cx5my's rig

Actually this has M10 perfect fit for motor's spindle, shorter version too.........

https://item.taobao.com/item.htm?spm=a1z0d.6639537.1997196601.20.75cb7484VodNN5&id=597417024004

OK, finally bought the external length 110mm CTC lever, where the bolt pin to pin would be just about 90mm+ a bit, luckily it can be adjusted to less.

M12 hole is bought instead of what I was told by the motor seller, because test fit a M10 and stuck can't go in at all. Lucky only bought one of everything to test..... TRULY .... shouldn't jump the gun.


2 units of 1000watts PSU also bought, instead of 6 cheap plastic 10amps PSU.....
https://item.taobao.com/item.htm?spm=a1z09.2.0.0.72872e8dBYr4ZI&id=629971051218&_u=kb8lko67a99
about US$45 each, actually also cheap hence also a little worry....... but it's industrial category build, so less worry than plastic PSU.

So going towards 6DOF costs up to US$398 (hardware only) at this moment for most of things needed, just need a few more brackets, nuts and bolts to be bought. Wires I have plenty laying around from the DIY Hi-Fi times, probably cost a bit for someone starting from zero. May be at the end hopefully not more than US$ 450 (parts with shipping cost US$93 up to this moment, wow didn't expect this) ........ a lot more than the planned US$ 200 ......... haihhhhhhh, couldn't resist the sim-pit and fell into the trap of when looking at individual part alone doesn't cost that much...... and forgot about the shipping cost.

wrong parts US$ 13, a short 5cm M10 CTC lever with tie-rods and 6 piece of M6 tierod, hahah lucky I suppose, small casulties when DIY..... not counting the Mega2560.....

[edit] now wait a week or 2 for parts to accumulate at Shenzhen warehouse then come to Hong Kong at least 5 days custom clerance and all, low cost slow shipping.

Hi guys.....

more dillema,,,, which layout to use ????

The way I try to follow the STEWART is not yet optimized, but already clearly see the motors sticking outside the intended frame..... was hopping they didn't..... don't look nice this way .......

hence trying the other type of layout, I don't know the term, 6 motors angle equally, a quick 3D check seems to look good all within the frame ........

after drawing 3D (I'm beginer), then realized the CTClever I bought is starting to enter the too long region,,,,, a bit regret, should have bougt slightly shorter i.e. 100mm or less next time,,,,,, and the realization of even with adjustability, their length clash with other CTClever when too long..... oh dear......

tie-rod-length adjustable between 150mm to 180mm, just clear the motor when maximum down, phew, lucky miss.

The parts are finally coming, estimated arrival June 4th, else the week after .

Anyway I want to make some update:-
1) bought US$ 5.2 for 2 pieces sheet metal (nett) 40cm x 40cm x 1mm thick with 10mm holes all over as base to mount motors , because I haven't decided how to mount the motors and can't afford to drill the seat-rig countless times and loose integrity ......... shopping of this piece took many days of thinking and rethinking and considerations.....
https://item.taobao.com/item.htm?spm=a1z09.2.0.0.2a6f2e8dCB7J8r&id=612763880937&_u=vb8lko6a9ad

originally 8mm holes were sought after and thickness but couldn't find cheap readily available, custom cut cost more hence gave-in and bought the quick and dirty cheap US$ 5.2 net

now I hope the 10mm hole will give some freedom, freeplay when mounting the motor as well as POT's L bracket, because they sure are not aligned naturally.......

2) also bought POT's L bracket, all 6 pieces for US$ 0.45, but will need to drill the 5mm track into 7mm hole to nut and bolt it in.
https://item.taobao.com/item.htm?spm=a1z09.2.0.0.2a6f2e8dCB7J8r&id=601020366973&_u=vb8lko60229

Just because thinking of a quick and dirty nuts and bolts plug and play way and shopping for the above two last secret recipe took so many days to search and search......

while waiting for parts, re-trying Mega2560, still couldn't work, then used a multimeter to check each output voltages and seems to have some clue, then tried, but motor still doesn't spin...... along the way plugging in Uno_R3 ensure motor spin then unplug, change com port in smc3.ini, then replug in Mega2560, swap them so many times, with so many loose wires...... now the motor only spin one direction ......... saw another member here said probably one chip of BTS7960 got shorted due to regenerative voltages....... oh boy ..... just when I thought I had all the parts ..... now looks like I'm short of a BTS7960...... well got to wait for the rest to come and cross test.........

