Showroom FlyPT - 6DOF Brushless DIY Actuators

@pmvcda Hello FlyPT! Sorry for my absence. I think soon I will be somewhat active again now that summer is over but I will still be very busy for a long time.

I'm considering starting over from scratch in the future (probably at least a year from now) so I wanted to ask about your build. If you were starting from scratch, would you use the same motors etc. you used in this build or do you think something else would be better?

I also have to look at the details in this thread on building the controllers, I'm a little afraid of that project! Could I use the Thanos AMC1280USB for these motors if I can't manage to build the controllers? Hopefully I can build them, but just in case.

Also, did you ever say if this setup can work directly with simtools? Just in case you fall off the planet after we build it, would we still be able to get by without your software?

Thanks FlyPT!
Trip

Hi @Trip Rodriguez ,

Have been in your position, some years ago. It was to soon... Hope you can go by all of it the best way possible and try to get the best while you can.

If it was today, I would probably go with the ODrive + BLDC motors.
Take a look at that.

There's also the pt actuators with Thanos boards, they seem good but it's way more expensive.

For the rig design, I would try to make it more compact and accessible. That's a challenge.
Actuators, I would try some good PVC pipes inserted one inside the other with a good diameter that can stand the RC bldc motors inside (the ones from the ODrive).
Or some aluminium ones.

Hi @pmvcda and @Trip Rodriguez

Im also interested in that, because I started to build my own FlyPT rig, with DB59C024035-A and BLDC-8015A controllers. I already own 6pieces of them. I designed PCB for ESP32 board followed @pmvcda plans and circuit diagrams. I want to public my plans and designs with 3D printed parts as well in the future. But if there is a better way to control those BLDC motors, I'd like to know
Thank you,

Best way is my way

Just kidding.
But I have full control on what I do.
What I said above, about the ODrive is more my curious side, because I think my rig might have some advantages over the ODrive (but also disadvantages).
But without direct comparison, it's hard to say.

If you are starting right now, I would suggest taking a look at the ODrive also.
But if you have the boards made and they are working, then keep on that track. I have some updates ahead on my solution.
Nearest one is using UDP for communication. Not sure if it's going to work well, but at least the FlyPT Mover has already the module to do so.

I'm making some rig changes to add new pedals and wheel. So I can't test it now.

EDIT:
Have some people making my rig. Some have problems with the boards. It's easy to change something.
If you get the boards working, please share the design. Never made it. Mine are full of wires, but they work.

I ordered 5 PCB's from JLCPCB.com but I have not build that. So I want to public after I test it.
Where can I find more about ODrive? What are advantages?
But if those ESP32 boards work well I will be happy with that

https://odriverobotics.com/

Search for odrive here in the forum.
A good example: https://www.xsimulator.net/community/threads/wickie´s-6dof-sim.13660/

Thanks FlyPT! I will do some research on the ODrive. Unless I'm mistaken the actuators that Thanos was demonstrating are too short travel for my taste. I presently have about 15" (380mm) of heave and I don't want to reduce it. If anything I wish it was even bigger!

I'm planning to build a really good G-seat so that I can use smaller angles on the sim. I intend to use smaller angles but keep the heave travel as long as possible. Since that is the plan, I can potentially enlarge my platform a lot because that won't change heave travel. The only problem then is it will take up a lot more space and it won't fit where I plan to put it so I'm not yet certain what I will do.

The other issue is that FlyPT Mover won't drive the G-seat right? I'd probably have to run two softwares which is using up much needed processing power for flight sims.

You can use Mover with the G-seat.
Right now there are two solutions:

#1 Add a pose and specific outputs for it.
That pose uses the same source as the one for motion, but you put inside the filters you prefer for the g-seat.
Then send it to an output.

#2 Use directs.
You can make a direct for each acceleration and output that data directly to an actuator.
You can name the actuators as GX or GY ... And use them in outputs.
The direct output can also be filtered.

