X-Plane 11 6DOF motion platform

I am happy with how the frame turned out, although I am now wishing I had gone for 30mm square and planned for the corner braces from the beginning. I'll wait and see how this one goes. At least now I can confirm the dimensions.

I've got it balancing on a bar stool - this will help me work out the protective platform underneath I still need to build.

The base has to be moved round 60 deg but this is the general layout. At this stage I am intending connecting the rose joints to the inner side of the bracket between two wheel nuts. I'll see if that gives a better angle.

I removed the bottom plate of each actuator and eased-out the corner holes using a 9mm drill. This means that I could fit the bottom plate to the u-joint using the M12 nyloc nut then mount the four actuator M8 threaded rods to the plate rather than the other way around. It would have been extremely tedious to have mated the M8 rods with the holes drilled to exactly the right size.

Getting excited

Managed to get a little bit of time this evening and things started happening quickly. Tomorrow, I will experiment to find out the best angle for the u-joint / rose joints.

The thing I found surprising is how solid it feels once all the actuators were in place.

Coming together a treat that man .. well done .. tick those boxes off one by one

Thank you everyone

The actuator orientation seems to be best when the bottom plate is parallel to the timber frame corner brace. It is hard to draw but I tried - see image 1 and the yellow lines. This gives maximum freedom of movement before the bottom plate hits the frame. By the way the bottom aluminium plate size (in fact all the plates) is determined by the motor mounting which in turn is set by the fan diameter.

The u-joint has two ways it can be orientated. I hope I chose the right one (image 2). I still have to put in the other anti-rotate screw.

I also orientated the rose joint so the inner swivel ball of the joint was in the middle of its travel before I tightened up the wheel nuts and M12 nyloc nut (image 3 & 4) .

I sat in the chair for a while and bounced up and down. It *seems* solid enough (I also made some brmm brmm noises for extra realism )

I have fitted the helicopter controls which took waaaaay longer than I am willing to admit. It is an awkward contraption and difficult to hold, measure etc. I wasn't keen on using a cantilevered method and preferred using 'stays' front and back of the main mounting panel (25mm high density chipboard). All the bolts are M6.

I still need to fit a heel plate under the pedals and need to be mindful of not hitting the actuators. That should nearly be it for the main hardware. Tomorrow I can start wiring.

I've wired up the motors and limit switches plus mounted the panic button. I used a 40mm PVC pipe to mount it to the frame using two self tapping screws.

The wiring is still fairly neat using the three modules idea plus the motor wire length back to the Sabertooth is reasonably short.

Tomorrow will be the self latching relay circuit and finalising the position of the potentiometer / belts.

that's looking so good ...

@baykah
Thank you! I hope you can see your influence on this build

This morning I connected a variable power supply to one actuator to see what would happen...


As some of you have helpfully advised the motor when driving up has quite a bit of momentum and more so when assisted by gravity downwards. I am going to adjust the limit switches accordingly.

Although I didn't show it in this simple video, my new clamp meter actually reads DC current so thank you @DEADBEEF and @SeatTime for your advice.

Edit: at maximum rod extension the bottom plate edge nearest the frame has a clearance of about 10mm so that is good news.

Well done. What a build. Just finished reading the thread!

@Archie
Thank you - much appreciated

Finished! Well actually for now, these things are always a work in progress

Tomorrow I hope to start the testing and maybe even get Simtools up and running.

@noorbeast
Thank you so much! That means a lot coming from you

After a day of testing, I've had mixed results

I made up a checklist for each actuator and worked my way around until they were all 'commissioned'. Next came the power modules and things were ticking along smoothly at that point. Then came the Arduino comms and SMC3 soft start sketch download - all fine.

I had encountered this before but it seems that the Sabertooth wants to be powered up first before the Arduino otherwise it doesn't get the auto-baud command, I'm guessing. At this point one of the power supply fans became noisy like it was scraping on the case. I'll have to open it up and see what the problem is. I know that one of the power supplies was pretty beaten up through the post not sure if it was that one though.

One interesting point was that every pot was wired exactly the same but the SMC utility wanted the second actuator of the pair to have its pot voltage polarity reversed. No problem. The first pair passed now onto the second pair - again the pot wiring on the second actuator had to be reversed. I think it was at this point that when I went to the second motor button on the SMC utility that it drove the ball nut down hard until it hit the limit switch. The bottom screws that hold on the rod assembly hit the bearing but didn't do any damage because the limit switch had cut out power by this stage. Nevertheless, I'll have to move the limit switch even further up. Pretty scary though In the SMC utility I had put everything at zero with the first two actuators so I'm not sure where it got all those crazy numbers - I think the PWM max and reverse were high 200's. The PID settings were also high numbers. Need to be careful when switching between Arduinos.

