - and greatly appreciate your assistance - thank you!
Hi Joseph
those params were used to build about 30 or 40 EDU450s and all of them (after the first one) just needed a test flight to verify they actually worked as planned, and no additional tuning was required.
The first build was basically flown twice before call it “done”.
First flight with “Initial Params” and INS_LOG_BAT_MASK,1 to gather the FFT data, then HNOTCH values were set.
The 2nd flight was Autotune with INS_LOG_BAT_OPT,2 to verify HNOTCH was set up correctly.
Then a couple of test flights to put it through it’s paces, but nothing remarkable was changed.
So this is good proof that following the tuning guide and initial params actually works. The caveat is that motors/props/ESCs/frame/intended-purpose all need to be matched, as they are with the EDU450 kit Some builds will never work properly
There is lots more knowledge and pointers in the tuning guide than most people realise, and it requires a few readings, and probably a print-out that you can cross off items as you go.
Shawn - Wow - that’s a lot of EDU-450’s! Why so many? I also have a TD-650 in early phases of testing - rock solid so far with just the initial parameter settings.
I’m looking forward to studying the whole tuning arena - and a bit of guidance from helpful and knowledgable people make the job a lot easier.
One thing that’s caught my attention is your suggesting to set INS_HNTCH_MODE to “2” per your suggested initial parameter list. According to the docs on Managing Gyro Noise, this option is for an external RPM sensor - such as what one might have on a traditional helicopter.
Since I’m using an Orange Cube, it seems to me that if I change this parameter from “1” (throttle position based) then I should set the parameter to “4” to enable the dynamic FFT determination.
This isn’t for an “external” RPM sensor it’s for ESC Telemetry RPM as you would have on BLHeli_32 ESC’s. This is the best reference for the Dynamic Notch Filter. If you don’t have ESC telemetry, or Bidirectional Dshot, that provides RPM then either Throttle based reference or FTT, depending on your FC, would be the alternatives.
Yeah, INS_HNTCH_MODE,1 is using the throttle as notch filtering dynamic input
INS_HNTCH_MODE,2 for ESC telemetry is the input. It will be more accurate and better handle changing payloads
Yes, you could use 4 for dynamic FFT - there’s a whole section in the doco on using that and I haven’t tackled it yet.
There’s even a scripted method…
Shawn you help me so much with my Hexsoon 450 as did Dave with stuff as well,my 450 is very heavy as well as I have hawkeye camera and vtx,RFD radio and both a lightware 11cLidar and one of theawesome lightware 45c rotating lidars,I had a mishap with my 450 but am just off to get it airworthy again.
With people like Shawn and Dave around and now you Joe as you ask sensible questions and it inspires me to try stuff as well.
Finally the Hexsoon 450EDU platform is awesomeand Farther Shawn of “HEXSOON” im off to purchase a BL heli 4 in 1 for my 450,thank you for all your help in the past and your help going forward you people are what makes Ardupilot what it is