And you will have a #3 in the future that in reality is a block of #2's

Awesome news! I'm excited.

There are longer actuators 400mm stroke or up to 1000mm stroke supported as well on the AMC-AASD15A... see here a DIY actuator with these AASD-15A servos:

https://www.xsimulator.net/community/threads/wannabes-4-dof-actuator-sim-project.13791/#post-185545

Virtually any type actuator is supported, short or long linear actuators, or even planetary gearboxes... see this video that explains the possible combinations:

Hey thanks for chiming in Thanos! Do you have any kind of rough idea of the total actuator cost for parts to build 400mm? Any idea how much additional cost for something like 1000mm or a little less?

You should better ask @wannabeaflyer2 about that... I can only comment or refer for ready linear actuator solutions from PT-ACTUATOR...

Do you know the price for the 450mm stroke PT-25 from PT Actuator? I will ask him about his build, thanks.

No, I'm not sure about prices for the larger stroke actuators and it depends if you want the universal joints on them or not... its better to ask them directly.
http://www.pt-actuator.com/about.asp?lngLevelID1=42

Make sure to mention you want them with AASD-15A servos on them (includes drives), unless you want to get the AASD-15A servos separate from Aliexpress (perhaps costs more this way...).

Well, I think I just finally nailed down the reason I get that "notchy" behavior in straight and level flight.

If I'm understanding and doing math correctly my motors can't go slow enough. Thanos mentioned this possibility a while back but I was zeroed in on sensors and forgot about this possibility until I was digging through old emails tonight.

My motors are listed as 360RPM minimum speed when using Variable Torque
My gearboxes are 60:1

360/60= 6RPM at output shaft.
6RPM/6= .1 RPS (rev. per second)
360 degrees * .1 = 36 degrees per second

That sounds about right for the speed they move when they are doing the obnoxious starting and stopping instead of continuous fluid movement thing. Again, not an issue for racing but I'm mostly a flyer.

So that cinches it, these actuators have to go. D=

PS- just to finish the maths:
36deg/s = 117.5 mm/s of actuator travel... that's the slowest it can go! If i reduce my torque arm back to 130mm radius that's still 80mm/s of actuator travel!

The AASD-15A servos and servomotor can go really slow and still hold position without ANY vibration. They will move only as instructed. No analog signals or analog feedback sensors to interfere either. All control signals are digital.

Besides replacing the grass in the garden by artificial one, here's what I'm making right now:





That dashboard was a pain to cut and open that hole... needs some paint.

Nice set up !!!
For your info and if you like a hard surge ( brake or gas ) like me, braze those parts ( black plates ) because for my mine , i have a lot of swing (see red cross). Behind also.

Yes, agree - re-enforcement required. When I had a frame on my 6DOF of a similar design - but made of box steel, the steel failed at exactly that spot and it was not the weld that failed.

Could you guys elaborate on what spot you are talking about? I'm a little slow on the uptake!
I was looking for "black plates" but didn't know what I was looking for otherwise.

PS- @Thanos I'm just glad you told me that way back in that email! I might have spent $200 on new sensors to try to fix it. When I went looking for the sensors you recommended in old emails I got reminded why I didn't buy them then (price) and hesitated... then saw your message about motor speed. I know it's a bit of a delayed response, but thanks for that info!

Also that info on those actuators sounds great. I fear that I will have to build my own linear actuators due to cost but if I do I'm going to use either the AASD-15 servos or the Clearpath ones. Seems like I can get more power for the money from the AASD setup so if they will be nearly perfect that seems like the way to go since from a financial planning standpoint I shouldn't be doing this at all!

On that note, I think I'm going to try and get my sim tuned really nice for racing and see if I can get a decent price for it complete and ready to go. It's a good racing simulator, just not a great flight simulator. I might make a "profit" (not including hours!) that way, if I sell the parts I'll lose quite a bit of money. To sell it as a ready-to-go sim though I have to do a good bit of work first. It will need a bottom hex frame that can go with it for starters..