FlyPT 6DOF/Stewart/Hexapod Interface for linear and rotating actuators

It’s another feature? Where can one switch it on/off?

Check boxes on the roll column. between range and flip.

Ok thanks,
Let’s take the sway in xplane.

Why smoothing and delay are so huge with value 1 of LPF ?

Because when reordering to use simtools order, the low pass controls are not reordered.
Copy paste the high and didn't changed to low...
So in reality you are using a low pass with 180 samples.... the first control.

In the next update, I have some bugs corrected in the filters.
DEMA and TEMA had some signals changed when simplifying the formula....

Could you add on program launch those two features plz. ? :
- Auto connect to the comport
- Auto connect to SimTools

Two checkboxes. You can put the warning dialog so one is aware about the consequences

Thanks

Version 1.7.9

Download in first post.

New version with:
-Optimisation and treatment of some bugs on the filters. Now DEMA and TEMA are working correctly.
They have less lag compared to EMA, but they are slower and go above the spikes.
-Corrected bug where low pass filter values where not ordered when changing to SimTools order type.
-Config file, now saves the graphics selections and the auto gain selection.
-Added Options to connect to SimTools and serial at startup (@hexpod ).

Well done.

The comport auto start works fine.

Would it be possible to modify SimTools autostart and make it wait for simtools to avoid a dialog “not connected, bla bla” in case when the game is launching after FLYpt execution?

Sort of auto trigger when simtools goes on

OK, I will add a check box with:
"Connect to SimTools when available"

Hi @pmvcda
Mate every time I read this post, it never fails to amaze me at how reactive you are to users Request, Software coding at this level , is not for the faint hearted and your workrate is Brilliant for us here and the pace at which this utility is growing is Brilliant .
Said it before , and will say it again . Thanks for the Hard work on behalf of the 6DOF sim Builders Crew ..Hats of to you Sir, Cos you Rock

There are still some errors and things to add...

For example, if you swap the transform, for translations to happen in the coordinate system of the top platform, you have to keep center of rotation at 0,0,0 or it makes a wrong transform (solved for next version).

Version 1.8.0

In the first post for download.

-Corrected an error with the quaternation, when we had center of rotation different from origin and translations on the coordinate system of the top platform.
-Took the chance to make some cleanup on the pose class.
-Added this:



Works like described. Only one time, or you wont be able to stop the connection with SimTools.

Enough updates for this week... my bosswife is not happy...

Switch and open a good bottle of wine, you deserve it ! ;-)

Hello @pmvcda Just wanna say that your interface is working fine. It's just that roll got inverted from version to version. I don't know why I did not felt it right away, but it is corrected. I still don't feel the acceleration, but I do feel the breaks. Still a lot of calibration, but it's the best run of Asseto Corsa so far.

Quick question @pmvcda . In the washout filter, the second value slow down the return to the zero center. In one example, you put it to 500 and this is what I used on my Elite Dangerous setup. But, what is the use of the first number. In your explanation, it seems to be for the high pass filter, but what does it do exactly? It smooth the high frequency?

I'm writing a guide to post today or tomorrow I hope...

First number is the number of samples used to calculate the low pass filter
Second one, the number of samples used to calculate the high pass filter

If you put 1 in the left one, everything passes
If you put 1 in the right one, nothing passes, because High Pass = Value - Low Pass
So for high pass, the bigger the value, more low frequencies can pass
For the low pass, higher the value, more high frequency can pass
For Band stop and Band pass, you use both values, what passes is what you have out of them or between them

The washout uses a different approach.
The value is filtered by this formula:

filteredValue=lowPass (highPass (unfilteredValue))

The high pass is limiting the maximum value of yaw and making it come back to zero (500 is working for me on LFS tests, but it depends on the track amount of corners).
An high spike value could pass (high frequency), but then you have the low pass to remove it.
So the left value in this case (low pass) is how much you want to soften the high pass.
If you put to much, you get a delayed response to yaw, because you are softening to much.
If you put 1 (none), a spike can pass and make the rig vibrate.
I would use 5 in low pass.

The roll over should be used with the filter. Alone is going to make yaw saturate (horizontal line in the graphic).
You can try also a band stop filter to achieve a similar result. Might work even better.

Sorry for fast response, hope there's no error here. I have some work to do...

Hey, thanks.
Again, I will try this as soon as I can.
Filter are not that hard to understand mathmatically. It's their application: What should I use that filter for? How much value should I put? Those are harder to find.

Version 1.8.1
Download in first post

Complicated week at work...
Well, not much here, just uploaded a build with some things I had in the background...
Main change is this:



MMF stands for memory mapped file (UDP for consoles later).
Might have errors....

Select SimTools for normal use.
Select the other options for direct data from the games (only two for testing).
In that case, use 180º in angles range, just like in the figure.
Mixing of surge in pitch needs to be swapped between Project CARS 1/2 and Assetto Corsa.
And maybe some other flaws...
There's no traction loss. It should come from the received data, but needs some changes/tuning.
So yaw is really yaw. Be careful. Needs filtering and anti roll enabled.
This is preparation for future big changes in the software... Nothing to do with current one.

Would it be possible to build a feature (checkbox activated) for "local coordinated" kinematics? (translations dependent on rotations) in opposition to what we have now: "world coordinated" kinematics (translations independent from rotations) ?

After a detailed discussion with @Dirty, this is something which could be highly profitable, especially for flight simulators but also for quality racing games.

As you have the access to the kinematics matrix, it is something which could be doable.

Cheers

This is fantastic news and just what I have been advocating for. Don't get me wrong Simtools is great software but there's really no need for us to have to use it with comprehensive software like pmvcda is developing except when telemetry can't be achieved natively. The best part about Simtools is its fairly easy to learn but its also limited in ways. Its sufficient most times but there's no reason things can't be taken to the next level when more information is available etc. And since, "At the core of XSimulator, is a belief that creativity and knowledge should be open and free" there's plenty of room here for Simtools to have some company.

I am saying again, for a good motion, the filters have to be applied before the kinematics (ON THE EFFECTS) and NOT after (ON THE DOFS). Unless FLYpt incorporates all the functionalities of the "SimTools's "Game Engine like" Tunning Center", "Axis Assignements" matrix, game plugins compatibility etc. there is a huge advantage to use the combination of both.

All we need is ema dema and washout functionality from FLYpt in SimTools and instead of 3 "Tunning Center's" extras, 12 of them. of course in upcoming "expert mode"

Maybe FLYpt could share his filtering knowledge with @yobuddy. In this case it will be a PERFECTION !

The future will show us what is going to happen.

Thanks