FlyPT Mover

If I'm right here Simtool Output testing can only be evaluated by whether the motors move or not when I move the sliders, but that does not happen.

Can you guide me more on the defualt profile, I'm not sure I got that.

As for the baud rate I tired both 500000 and 115200, the reason for testing with 115200 was to match the sabertooth baud rate, it was one of many trials.

@pmvcda & @noorbeast I have good news here, I found the problem and got it fixed, it turns out that I was using the the windows utility for tuning the PID in an incorrect way. In simple words I thought the utility is only to drive motors manually till you find the right values, but in reality it turns out it will re-write those values simultaneously to the arduino EPROM, so what I was doing at the end of each calibration session I was droping the power down to 0 to stop the motor and move to do something else, this means I was basically telling the arduino to use zero power to drive the motors.

Next I will do some final adjustments and see how it goes...... but I have both MOVER and Simtool working now.

I'm going to make some tests today, with other PC's.
But do you have the full version of NoLimits2? It works only in the full version. I think the demo as no telemetry.

Are you using the parameter --telemetry to start NoLimits2?

Hi , yes for all . As described ,it works great with Simtools plugin and simtools using your first flypt hexpod software to drive my sim rig !

Could you go back to ip 127.0.0.1.
That should point to your own PC.

I have no problems from my side. Just tested with another PC and everything was OK.
Going to test with another one.

I did this and it works ! But sometimes it doesn’t work , no data sent , and i don’t understand why ( check boxes connect and auto detect ?) . If i find something why I Will report here !

That's so strange...

Do you have SimTools (game manager) running at the same time?

No

Of course not ;-))

FlyPT Mover Version 2.6

Download in the first post of this topic!

Changes/Fixes:
- Corrected the logistic filter again. It was bugged on the last build.


So now, we specify the maximum value and how "fast" it goes to that value.
Setting an "how fast" of 1.0, is the fastest possible, without getting results above the input (see above image)

Bellow 1.0, goes slower:


Above, works like a gain, climbs fast, but it's limited to the maximum range:


- Corrected problems to open files. Hope it's working now.
When opening a file, we read everything. If something is not compatible, it's not going to open, but all that's compatible, opens.
- We can now merge files. So you can keep a file just with you rig setup and merge a setup of another user with your rig.

- XPlane allows the selection of the type of landing gear.
What shows in the options, is the order defined by XPlane for each wheel.
Selecting the correct setup, produces a pose based on the load of each wheel, so we can have rotation besides heave.
If you select "Basic 3" (the default), we use just heave. That's a mix of the first 3 wheels.

Same for the other basics, that use just the number of wheels specified in the number.
- Reworked XPlane controls.
- Added separate filters for lateral and longitudinal accelerations while on the ground (to remove spikes when braking)

- Some minor bug fixes all around Mover.

- Some internal changes for future features...

Download is now available. (sorry)

Instructions to setup XPlane11 for Mover 2.6:

In Xplane, you need to select what you want to send through the UDP port:


In the left, you have the numbers to identify the info, in the right column, check the ones you want to use. And you should check:

#3
#13
#14
#17
#36
#37
#63
#66
#67
#127
#135

In the IP, keep 127.0.0.1 if Mover runs in the same PC, else, insert here the IP of the PC where you have Mover.
The port, should match the one used in Mover. Be careful here, XPlane has many 3rd party applications, be sure the port is not in use by another application.

When you add the XPlane source to Mover, it shows two windows. They are two sources, one when the aircraft is in the air, and the other one when it's on the ground (the one with options).



They display the received values, altered by the transition made from air to ground or ground to air.
Those transitions can be tuned bu changing the respective speeds. A value of zero generates an immediate change, while a larger values, makes the transition slower.
I would say that from air to ground, transition should be fast, while ground to air, should be slower.
Looking at the source, we can see two fields with the amount of each source used on the transitions (%). They are yellow when the transition is occurring.

