My 3DOF project is born!

+1 Awesome. The motion is great considering its moving everything including the box structure.

@bsft:
Hi bsft, my initial task was to have my platform plugged to the 220v of my home.
I have 3 220v to 12v converters ready to use but in finish I decided to NOT to use these because the electricity situation on my city it's not the perfect one. Here very often little voltage drop happens and I would to avoid to give that to my poor, little jrk!
Now I don't sure if the UPS in the link you providing have the protect voltage and earthing, if yes so the situation change and about the use of these UPS I can say why not!.
Sure using the batteries you have a good, easy and stable electrical scenario. Ready to use and without all little/medium/big problems that can happen using the 220v.


@fermevc and Fahim:
Thank you a lot dudes! : , the structure it's really solid considering that the 2 main lateral panels are attached and anchored to the main iron structure with a good weight balance (in the max inclinations scenarios it's all ok with no issues.)
All the stress structure+cabin tests went very well with no particular problem with that.

Hi everibody!
Finally the platform can be used without problems, all the major pot/motor joint system issues are solved and now I can use my platform continuously easly!
2 days ago over 2 hours testing it without problems about temperature (heatsink mounted) and mechanical situation.

Finally I can concetrate myself to the motion optimization and today I tested ONLY the vertical force from rFactor (so I deleted the longitudinal and lateral forces).
The track is Japan Suzuka and I find some strange behaviour about the 3 motors.

In the profiler I set for each legs the same Gforce_1:1 math plugin (max = 600 / min 60 / split_min = -60/ split_max= -600) but sometimes I can notice the legs make little/medium/big motion differences. (in same moments I can see the left leg going up and the right leg going down!!! )

The 3 jrk have a personal pot configuration.
In other words I don't made a single pot configuration to load on all 3 jrk but I made the learn procedure (the procedure to set the max and min pot value) with the first jrk, later I made the learn procedure with the second jrk and in finish I made the learn procedure with the third and last jrk.
now you can tell to mek this is the problem, you must to use the identical feedback setting for all the 3 jrk...
but... when I set in the platform start the 64 position for each legs so I obtain all the 3 legs going in the same identical position so I don't know if this is the problem....

You can notice that in same moments the 3 legs making different things when (if everything was correct) the legs should do the same identical movement since I am using the same Gforce_1:1 math plugin for all 3 legs!!! :biw:

All the jrk configuration utility settings except the feedback setting are identical for all the 3 jrk (so the input,motor, PID settings are the same for all the 3 jrk)

Here the vids:

motor duty cycle = 600/600
motor accelleration = 600/600
max currency amps = 12.49
dead zone = 50 (all the previous tests was 80 so now the platform can make more precision movement)



First ONLY VERTICAL FORCE test
Effect 26, vertical force values-----> max = 600 / min 60 / split_min = -60/ split_max= -600
http://www.youtube.com/watch?v=GPX0QvB2JHs&feature=youtube_gdata




Second ONLY VERTICAL FORCE test (sorry, the cam battery die soon.. just 30secs)
Effect 26, vertical force values max = 700 / min 10 / split_min = -10/ split_max= -700
http://www.youtube.com/watch?v=lTjvlpwKm8E&feature=youtube_gdata


Somebody can explain why the 3 legs doing these different movements? Thx!

that the only reason i built 3DOF in one controller. windows is not real time OS..lagging and delay is their closest friends.

2 hours testing is good but the real test will be continous use over many months. Then you will know you have a really robust design. Keep up the good work.

the 3 jrk have a personal pot configuration.
In other words I don't made a single pot configuration to load on all 3 jrk but I made the learn procedure (the procedure to set the max and min pot value) with the first jrk, later I made the learn procedure with the second jrk and in finish I made the learn procedure with the third and last jrk.
now you can tell to mek this is the problem, you must to use the identical feedback setting for all the 3 jrk...
but... when I set in the platform start the 64 position for each legs so I obtain all the 3 legs going in the same identical position so I don't know if this is the problem....