still so many concerns and worry (mostly I forgot 6dof holding the 100kg load up at equilibrium even when not moving at all is using power, I start to realize now a lot of power and continuosly just to be there not moving, oh dear, somehow I forsee .......... this is going to so big 6dof screw up ..................

well it's beyond the point of no return, everythings bought, just keep soldiering forward .........

already thinking of fallback plan to must use universal joint, i.e. 3 motors 2DOF+traction lost and be done with. This is the sure least power burden way....

OK, see you guys in a week or two .....

Hi guys, some update, materials came June 7th Tuesday evening, start unpack and mockup build,

• 8 holes out of the 28 10mm perofated metal net needed drilling, but ended up using a metal sheet cutter instead, because can't rely on exact 6mm holes and positions, needed a lot of room freeplay to move adjust match to other bolts points etc.....
• 4 out of the 28 holes are blocked by APIGA frame thick metal and are skipped completely.
• power cables 1 meters+ each extended, not shown in picture
• The motors are placed tighly side by side, unlike 3D simulated with a gap, because realized CTC lever only turns one side, i.e. 7o'clock to 11o'clock one half of the face.

June 8th & 9th evening, only manages some wiring, dam, it's hard, wiring one BTS7960 was fast and easy, when all 6 put together, CONFUSED the hxll of me...... took whole night just to reach this point, Still need to figure out how to secure them, else sure accident moving then short circuit...... don't dare to power any of them on till everyone is secure, no quick-test allowed.

No picture, 2x 1000W power supplies wired up the wall plug sockets and 3x 1 meters+ long preparation for BTS7960, 24v PSU can undervolt t0 12v, hooray. Pending soldering all the terminations 24x oh boy.

June 11th Saturday night complete the 6x motors to metal net with CTC lever, (after wrestling the metal net and bolts, lazy man short-cut quick-&-dirty method not paying dividen, not fast not straight forward, whole day spent.)


June 12th Sunday, only managed these 6 BRACKETS drilling 8mm hole using a 6mm drill_bit (wrong bit), toook 2 hours just to find a way to hold them hold tight in order for drill to grind it,

• Biggest hurdle was manages to unbolt 4 out of 6 screws on the motor to secure the pots' BRACKETS, may be 2 hours wasted, countless attempts. No picture yet. I recall 2 came out immediately, no effort, the other 2 took forever...
• Last 2 out of 6 screws not yet loosen, around another 2 hours wasted, countless attempts, the heads already worn 50% and worry might be stuck forever. Even fire from lighter was used to heat the outside casing in hope expansion would loosen it up but no luck. Will try again later.
• Might resort to using bottom layer screws, luckily unbolts immediately and easily like 1 second, not the best of choice, but will work if needed.
• My entire body aches just from these.......

Next got to figure out how to attach the motor spindle to pot, have a couple of ideas, lazy man style quick and dirty again, but might end up lousy messy and regretfully time consuming.....

More next, got to wire up all the pots ...... with 6x 1meters++ long cables to Arduino Uno.

Still need to buy a 2nd Uno R3, buy a replacement BTS7960, buy 6 diodes/rectifier protection for BTS7960 against back EMF???

and more ...... that I haven't considered/planed..... to be reported in a week or two......

wrong keyword, "back EMF" will always occur when inmotion, hence I think I don't mean back EMF, but instead I think some of you guys has been talking about actual voltage/current generated by the motor itself when it's motion is sudden and large, i.e. jerking/shaking and electricity is sent back to the motor controller, in my case the BTS7960 and killing it, especially when the voltage/current exceed it's threshold......

obviously I don't fully understand it and am still studying this electrical topic.

currently thining of buying the diode circuit only CNY￥2.5，60V 10A
https://item.taobao.com/item.htm?sp...VYHDIh&id=602177108588&ns=1&abbucket=9#detail
its says no polarity, which is what we need I think, because motor generates +/- depending forward/reverse. I can't simply buy a fixed forward polarity reverse voltage/current protection as used by solar/wind power generators.


can't use





can't use these 3, because they have fixed polarity.

