New 3doF platform construction

[Roland]

Last couple of months I have been working on a new type of 3doF motion platform construction. The idea is based on Inmotionsimulation 3doF type, but I made several modifications, to achieve cheaper motor drive system with fast reaction speed and integrated weight balancing scheme.
A bit more complex than my previous platforms, so I have first build a wooden 1:2 scale model to test the functionality, which looks quite OK.

The final 1:1 unit will be build from metal, aiming for 200kg payload.

You can see the 1:2 scale model in action at:

http://www.youtube.com/watch?v=NtzIzNf3crQ&feature=player_embedded

I will also show this unit at upcoming FSweekend in Lelystad, the Netherlands
(This one fits in my car)

[Joe Cygan]

This is very exciting stuff! Thanks for sharing.

Joe

[IanH1960]

Cool!

Neat, compact design - can't wait to see the full size version.

Looking the motion it looks a wee bit "stickier" than your other full size platforms so far - have you had a chance to analyse this out yet? I guess it's in the servo loop tuning?

Ian

[autocadplease]

I really appreciate you sharing this!

[flymo001]

Roland....what can I say......EXCELLENT!!!!! :cool::cool: can't wait to see the 1:1. :lol: Saw the other base in action last year....that was fantastic, so will say hello this yr. See you there!:-D

Cheers, John P

[Roland]

Ian,
The "stickier" motion is caused by the insufficient gear ratio and the rather coarse cogging characteristic of the motors. I build this unit mostly from scraps out of my junkbox, so could not select optimal components. The real-size unit will have bigger gear ratio, and uses low-cogging motors, so should be fine.
I will also use belt-pulley for the motor-1st gear transfer, to get rid of the chain induced noise.

John,
See you this weekend at Lelystad!

[wannabeaflyer]

Hi Roland just to say as a long time follower of your work , you never cease to amaze me with the stuff you do , always very much with the homebuilder in mind and it all comes with great and easy to understand assembly and methodology , so just my way of say keep up the great work those of us building or hoping to build motion platforms could not find a better Combination in your ideas and Ians Software we definatly need guys like you ..

[kermit]

Hi Roland, If have looked on the build for fun site,
I`m also interested purchasing the 40SPU-1 drive system card with the speedcontrollers,
Do you have experience with lower voltages on the speedcontrollers MDo3,
I can buy actuactors with 2250 newton(about 225 kgf) 38mm/sec at 12 volts 20 amp, but the controllers operate at 24/36 volts,
But I`m not a electrician ,I can put a battery in a radio and thats it:-D
Will it affect the speed or working capacity
greetz
ps I will attend the fair also

[IanH1960]

Hi Kermit,

I'm sure Roland uses drive hardware of his own design and not my 40SPU-1 cards etc, however I can answer your question on the MD03's. The manufacturer originally advertised them as 50V controllers but recently "clarified" their ratings at 24V. I have used them at 36V for some time without problems. However you can power them with voltages as low as 5V if you want - the voltage output range is then simply limited to that maximum.

They will operate with a 12V motor side battery supply.

A quick note on your actuators - 38mm/s is on the low side for a motion platform drive. Some I've built were at 400mm/s. If you only want slow speed movement then you might be able to use slow actuators - but this then rules out things like touchdown bumps or other vibration type effects or faster impulse type movements.

Ian

[kermit]

Hi Ian,
thanks for your reply,
I have already contact you thru mail regarding the actuators ,
I had just another question regarding control volts and main motor power
greetz

[Wendy]

Hi Roland,

This is great. You are definitely the expert on moving platforms.


Regards Wendy

[Roland]

Regarding controllers, supply voltage and matching motors:

There are a number of parameters that are important:

1. Motor DC resistance (in stalled condition)
2. Motor inductance value
3. H-bridge (controller) PWM frequency
4. H-bridge supply voltage

Motors with low DC resistance (normally low voltage 12V types) will draw considerable current during platform sudden motion steps. When applying more supply voltage than rated, the current will increase. (i.e. a 0.5 ohm motor will draw 12V/0.5=24Amp max at 12V supply with H-bridge fully on. At 24V the current becomes 24V/0.5 = 48A! the controller H-bridge MOSFETS may not be able to handle that kind of current.

The motor inductance value together with controller PWM frequency determine the current ripple through motor and H-bridge, even during steady state. Low inductance means more current ripple, and low frequency also means more current ripple. Too high current ripple heats up the motor and bridge during operation.