You can notice that in same moments the 3 legs making different things when (if everything was correct) the legs should do the same identical movement since I am using the same Gforce_1:1 math plugin for all 3 legs!!!

that the only reason i built 3DOF in one controller. windows is not real time OS..lagging and delay is their closest friends.

Could be delay , as I am running two jrk's on a frex, so for vertical force I have to put in the same number of force for both so they should react the same. I will have a look further into this myself. Although , you have a flipping BIG delay if thats it. Just to confuse mine further, one motor spins a different way to the other and one is a bit faster than the other. I do not really notice though as I have 3 forces running at once.
As for your's , the base is flexing up and down and may be causing an issue with the motors reacting correctly, as the base moving could be causing the pots to read slightly off, hence the motors not moving evenly. Perhaps make the base flex free and re-test.
:hi:

Salute!... :
DO put some flight simulators in there... that's where it's at!!!... good work... kinda looking forward to having to keep a few airsick bags in the cockpit myself
clear roads, and/or blue skys to ya!

Ahh I was thinking about the use of a controller but now seems necessary.. Today I will to try to raise the task priority of math engine and main program to get a better syncronism of the 3 legs. Theoretically I will obtain a major CPU work for the math calculation and maybe (I hope) a better legs work. But one time I noticed that raising the task priority I obtained rFactor with strange audio pauses.. In fact if I rise the task priority at the max level sure I reduce the task priority of the game.. . Today I will do some test about

Yes dude, this is just the beginning Sure finally I overtake the bad moments when each of the tests = born of one BIG problem.

bsft please check if your jrk configurations are the same for your 2 jrk, motor duty cycle, accelleration, break, max currency and if you are using 2 same model potentiometer (I have the left and right pots of the same model but one is 50K and another 100K, but I don't think this can be an issue) and check If your math plugins in X-SIM are identical, values and intensity.
If all equal then you have my same problem
Today I will work to raise the solidity of the under dais; in fact your reasoning about seems really correct. That dais oscillations in fact change the motor height and balances and this can cause the not correct motors/legs behaviour.

Yeah dude, I am working hard to make an advanced flight simulator x profile. I watched many 6 dof flight sims on youtube but (thanks to one my friend, real 737 pilot) I notice that they don't make the right behaviour.
To simulate a car in a track is really more easy that know, undertand and make a decent flight simulator profile. Sure it's a really fascinating challenge and probably I will go to simulate IL2 or LOMAC (Flaming Cliffs 2) that have a truly better and realistic flight model than flight sim x.

AldoZ,
What type of spring , or what application was it originally for, did you use for the center support?... would you have some specs on it?
Thanks

Hi, I take a Fiat Panda frontal tyre shock absorber, here the Panda pic :


I removed the spring from the shock absorber and I make a custom body design; square cylinder and a square piston. I use lithium grease to make the piston penetration ( :lol: ) more easly sliding.
I weld an universal joint in the top of the body (my universal joint permit me to have more of 30 degrees inclination). This universal joint it's an internal mechanism of a van wheel (the internal mechanism that joint the steering wheel to the wheel (tyre).
I weld to the custom body a new iron component with 4 perpendicular iron legs. Each of these legs have a 4mm diameter hole for fit a 4mm diameter bolt. These bolts joint my new custom shock absorber to my under dais.
The spring is really rigid. I might add more weight to my platform and I would still be far away from a situation in which my spring is too many crushed.
When the body was alone, not mounted in the platform, so I tried to push the universal joint down with all my force for understand how many rigid the spring was and I noticed that I could push the spring just for a 2cm travel and just for half second. This can explain to you how many rigid the spring is.
:hi:

Thanks, AldoZ,
That ought to be an easy and cheap aquisition for me, since i live down the road from a junkyard specializing in foreign cars
(foreign relative to me :rofl: )

Hello Aldoz,
bsft please check if your jrk configurations are the same for your 2 jrk, motor duty cycle, accelleration, break, max currency and if you are using 2 same model potentiometer (I have the left and right pots of the same model but one is 50K and another 100K, but I don't think this can be an issue) and check If your math plugins in X-SIM are identical, values and intensity.
If all equal then you have my same problem
Today I will work to raise the solidity of the under dais; in fact your reasoning about seems really correct. That dais oscillations in fact change the motor height and balances and this can cause the not correct motors/legs behaviour.
Regarding your comment, The jrks are set EXACTLY the same, just one motor is a bit faster than the other. The pots are also exactly the same, both 5k linear single turn. My issue is probably just different motor speed. Not of great concern to me though. I may fix it sometime....I am having too much fun to bother!
Cheers, David.

