Showroom FlyPT - 6DOF Brushless DIY Actuators

That sounds great. The load test will be very interesting. This may very well be the excuse I need to buy a 3D printer.

i ordered up a ESP32 - just trying to get it configured

did you find any tricks to get it to count an interrupt on one processor and run the main loop on the other ?

For a small guide on using interrupts, take a look here:
https://techtutorialsx.com/2017/05/09/esp32-running-code-on-a-specific-core/

I have not tried yet...
By default, it seems that all interrupts run on core 0 (https://www.esp32.com/viewtopic.php?t=3689), so we can use core 1 for other code (setup and loop run in core 1).
So from above, it automatically separates execution in the two cores if we are using interrupts.



EDIT: Also this link: https://www.hackster.io/rayburne/esp32-in-love-with-both-cores-8dd948

OK, changed the code to run the position function inside core 0.
It works well.
Is it faster? I don't know. But it should be.
To use another core make something like this:

Code:

TaskHandle_t TaskA;
TaskHandle_t TaskB;


void setup()
{
 // Start runing tasks... (in any core we want, we can make a task for core 1 with higher priority)
 xTaskCreatePinnedToCore(
   TaskUpdatePosition,     // Task function code
   "UpdatePosition",       // Name
   10000,                  // StackDepth, size
   NULL,                   // Parameters
   2,                      // Priority 0=low high number=high
   &TaskA,    // CreatedTask
   0);                     // Core where the task runs (0 or 1)

 xTaskCreatePinnedToCore(
   TaskGetSerial,          // Task function code
   "GetSerialInfo",        // Name
   10000,                  // StackDepth, size
   NULL,                   // Parameters
   1,                      // Priority 0=low high number=high
   &TaskB,                 // CreatedTask
   0);                     // Core where the task runs (0 or 1)
}




void TaskUpdatePosition( void * parameter )
{
  // Must run task code inside loop...
  while (true)
  {
    updatePositions(); // Code runing inside task
  }
}

void TaskGetSerial( void * parameter )
{
  // Must run task code inside loop...
  while (true)
  {
    getSerial(); // Code runing inside task
  }
}


void main
{

 // Some code runing in core 1...

}

thanks for sharing that

its good that it defaults the interrupt onto one core !

Got some time,
it works with 3 actuators!
But I need to clean the mess :

OK, I got some parts and started cleaning the mess:


Well, more or less...


One done, need another one.
And 6 looms/cables...

1 of 6 made:


I'm also printing the remaining parts to hold the actuators, since I got the base:

Very tidy engineering. Can't wait to see some load tests.

Well, I made it today:


Used some weights to hold down sun umbrellas.
They have 26kg each.
Unfortunately, I had no space in the phone to make a movie, only pictures.
I will repeat and make a movie, but the weight was a bit unstable.

I had 3 weights, and it lift the 3 so 3 x 26kg = 78kg more than enough...

That b&d workmate looks to be in just as good condition as mine

So you were lifting 78kg with one actuator? That's great! We'll get a sense of the speed when we see the video.

I have to say that I really love that Workmate. She was stored in the outside for 3 years, at rain and sun, but she's still useful, but I have to replace the wood...

And, I made one of those stupid mistakes!
Until now, since I have been testing the actuators without big loads, I had no protection for the power supply, so in the hurry to see the results, guess what...
78kg going down and some bad PID values = Lot's of EMF.
So I have to send the power supply for repair

What should I use? Diodes? Witch type and values?


(Taken from: https://www.designworldonline.com/using-power-supplies-with-dc-motors/ or https://blog.uk.tdk-lambda.com/uk/2...-external-diodes-be-used-with-a-power-supply/)

HELP!

Motor specs:


Power supply (in red):

Oh nooooooooo. Thanks for pioneering this new tech for those that hope to follow.

Any luck with that PSU? I've been following this development with great interest as have many others i'm sure. We should club together and buy you a new one so you can carry on and produce a reference design for us

The PSU is being repaired. Must get it this week.
Meanwhile, I got some diodes, made the new cables, replaced the pots with resistances in the actuators, printed the supports for the actuators...

Top mount:






Bottom mount:






Waiting for the PSU to make a new load test with video.

Looks great. Not long now before I can justify buying a 3D printer. The 3D printed brackets/supports are a big part of the appeal of this build for me as many other actuator designs seem to require complex machined parts. Its also great that you can share the design as I have very little experience of 3d modelling.

Tell me - does it use lots of filament to make blocks that big? Most of the 3d printed stuff I've seen is quite small. These look chunky. How does it equate in terms of cost of materials?

Thank you very much for the information . I am a bit transparent on the forum but I'm currently mounting a platform almost identical to yours with drive and similar engine and I had not imagined the return of tension. I'm reading you in time ... thank you!

Just received this from China:

DC 12V-36V 15A 500W Brushless Motor Controller

Bought it for testing.
It's cheap, less than 15€, but no heat sink.
Can be used for BLDC with hall sensors.
Will test when I get my PSU back.

Bought here: https://www.ebay.fr/itm/DC-12V-36V-15A-500W-Brushless-Motor-Controller-Hall-BLDC-Driver-Board-Support-Ha/163131509802?ssPageName=STRK:MEBIDX:IT&_trksid=p2057872.m2749.l2649

Hmmm,....

I calculated those values as well for my own design, but come to a different result:
If you apply 0,6Nm to a SFU1605 they should reach a push force of 754N (75Kg). Or did I miss something? The speed of 292mm/s is correct though.

...could it be that you used the diameter (16mm) instead of the radius (8mm) in your calculations?