Or am I big wrong..... welcome to point out my mistakes.

Hello, good morning, hooray, 50% progressed ......... quick version story is just need to buy a 2nd Uno_R3 and lucky didn't break (more like haven't broken yet, this day will come I'm so worried) the first BTS7960 while testing with Mega2560 to find dislocated pin vs. Uno_R3.

just 3DOF out of 6 quick video



Below long version story: -

Last weekend stuck at removing a screw to mount the POT's bracket (wasted few hours achieve nothing):-


then borrow BIGGEST screw driver from office, still doesn't work.
use SAW to file a straight grove in hope to use FLAT MINUS type screw driver ...... too hard of steel to file, gave up.


finally tried CLAMping pliers and WORKED !!!! in few seconds, OMG can't believe it, was so worried the head might tear off hence didn't try at first.... so regret now, should have just done it last weekend....... could have saved a couple of hours.......


next how to attach something to read motor's angle, couldn't sleep well a few nights....... so near yet so far ........ then finally last saturday just force try below idea and luckily it worked "quick and dirty" wins one round. Mineral water cap with takeaway hot drinks wood stick......


No picture next. The wiring was so hard, took 2 full days (and more) just to get 3DOF..... mainly because not enuogh different colors wires........ force to use repeating colors wires took a lot of time trying to seperate their functions, i.e. red is positive, next best thing is orange. Black is negative, next best thing is brown which is originally attached next to black.... etc etc

Hello guys, Some questions need your help, where to find a reference 6DOF stewart axis assignment table, i.e. which axis should be assigned to pitch, roll, etc etc..... (millions of thanks in advance)

although last night I managed to figure out roll, pitch, heave, sway, that's only because it's 3DOF, I could not imagine when all 6 axis in motion, which should be which, hence thanks a lot in advance for the help to be received,

I probably would have read it already somewhere, but recent weeks of actual hands on have completely depleted my memory and am very brain-power deprived, hence a little nudge in the right direction is much appreciated.....

in the meantime my head is preparing how to do the travel limits, tried it during one motor mockup, couldn't get it to work...... for the moment I'm cheating using 255 chipping and 255 max to limit the CTC lever angle range, then push the level to the max knocks the POT's bracket very weak (version frame below very strong), then manually rotate the POT's angle to match the CTC lever to over the MAX into the "motor_OFF" state, so to prevent CTC lever over powering the bracket and damage etc.

apart from shopping for 2nd Uno_R3, I'm also partially in "REVENGE" mode to get 36 color coded wires, 2 or 3 meters would be more than enough, CNY$48/meter. This is the most UNEXPECTED MUST have part/material, could have saved me around 12 hours pondering of my limited 10 colors only how not to clash functions........

but before "revenge", let's focus to get it fully operational first, which means NEED some breakout SOLDER circuit board, no more PINS plugs, because motor 2 and 3 didn't work last night was because of LOOSE pin for both ENA&B x 2 motors, as well as wire BROKEN away from pin due to wire stripper, etc etc..... lack of planning, lack of experience, lost times again trouble shooting, re-wiring, (around 2 hours).

OK, what else have I forgotten,,,,,, somehow it feels like I missed something again ........... welll to be reported next time....... Thanks again in advance to whom ever gives me to nudge forward........ much appreacited.

last night 1st time 6 axis moving together ...........

not finished, not ready, still getting grips to head and tails of aligning CTC lever unorthordox way.

June 25th morning, just put chair on top to find out where can plug and play nuts and bolts.....


and trial plug and play the front tie rods


rear two holes see through sky WHITE ROUND


then a check before power on...... or was it already power on ???
this looks right, but is wrong because lever length misalign fron two with center two, so this end up just view for pleasure, not working in reality.



after install all, power up test, all crashed trying to go to equilibrium level when total free there was no chair, which now with chair limits their angles against each others and some CTC lever smashed into the POT and broken the tiny WOOD to read angle.


the chair either fell down when power off or power on to go equilibrium and smash BENDed the POT's bracket ... oh dear .......


Just a quick check overall - then off to lunch,,,,, because stuck couldn't figure out how next , ahhahahha.


one of the checks of heave holding position shaking.......


another check after wrong settings in axis assignment and %%%%


more trial and errors to come.

highlight rear tierod connection to motor.
anti-clockwise motor reverse POT's (+) & (-)


The right pot's signal was haywire last night, hope is my bad soldering, tonight rework.