Hi bsft, if your Issue is a different motor speed so there must be a reason.
In the jrk config utility you can set the speed of the motor with the duty cycle parameter. This option reduce the voltage of the motor. If your jrk have the same values so probably your wires may have some differences. If you are sure about ALL is equal so probably your slowest motor could have an electrical problem.
You can overtake at the moment the problem getting down the duty cycle of the fastest motor until you reach the same speed of the slowest motor!

Thank you very much Raceray! :

Hello Aldoz, thanks for that. I will have a fiddle later in the week. As I mentioned, not a great concern for me, despite my melancholy nature of needing somethings exactly right. I am using two different motors from two different vehicles. I will eventually replace them with two new ones of the same make.
Cheers, David.

Can I get a copy of one of your game profile? I just finishing building my simulator but using K8055 + Sabertooth. I am trying to figure out how to build the proper game profile taking in consideration the reference input from the pot. The math I am kind of getting but the output is the part that I am still struggling with.

Thanks

Does any one have K8055 USO output numbers? This thread has only JRK's info.

Hello all, here the stress test videos from today and yesterday.
No cabin this time, too many hot here.. and needed more room space to work good.

Many tests without problems!
I reduced a bit the lateral inclination using the percent limiter plugin because sometimes happened that a leg exceeded the maximum height point making the jrk stopping the motor.

This happened for motor moving inertia + PID issues. So now I obtain a very more stable motion system but with some little less lateral inclination (10% less).
My actual goal is to obtain a system that NEVER get problems. So I am trying to find the right mix between inclinations and reliability.
When I will obtain the perfect mix so I will point to obtain more inclination changing the motor brackets with more lenght brackets.

The 3 following vids using 3 different profiles:

* The first is the old good profile on suzuka track.

* The second a new profile with a better simulation of the little vertical forces + a better simulation of big vertical forces (example hitting a curb) in Las Vegas city track.

* The tird using the old good profile with some vertical forces adjustments. This video is 3 intense laps in the Las Vegas track.. really cool city track, accellerations,breaking,accelleration,breaking!!! a very good stress test!

http://www.youtube.com/watch?v=jmNmz3WRNIs

http://www.youtube.com/watch?v=HHMcmj_fbgo&feature=youtube_gdata

http://www.youtube.com/watch?v=bWcOqGz7DGs&feature=youtube_gdata




Hi frviana, IF you want I can provide to you my last profile (made for simulate the little track vibrations)but you have to know I am not using K8055.:
aaaaa_ULTIMO_9_C.rn2

Hello all again!
Today I had shoot some on-board videos!
The track is Monza (Brianza) third version... I am trying to do a good racing skill performance........ :rofl:

I was trying to go ahead with the profile, adding features and features but I shoot this vid with the old good setup beacuse with the newest I get still sometimes my legs going over the 100% travel, making the jrk stop the motion.

The profile of this video is not too many complex (and not exceptional) but ensure to me a really good performance without problems.
unfortunately my stern motor battery is discharged before I can shoot the valencia city track video! Really tall curbs and big vertical hit!!

I notice also that my stern battery discharged increasingly in the time and I need to recharge it increasingly too .
Sure now I am making many little tests (more than in the past) but I think this battery is not more in good status. One thing is sure: these car battery are working outside their assumptions. These batteries must to work with a car alternator that continuously give to it the electrical charge. In my system the batteries are without somethings that recharge it when the platform is on so I am thinking about the bsft past post:

Yes I am getting the idea of use another way of power supply for my 3dof.

here the vid, bye all :hi: :
http://www.youtube.com/watch?v=PMsI5JXUQa0&feature=youtube_gdata