I normally use 50V rated DC motors, DC resistance about 1 ~ 1.5 Ohm. Inductance about 1mH. These run fine from 24V - 36V battery supply, (peak currents about 36A). I set my PWM frequency to 20kHz, so you won't hear any noise. Ripple in this case is rather low, about 1A pp.
Note that power loss in these systems is mostly coming from conduction losses, so selecting higher voltage/lower current application is better than selecting lower voltage/higher current (even though power is the same: P=V*I)

Note on controllers:
They sometimes specify 2 voltages: one voltage for the small signal section (i.e. 5V or 12V) and one voltage that is the supply for the H-bridge. It is this voltage that gets chopped and drives the motors. This H-bridge voltage can indeed be varied from smaller to larger values, to get smaller or larger max motor current.
But if the controller has only one supply (small signal supply is derived internally from the main supply), there may be a minimum and/or maximum limit to the supply.

Check the controller spec carefully! Motor specs are sometimes hard to get, so measuring is a better way to know if your type is suitable.

[kermit]

Hi Roland,
Thats what I mean, I`m a complete noob :lol: about currents , resistance and Inductance.
I have more understanding in hydraulic or airdriven machines
I can connect a motor to a controlcabinet ,
If I have a drawing but dont ask why you need this type of motor for that kind of machine
offcourse I know by experience that a certain type of machine needs a least a certain type of electric motor but the story behind that is a questionmark
I can however get all the electric hardware I need, but havent got a clue about electric parts.
I can mechanical build you an actuator if I have all the data ,
For that reason I`m looking for someone who can tell me what I need electric wise.

The bottomline is I want to use the 40SPU-1 card with the MD03 controllers, It looks great and seems easy to install.
I building a platform size 1600x1200mm from 50x50mm steeltube gross weight incl Cessna cockpitsection and person aproxx 400 kilo.
The platform will be connected with a cross coupling with the central gravity zero out(balanced) on coupling aproxx 60/70 cm above ground.

As I understand the MD03 uses 24/36 volts to feed the motors for moving the platform.
My question is :

Can I use this current as a( in Dutch stuurspanning ) control current and use 220/400 volt for feeding the motor itself with out blowing the MD03 out .
I can get motors for building dicated collumn actuators up to 2 KW 220/400 V 12/24 control current 50 HZ.
However will it work without a meltdown?
This is my question in a nuttshell, my wife says Ì`m nutts:grin:
but enough said bla bla bla
greetz Henk

[Roland]

Hi Henk,

The main reason that there are not so many homebuild platforms out there is because building one requires a number of skills that are are not often found in one and the same person: Mechanical, Electrical and Software. (and an understanding partner)

As far as I know, there are no ready for assembly kits avaliable that can be put together without above mentioned knowledge. Ian's plans go a long way, but still require some skills.

Difficulty increases when building bigger and heavier platforms. That's why I have kept the size and payload of my platforms small, as I would run in some serious issues when trying to build bigger.

Like you say, you are good in hydrolics, but not in electronics. The only thing I can advice you is to learn the skils you don't have yet. If you are trying to build a big size rig without electrical knowledge you're going to run into some serious issues.

I have tried to put most info on my site as clear as possible, but it is defintely not a plug and play guidline, so needs lots of experimenting.

I hope I did not put you off, but I'd say start simple, and after you gain experience, go for bigger projects.

[kermit]

Hi Roland ,
thank you for your reply,

Thats why I place my questions on the board, maybe someone can shine a light on the issues I`m struggling with.
like:
Can I use this current as a( in Dutch stuurspanning ) control current and use 220/400 volt for feeding the motor itself with out blowing the MD03 out .
I can get motors for building dicated collumn actuators up to 2 KW 220/400 V 12/24 control current 50 HZ.
However will it work without a meltdown?

If that is the case I know what to purchase or if not what then.
Experiment costing a lot of time and money and I don`t want to invent the wheel again.

I see a lot of platforms all over the internet , but when I ask to explain in plain English or German/Dutch for that matter.
I get so much technical info , after a few lines I`m lost,
You have to have an academic degree in electronics just to understand what they talking about.

I`m taking a lot of effort to make a platform, but electronics is not my cup of tea.
I will talk to you on the fair, tomorrow.
greetz
Henk

[wannabeaflyer]

Hi Roland can i ask that age old question again regarding backdriving, i have looked at your saruss design but was not quite sure if this was a factor as i could not see a worm and pinon setup .. i appreciate the design will lift the calculated load but wondered about the non powered state or park position of the platform when no feedback signal is applied .. dont mean to be awkward its just that over the years of my dreaming waiting and still tinkering with platforms ( None finished yet LOL:-) backdriving of the lift actuator always seemed to be a factor, in the ideas i have looked into or seen elsewhere .. does the feedback circuit always hunt for positon and in so doing keep the motor energised and hold the platform in a steady state ?? love the stuff you do and as always no critiscism intended just trying to learn a bit more and get another perspective on how things work ..:-) oh by the way did i say i love your ideas :-) Ps Nice explanation in laymans terms about motors now thats the stuff i like .... it eventually seeps through the dark matter thats my brain and i get that eureka moment ..Cheers Roland

[Roland]

Just returned from FSweekend show, which was really great! Very nice to talk to all fellow builders and many other interested folks. Saturday evening we had a wonderful dinner and lively discussions. Thanks Matt!