So near yet so far......... not finished because quick and dirty lazy men method backfired, so much rework......... big big lesson learnt ......

Motors' forces so strong bended 2*1mm metal sheet so much, later found out I didn't tighten the M8 bolts on the far edges to hold metal sheet to rig frame ....... completely forgotten about this .......... after tighten is OK (no picture because just tighten)


Front flex so much because POT's signal suddenly irratic and causes motor to go haywire crazy finally CTC lever smashed onto POT's bracket again ........ possibly causing intermittent irratic signal....


Somehow I'm getting more and more thoughts of releasing the motor's where they are now and slowly shift them to new position that naturally flows with the CTC lever and not bend/force them so much.....

Motor 3's BTS7960 got up to 75.9 Celsius direct on the big chip MOSFET???, only 14 volts input, not even the full 24v. On the side direct on the heatsink opposite side got 53C.


Motor 6's also got hot, (middle unit), 64.9C.


Motor 3 and 6 are at the rear seat and are both no.3 in SMC3 with the 31kHz Fpwm, KP350, PWMmax200,

While other motors are 20kHz Fpwm, KP300, PWMmax150 are only warm.....

Apparently these problems of OVER-SHOOTing ctc_lever are because my mistakes of power cable to Motor 5 and 6 swap wrongly, but reference POTs are 6 and 5, completely wrong way respectively.

hence report of last weekend's effort / achievements / problems to solve are:-

1. Found and correction of power cable motor 5 and 6 versus POT signal 6 and 5 respectively. [me big IDIOT ]
2. Tighten down 4 x M8 bolts of the metal net to bottom rig frame + some quick and dirty cable ties to assist holding them down not bend up so easily, July 9th night tested my weight load SAT testing found very helpful no more bending.
3. And tested sitting with full power HEAVE max height ...... and ..... hahahaha it completely could not hold my weight 70kg, the chair just SANK completely so quickly 1 second, but not full slam dunk which would crush the motor's casing below, because I suspected it would not hold at all, hence already prepared my legs to decelerate the fall in hope the reverse-voltage does not kill the BTS9760s [but still surprised it was weaker than expected].
4. It finally ALWAYS could HEAVE up and down, literally 100% repeatable..... no longer haywire going wrong direction because motor 5 and 6 powered wrongly.
5. It finally PITCH and ROLL correctly 100% repeatable........
6. 4. & 5. achieved because followed yobuddy's 6-axis sample more closely, obeying rules like: -
   • not lazy on POT's direction, i.e. SMC3 up then motor must up, Should not rely on using SimTool's reverse feature.
   • optimized axis assignment, where both SMC3util's motor 3 are at rear seat, previously did not consider to match them, because around 1st milestone June 24th rear seat was motored by a Fpwm 20khz plus a Fpwm 31khz, and I think I notice the CTC lever up and down angle/distance are difference, I don't deny some of it may be my psychological misconception.... in fact are more likely simple mismatch of POT to CTC lever between the 2 rear motors, mechanically not SETUP perfectly mirror.
7. My brain could not figure out SWAY SURGE YAW, hence not yet done.
8. signal wires loose, causing intermittent no motor movement,

• sometimes POT seems to be bad or dead.
• sometimes POT signal wire loose.
• sometimes BTS7960's ENA1 ENA2 PWM loose
• the convenience of pins and sockets are blasphemy ....... must SOLDER, pending new wires to come..

​

ohhhh so near yet so far ..............

Completely agree with you, instead of pushing the chair up, they now just try to tear apart left to right, I started suspecting this problem since last weekend......, hence last weekend adjusted the tierod on the ctc lever closer to motor's spindle.... in hope for a quick and dirty solution......

In fact I thought to relocating the rear motors instead of side by side in the middle, move them to far out edge by edge ...... but before making any big architectural changes, am going to try to make this setup all the work, learn all the mistakes and make a fundamental corrections in the future.....

Hello @cx5my ! In place of mineral water caps, try to use such flexible couplings.
The water caps certainly look more fun)))
https://aliexpress.ru/item/32970922...2588684&spm=a2g0s.12269583.0.0.419855efXrXVXd