Edit: another really important point is don't push the motor or power wires into the Sabertooth screw terminals too far or they will touch the aluminium heat-sink causing a short circuit

I spent the next few frustrating hours trying to figure out why the last pair were refusing to budge. In the end I switched Sabertooths with a working pair of actuators and the Sabertooth was fine so it had to be the Arduino. The SMC utility was getting reliable feedback from the pots so the comms must be OK. I tried downloading the Blink sketch and that was fine. In the end I put it on the bench with a CRO and it all seems to be working fine with a nice stream of bits flowing out of pin 13. I even tried putting on some temporary pots to see if that was the issue - still no go.

Oh well, I have two more Unos on the way but it is going to slow things down until they arrive.

Last night I got my key for Simtools , so I will fire that up tomorrow and see if I can get the other four actuators moving.

Glad you’re getting it sorted, though! It seems most everyone hits a roadblock at some point or something doesn’t work like we expect.

One thing about the Arduino/Sabertooth thing is maybe you could just add a delay at the beginning of the sketch so the Sabertooth is up before the Arduino starts sending data?

@Zed
Good idea! That will help with the start up sequence.

The crazy numbers came up again (image 1)

As I couldn't bear waiting for the other Unos to arrive, I went to work and retrieved one of my prototype robots (image 2) I use as a demo for my students. It uses an Uno and motor drive board so it was going to make the ultimate sacrifice for the cause....and what do you know ... it didn't fix it I could move the actuator by rotating the coupler and watch the green feedback line move on the SMC utility ... what the

Lets work through this logically:
1. Replace the Uno and reload the sketch - no go
2. Clip off all the cable ties and visually inspect all the connections - checks out OK
3. Remove the motor wires from the Sabertooth and energize the relays and attach the variable power supply directly to the motor wires - motors work fine
4. Re-check the wiring (again) between the Uno and Sabertooth serial inputs with a multimeter from the pads on the Uno to the screw terminals on Sabertooth - gnd and pin 13 are through to the screws OK
5. Isolate the LED power supply and use the battery only just in case it was going into current limiting - no go
6. Maybe the comm port through the hub (port 9) tried another cable into the back of the pc directly - no go
7. Connect Uno serial output from pair 1&2 with long wires to 5&6 Sabertooth input (very risky no feedback) - success it moved
8. The difference between 1&2 and 5&6 was either the port or the cable. Could it be both cables I tried were faulty? I pulled out the short blue cable from the box that came with the Uno and used a short USB extension back to the hub and what do you know it worked! The cables I was using were from printers so I don't know how different they could be?

Edit: after step 7 I moved the Arduino from pair 1&2 to 5&6 after fitting the 0.1uF filtering capacitors

I will do the brain replacement surgery on the robot tonight

While I had the Uno unplugged I fitted the noise suppression capacitors (image 3)

Edit: also fixed the noisy fan with a couple of spacer washers to move it away from touching the case

Moving on to testing with Simtools.

Just glue some plastic sheet behind the sabertooth terminals to ensure it cannot short out - although why this has not been picked up by the manufacture is beyond me .

@SeatTime
It surprised me too. I didn't actually short anything out but when I was re-connecting / swapping Sabertooths to find out if they might have been the cause of my problem, I had unscrewed one of the motor terminal screws a long way out and to my surprise the wire went all the way in an would have been touching the aluminium heat-sink if I hadn't been standing directly over it looking down.

Just when I thought I was clever and had worked out the cause of actuators 5&6 not working...

Anyway, the USB cable was only one part of the problem - the other part was some apparent difference between Arduino clones. Take a look at the image below. The Uno clone on the left works (now for actuators 3,4,5,6 but the one on the right doesn't (I also had another one of the clones on the right and it too doesn't work - the one that came from the robot). Note the missing crystal on the right clone. When I was trouble shooting I had swapped the working Arduino from 1&2 into the 5&6 module and it was only this morning that it became clear there was more to the issue.

So still waiting for the other Unos to arrive. I also ordered another two like the ones on the left from a Melbourne supplier which might be with me by the end of the week.

Happy days

Still testing.
I have installed Live for Speed and Simtools.
I have patched LFS successfully.
I am concentrating on actuators 5&6 only at the moment and can control them manually using the SMC3 utility:
Arduino SMC3 ver 0.80
Windows utility ver 1.01
UDP port 20017
Comm port 7

I have set the axis limits all at 10% for safety. I have gone to Output Testing and chosen (a) Axis Output.

I have set up the axis assignments as per the attached images but I am not able to get the GE to communicate with the Arduino. Are there UDP or some other settings I need to set up in GE? @yobuddy @noorbeast