Showroom FlyPT - 6DOF Brushless DIY Actuators

If it happens again, with the new versions, please tell me.
Try to remember what was running and modules used. It will help.

ok will have a look at this !

Thanks so much, that seemed to work fine. I'm pretty sure it's because my motor is mounted upside down on my design of actuator so I should be able to wire it up backwards but for some reason it dosen't like it! Anyway now to tune in some PID and all will be excellent. Busy building actuators here now anyway so got time for that. Even watching one actuator move whilst driving some laps on LFS was exciting

Tested 2.9.7 with the 2 boards Thanos and esp32 .
Got the error transmission and the message . But curiously the esp32 disconnecting by itself ( not Thanos board ) . And after the app didn’t freeze and I could follow only with Thanos board . Will check the new release 2.9.8 !
Below a screenshot .

Trying to re-connect but mover 297 freezes , I click on the connect button , just show starting but mover freezes !
Install Mover 2.9.8 and exactly the same problem . To make mover working again and I could after make a few turns with both boards connected , I unplugged the esp32 and plugged it again and it worked .
If that could help

are you sure of the force calculation

0.6 N.m thats 600 N.mm / 8 mm thats 75 N arround 7.5 Kg

I'm using an sfu 1605, that's 5mm per turn and 16 mm radius. So it's right, you are ignoring the ball screw reduction

Update, will need to upload a video at some point but I ended up making my own circuit board and got it made by one of these chinese companys. All soldered up, assembled 3 actuators (taken time during lockdown due to other jobs and the OH being around all the time now) and it works, have 3 actuators connected to mover and doing what I tell them to.

Welded most of the top and bottom frames so just need to print a few more parts and then assemble and solder the last 3 actuators. It is a lot of soldering.

But it's great that it is working. Thanks @pmvcda for this great design.

i am afraid i didnt get what you did with the positioning part of the code

line 7, 8 and 10 down

Hi Flypt ,
I always test my pneumatic actuator for acceleration
I quoted the message you made explaining how to setup a direct or multi direct for acceleration .
First was is the difference between direct and multio direct input ?
in my setup i tested with also simple direct , just to test .
I tested with asseto corsa competizione .
Below in the first pic my setup (simple direct ) .
The problem i'm facing is that i setup on my actuator a range from 0 to 30 , that means according your quoted message that the max acceleration / decceleration could be 30 m/ s2 .But with this setup and i checked in the input plugin , the max acceleration won't go over 10 m/s2 , but my actuator reaches anyway the max limit switch .If the maximum travel actuator is mapped with 30 m/s2 , i didn't never reach the max limit switch with 10 m/s2.
Look at the other pic attached showing the graphics view wtih max acceleration value 7,xx m/s2.
Normally it should be for me more than 7,xx because above one G it's not correct (may be a scale down value ?).
There is somethiong i didn't yet understood why my actuator reaches this limit .
An another thing i didn't understand also is when i move the slider and test manually my actuator , i reach the limit switch above the max travel on the window .is it normal ?

Last point i got some serial error data (esp32). A good thing is when waiting mover seems to work again (not necessary to close it ) . And when I got the error and the serial has been disconnected by itself , i took a screen shot from task manager . If that could help .
Thx

Hi,

Simple vs Multi
Simple direct, allows you to move an actuator with one of the source values. You can apply filters to that value.
Multi direct, allows you to add the effect of multiple source values and apply filters to each one of them, and also to the final actuator movement.
So in reality, the simple is a special case of the multi. It's there to simplify, but it's causing more confusion than helping.
Also, simple calculates faster then multi, but it is not significantly faster.
Honestly, I already thought of eliminating the simple and might do it.

Reaching the limit
That's strange.
Are you sure you have a good actuator calibration?
Let's consider we have a 8bit actuator resolution. Goes from 0 to 255.
After calibration, 0 is the minimum and 255 maximum travel.
This is just a remap of values from one range to the other.
When you say value is 0 to 30m/s^2, you are saying it should send bit value 0 for 0m/s^2 and bit 255 for 30m/s^2.
If you are reaching limit with 10, something is wrong on the actuator calibration.
You can see the bit output to the actuator in the window. And it never goes above the maximum bit value.
If you reach limit with 30, it's normal to touch the switch, since in calibration, the maximum is defined by touching it.
Unless there's a bug somewhere. There was an error in the mapping, but should be already solved.
Just tested your values and they seem ok.

The serial error:
What is your calculation rate?
And in the outputs, can you show me a shot of them.
Are you sure you did not disconnect the cable with the rig movement (just asking because it has happened to me).

It's looking at the table for how many steps we jumped since the last read of the hall sensors.
At fast speeds it can jump more than one step.
Seems confusing.
The table is a map of hall sensors readings, where you enter with last hall read and current hall read.
The result is how many steps there are between them.

you will find attached the screen shot for the calculation rate and no disconnection from cable.

tested one more time the actuator and alwalys for me something wrong .
First the max limit switch reached above the travel of the slider when playing with the mouse .
when touching the max limit switch the motor goes in the other direction quickly (seems ok ) . but after that it's like the origin is completly wrong .
Fot exempler when i move the slider again to zero , The actuator isn't physically near the mini limit switch .
it's like an origin offset !
When you use a rig for the pose you have security to make that yoiu can't reach the limit . With direct input is it possible to setup the same ( dead zone ? etc ...)

i tested and after the calibration sequence (which run well from max to min limit switch) i reached the max limit switch when i was to the position bit output 19770 (not 32767).
May be there is somthing in the esp one actuator program (mover seems to be ok )
edit for some rerason i can't explain i hazve some trouble to make properly the calibration sequencve . Sometimes the actuator stops before hiting the mini limit switch . But for testing the calibration sequence was ok (my several posts before this one )

made another test .
First calibration sequence OK.
connected mover
direct input window and moved the slider
reached the max limit switch at 18946 (look at the pic )
And after that the behavior of the actuator became crazy some forwerd reverse turns at high speed .
Si in confirm that there slider travel not matching to the physically stroke.
And after hitting the max limlit switch behavior crazy and the position slider still the same on mover but real actuator posdition different from position slider . it's like the mapping is offseted ??
Edit 2 : ok i understood that manually if i don't move the slider unable to gety back a normal situation . To make it working again i have to disconnect esp , move back the slider to zero and i 've got a new proper situation . And in this case i 've got again the problem i hit the max limit switch at bit output 19000 like always . Problem is always here .

Flypt , i was thinking about the behavior . Is it normal that when the actuator hits the limit switch , the motor goes at full speed backwards ?
Why not just stop the motor ?

I' m busy this weekend, sorry.
But, can you try to move the actuator with a rig?
You can use the linear actuator rig to test. You need to add a pose.
Try to do so and see if it reaches the limit with the same values.
I can test it, but probably only Monday.

Yes, it is normal.
It works like an auto correction.
If you loose position, he recalibrate with the switches.
That's why I say the problem is on the code, or even in step count.
I have to see it better

ok thanks .
i peharps found the problem . My link between the screw and the motor shaft maybe slips .
will investigate on this !

Made a new one link between motor shaft and the screw ( which pull and push the actuator) , seems to got no slipping. But anyway always strange behavior .
Will try to use a pose as you suggest in order watch if there is a difference .