FlyPT Mover

Here are my thoughts:

• It's great that you have deactivated all other channels! It's much easier to tune a single channel at a time.

• The F-16 is hard to tune, because it's such an agile fighter and FBW on top. There's a reason why fighter aircraft normally use fixed base simulators with some moving parts in the seat and harness only. Stewart platforms just lend themselves more to airliners and general aviation aircraft.
• Try a LP(HP(Value)), the HP part will return the platform at the end of a roll much faster. It may even send it off to the opposite side, that's OK.
  
• Try without the logistic filter first and get a good baseline compromise setup this way. Then add the logistic filter and see how it changes. It will make small deflections bigger and big deflections smaller. It sort-of "compresses" the dynamic range of the signal. Thats why it can be used as method of signal compression in digital audio engineering.

What's that joystick you're using? Is that an FSSB R3?

@Dirty and @dureiken ,

Added recently some new filters.
There's a WASH and RWASH.
It washes the values to zero and should not go beyond zero.
RWASH is specific for rotations. It goes over the -180º / +180º.


Some other parameters without the low pass:


In action, filtering a square wave with noise:


Made it thinking specially on traction loss for cars, but could work as a low/high pass replacement.

The parameters are:

#1 The value we want to filter
#2 Smooth entry
#3 Smooth exit
#4 Amount of values history to use

Please try it. And remember, use RWASH in rotations.

Edit:
It might need a gain, because amplitude response is related to the history of the values.
Testing with a square wave might not be a good idea...

Yes it is

@Dirty @pmvcda I share here my tests for others, if it can help

I tried RWASH with position, I really like the behaviour, except the "spikes" due to full roll. Rollover seems useless. It would be awesome without that ! Entry and exit filter are exactly what I want.



I tried LP/HP as Dirty said :



I like the behaviour in entry but the rollback effect at 0:18 or 0:23 seems too exagerated for me. Could I keep limiting just the way it goes at the opposite ?



I tried with HP 5000 instead of 2000 but it's nearly the same. And with LP 1000 the return to 0 is too smooth for me.

To sum up : RWASH in position without "spikes" due to rollover would be awesome.

Thanks for help and time btw.

You can also try the variations of LPF and HPF in mover that I find very interesting

While properly set, they are supposed to stick closer to the original signal without producing overshoots

I used RWASH by itself and it works perfectly in a full roll - no spikes.

@pmvcda, can you sketch out exactly what the WASH filter does? Is it internally implemented as a HP/LP combination?

It's a combination of:

EMALP
ADDX that adds the last n values
RDELTA that gets the difference between the two last values (R for rotation, because of the -180/+180º gap)

So you get first the delta between the two last values
Apply a EMALP in it (it's the entry smoothness)
Over those two, use a ADDX to add the last n values obtained on those two filters (history)
In the end, another EMALP for the exit smoothness

Something like:

EMALP(ADDX(EMALP(RDELTA(VALUE);entrySmmoth);amountOfHistory);exitSmooth)

Those are new filters.
RWASH and WASH are the same, but R is for rotation
Same for RDELTA and DELTA

They are not documented in the online manual yet. Lack of time...
I want to test it more. Maybe some variations on it could be better.

@Dirty , forgot to say, that RWASH and WASH should be used with position, not speed or acceleration.

Could you share your mover profile ? thanks

ok thanks

do you fly jet like F16 ? or is it in FS2020 ? thanks

Its for DCS F15 and some others. Works also OK in MSFS 2020, with some changes.

Some news for next version.

Been working on the outputs, concluded a major update to optimize them and ensure a constant frequency on the data output.
Also changed calculation modes in the rig module that will influence all the rigs. Let's hope for the better.
Thank's to @hexpod for the help, testing and givind some tips important to find the light at the end of the tunnel.
Here's a video showing it at work:

Constant motion, even out of range.
More when it's released.

Also working on this:


and some new sources.

Hi FlyPT
such a great improvement! In the beginning it took me long time to get rid of "sticking" actuators when out of range. But it looks you solved this in a better way! I am looking forward testing the new version.
Wickie

Hi @pmvcda ismit now a good time to get one of these WT901C-232 modules , seem like your almost ready to release that update to Mover , seen these on Amazon at reasonable price so had to ask .... this version is 9DOF can the 6dof work with the new motion compensation module in mover or does it have to be the WT9xx version . Cheers

Still testing.
Don't get one yet. I will post what I think when I have a better idea of the results.

Hi @pmvcda too late lol thought for once i'd get in there before the mad rush begins , your track record is good mate , so its a gamble i had to take. picking it up tomorrow from local amazon drop point, so worse case it's return to sender Got my Quest 2 , My Occulus and hopefully with this option when its ready. it will make a nice difference .. No Problems ( Lol i did not want a repeat of the RTX XXXX cant get one for love nor money situation )