Showroom Grigory’s 6DOF 24V DC

Hello!

As the DIY motion-sim topic, which I just recently discovered, turned out to be so addictive that there is no longer any spare time on weekends for anything except cutting, grinding and welding metal tubes, I am opening this thread to share my progress in building a 6DOF rig.

I do not have any experience in this, but with a wealth of great advice available on this forum I hope that I will not make too many serious mistakes in the process.

The idea is as follows: this will be a Stewart platform with 24V 500W DC motors, IBT-2 drivers and arduinos. It will be VR oriented with Rift CV1 for display.


The parts list:

1) Six Transtecno motors (350W nominal / 500W short peak for short periods). Nominal torque (S1 as the manufacturer calls this mode) is 1.12 Nm, peak torque S2 is 1.57 Nm.



2) The gearboxes are 30:1 with resulting torque between 33.6 and 47.1 Nm:



3) Lots of 50*25*1.5mm and 20*20*2mm steel tubes and 50*4mm steel band

4) Some 50*20mm aluminum tubes

5) Six levers are cut from a solid block of aluminum. CTC is 150mm.



6) Six 12mm threaded rods

7) Twelve PHS12 rod ends

8) Three gas springs with extended length of 600mm and travel of 250mm. The force is about 350-400N

9) Six 6127V1A180L.5FS7 position sensors

10) Three Arduino UNO’s

11) Six IBT-2 43A motor drivers



12) Power supplies are not yet purchased, but I am thinking of using four of SDR-960-24 by Meanwell. Each unit supports up to 40A nominally and 52A peak. Units can be connected in parallel and share currents with combined capacity of 90% of the simple sum. The plan is to have two pairs of units, each pair supporting three motors.


Each motor consumes 21A in S1 mode and 29.4A in S2 mode, so three of them would take 63A/88.2A. A pair of SDR-960’s will nominally support 72A (40A*2*0.9) with 93.6A in peak.


I am wondering if I am too optimistic and in reality I would need a separate power supply for each motor. Hope not, as even in this configuration each pair of PSU’s will consume over 12A from 220V line. That means that I already need two separate 16A circuits. In addition, each PSU costs over 200 USD.


The whole rig would look like this:







Position sensors will be mounted like this, with tight rubber tubes serving as shaft couplers:



Some of this has materialized already:




The initial plan was to use the seat from simplistic static driving rig I already had, but after a successful experiment to prove that making welding corrections without removing the seat first is a bad idea I will probably need to get another one

Excellent start!!

I'd say that the Hbridge may be too weak to absorb peaks of current. H-bridge is "only" 50% above rated current

motor 29A nominal
Hbridge Maximum Current : 43A

indeed!

Be aware that a bug can lead to a lever full turn, breaking your pot mounts...

Thanks!

The drivers support 60A for 10ms, so I hope they will manage.

True. I intend to initially place them on the other side of the motor shaft and only move them inside once everything is tested.

I have a slight suspicion this seat was not approved for real racing use.

That's plausible It took it about 30 seconds to get to this condition before I got back with a water hose

if not, maybe you can double them in parallel (like motomonsters)

BTS7960 ICs used in these drivers have built-in current limiting. I am wondering if that would help.

Wow @Grigory fantastic work and something were seeing more and more is People having a go at 6 DOF very encouraging with the level of detail and presentation ... this is by no means a ctritsism but to save your feed back device ( even though it may be 360° device the support arm will get destroyed if for what ever reason the motor runs away or tries to rotate more that 180° ... if your gearbox has dual shaft then maybe it would be better to Put the feedback device on the opposite end to your current CAD sketch .. ( i have come close to loosing a few while testing so that's the reason for passing on what may or may not be a tip ) excellent work and once again thanks for putting it on this forum

Thanks!
Indeed, I am going to mount that on the other side initially and only move it inside once everything is tuned and tested. The reason I want it to be inside is to save space a little.

Mind that these are "cheap" Hbridges
Industrial solutions are much more expensive

You can add some quick fuses @40A to prevent some smoking episodes...

Here is how I plan to wire everything:


As there will be two power supplies each serving three motors, I think it makes sense to have each power supply take just one motor from each triangle side. I believe this should decrease chances that all three connected motors have current peak at the same time:





I would certainly need two separate 220V circuits with 16A fuses. As recommended by @RacingMat, I will also put 40A fuses before IBT-2's:



I want to have a safety mechanism to make sure that the whole rig stops if any motor loses power. Using "DC OK" contacts on SDR-960-24 PSU's to see if any of the PSU's stopped providing 24V and reading voltage after 40A fuses to see if any of those fuses was blown should do the task. Arduinos will halt the motors if there is any alarm:



Logical connections will be as follows:




Does it all look viable?

E-stop will be wired like this:

Good idea to use PSU "DC OK" contacts in serie!

Hi Grigory, Very wonderful design, give me a lot of inspiration.

I'm working on my 6DOF rig also, but I have no weld tool, so I need find someone help me to weld the metal tubes, but I don't know how to use AUTOCAD, so I still have many doubts about the detailed dimension.
Now I'm making my V2 wooden prototype to find the dimension. It's a clumsy method, looks it's not inoperative.

Your sharing give me a lot of inspiration, but I have a question about your motor location, the rod ends have angle limited. I think there is no enough space. Maybe I'm wrong, do you have some special considering for the motor angle?

my prototype, as I checked the base will be Non-equilateral hexagons. like below :

if use equilateral hexagons base, the length of side need be longer, occupy more space.
my current length is about 600mm, how about yours?

Hi Mike,

I am not sure I understood the problem. My rod ends have 45° of angle range: 22.5° each way from neutral. As rotation around the attachment bolt is unlimited, I tried to make the plane of motor shaft lever parallel to the attachment plane on the moving platform. It seems to me that this way I will use that unlimited DOF more and the ones with 22.5° limitation a bit less.

On the top view it seems like the rods are far from neutral angle, but it fact they are not:


I do not have issues with movement when running simulation, but in any case I will be able to rotate/re-positiont the motors when testing the actual assembled rig.

Here is is:

First let me say nice work so far. One possible issue i see is the mounted angle of the motors, in relation to the cockpit, appears to be putting an awful large amount of lateral force on the rods and motors. I suggest turning your motors so the rods/levers line up with the lines I have added. I believe this will reduce stress on your equipment as well as taking advantage of the direction of travel.

I would also suggest rotating the levers 180 degrees.

this is may illustrate a little better what i am saying.

Thanks. Indeed, I just did not think about this. I certainly do not want to put too much pressure on the lock rings on the motor shaft.