Landing gear option is a way to specify the type of landing gear.
This is used to generate a pose from suspension.
The options are the sequence used to define the landing gear in XPlane.
For example F RL RR, means it uses 3 wheels, where the first one is the F=Front, second RL=Rear left and RR=Rear right (FL=Front left and FR=Front right).



Choosing one of those options, calculates the travel in each corner of the rig, generating heave, roll and pitch from that data.
We also have Basics. with those we choose one depending on the number of wheels, but they just generate heave.

There are also two filters to filter the lateral and longitudinal accelerations when the aircraft is on the ground.
This is used to eliminate the spikes sent by XPlane (bug from XPlane?).
In those filters, I recommend the use of the the filter SPIKE (since 2.5).
The spike filter is used to remove big changes in values.

We use two values to define the filter:
First one is the diference accepted, for example 3.0
Second one, amount of diference used, for example 0.05 (5%)

To explain it better, maybe use an example with the above values:

- We receive a new value, for example 2.3
- We compare that value with the previous filtered value (it was for example 2.2)
- The diference between both is 0.1, since it's smaller than 3.0, the accepted diference, the value passes, and the filtered value will be 2.3

- We receive a new value, it's 10.6
- Previous filtered one was 2.3, like we saw.
- The diference between both is 10.6-2.3 = 8.3
- 8.3 is more than 3.0, the accepted diference, so the value is not accepted.
- The filtered value is going to be 2.3+0.05*8.3. What we see here, is that the 0.05 is 5% of the diference, this means, that if the value is out of the accepted diference, we use only 5% of that diference.

Ok, that's all for now.
Going to make the final filters document. Was waiting to have the filters bug free (logistic).
In the next version, I will start to include a manual in PDF.
It will be written with time, and sections will be added when they are ready.
It will be also available for direct download on the first post of the topic.

Well, I forgot this:

Extra XPlane data:

With the creation of the XPlane source, some extra info was added to sources.
We can receive the following info now:

-Stall
-Engine torque
-Flaps
-Speed brakes

besides the usual one.

That info can be used to generate vibration on the rig through the use of the pose "Noise of value".
What it makes is generate random noise from a received value. The received value alters the amplitude and frequency of the noise depending on it's value.



Select the sources and value you want to generate vibration.
You can use a filter to change the behaviour of the value.
Select the amount of rotation and translation used in the vibration (maximum amplitude)
The "use value range from" and to is to specify the range of the selected value that defines the minimum and maximum vibration. Beyond those values, vibration is constant.

Play with the toy there's some interesting cues there, specially in aircraft's.

- Suspension code is excellent !
- Noises for RPM and stall are great
- Torque on roll with additional pose works fantastic
- I can attenuate the longitudinal spikes

Bravo

Unfortunately I can’t get the lateral spike working. Really strange behavior.

We need to use a pretty big magnitude on the longitudinal and lateral cues in order to build a well working coordination channel.

This is amplifying the xplane disease which makes the spike attenuation really hard.

To construct a good transition between the rotation and translation for coord. channel, we really need to feel the translations. In order to achieve this in our limited space we have to exaggerate them using a value between 30 and 60 and not to much of LPF.

Those spikes can have a magnitude up to 0,5 G.

You imagine when you’ve achieved a good translational response those kicks make you jump out of the seat.

A nightmare.

I m sure you understand the dilemma :-(

We are going to tame the beast...
Believe me, I can be really persistent.

Also, looks that I might have something wrong on the landing gear. I have some complains about a disparity to big between wheels. Must be something stupid I left behind..

lol

Well... with the recent additions I must admit it’s the best motion experience so far.

For the moment I am just adding more orders to the LPF on the sway and it’s definitely enjoyable. Pity I can’t add more abruptness to the lateral accel because it’s really produces a great feel.

In between maybe you could attack the overspending turbulence for flaps and maybe landing gear and speed-brakes so they stop the shakes when operating in low speeds and on the ground. ;-)

We have the flaps percentage, we have the airspeed it should be doable.

Back than @Dirty sent me this picture. For the Cessna the limit is 100kt


Best.