The prototype has proved the concept, as it is still in one piece after 2 days of continuous operation in heavy turbulence settings and some shakey X-plane sessions.
Thanks everyone for the great comments and some really good suggestions for the bigger version.

To come back to the back-drive question:
I have always used straigt spur-gears for my drive train, so there is no worm-gear involved. This means that the motor will back driven when the payload is not properly weight-balanced. The bungee cords will need to produce sufficient up-force to cancel the platform weight down-pushing force back to the motor.

In my servo system, the position feedback signal will be continuously compared to the input signal, and the motor will be driven in such a way to make the position feedback signal equal to the input signal. If you would try to push the platform down in this equilibrium situation, the motor drive would immediately react and apply power to the motor to keep the current position. So weight unbalance will create considerable steady state power dissipation to the motor, heating it up and draining the battery.

Some worm gears cannot be back-driven, in that case you could do without weight-balance. But it means that the motor will have to work much harder to push the platform upwards than when going downwards. Most likely, there will be an unequal up/down speed, and also considerable more wear in the worm-gear without some weight-balance scheme.

The reason I'm not using worm-gear is because I'm not familiar in specifying this type of gear and construction. (I'm an electrical guy). I also have the feeling that backlash is much more difficult to control in worm-gears than in my straight spur gear drive train.

BUT worm gears seem to be the choice of many commercial platform makers, so there must be an advantage to it. I welcome anyone to present some info on making succesful actuators with worm-gear that don't need weight-balance and still can achieve fast response, equal driving speeds and minimal backlash.

[kermit]

Hi Roland,
Nice talking to you on the fair,
You put me on the right track as I was going round and round,
Now I know, what I have to ask the electric guys at my work to fill in the gaps and details.
If you want to know more about actuators and wormgears take a look on this site.
http://www.elsto.nl/index.html
This is a part of the Holding I`m working in.
greetz

[Roland]

Thanks for that link Henk,

And indeed it was nice talking to you. Looks like the colleagues in your company should be able to give you some expert advice on transmission system drive. If they are interested they may even join the team!

The PDF files in the link you gave do show some servo type worm-drive gears. Seems that backlash is solved by preload pinion shaft, that can adjust backlash to zero. But it sounds like this is a solution a bit outside the scope of the DIY-er.

Still not sure how these worm-gear driven units would operate under a-symmetrical load. (like an unbalanced motion platform)

[kermit]

Hi Roland ,
In the winches and capstansindustry I`m working in ,we`re using frequentie-inverters and sensors, But as you say a bit outside the scope of the DIY-er.
I`m only making attachments for the sensors and drill countingholes for the sensors.
We have a specialist who`s main job is to install and ajust the parameters within certain bounderies.
But daar I have no kaas van gegeten.:D
Also the backlash is partial depending on the inertia of the wormwheelgear itself,
At a certain ratio you can have some kind of backwinding up to a absolute stop for tonnes of weight.
greetz Henk

[IanH1960]

Still not sure how these worm-gear driven units would operate under a-symmetrical load. (like an unbalanced motion platform)

One way to think about worm gears is as other gearing but with lower efficiency - and even lower efficiency when back driven. Indeed if the conditions are right (small worm helix angle, high friction conditions etc) the back driving efficiency can drop to zero - ie the box locks. However locking is not a universal characteristic of worm gearing - many can be back driven.

Where this is the case the torque required at the motor end to "balance" the driving load can be a lot less than it would be in higher efficiency sets such as spur or helical gears. So you have a circumstance in which load deceleration and load holding is made easier for the motor. However the price to pay is that the internal sliding friction effects that help here also hinder when it comes to driving the load - the input driving torque needs to be higher than it might be because of the inherently poor worm gearing efficiency.

If the loading is unbalanced then worm gearing assists decelerating and holding it, but also makes it harder to accelerate and drive the load (though not by the same ratio). The question arises is one worth the other.

I think one reason for specifying worm gearing is because it provides a convenient and possibly cheaper way of achieving a large reduction ratio than multistage spur reduction or epicyclic gearing. If I had the money my high reduction box of choice for a platform drive would be epicyclic- compact with good efficiency - that and a well balanced platform to keep the actuator loads to a minimum in any case!

..probably just confusing things now - it's late!

Ian

[kermit]

Hi,

A wormwheel gear has always the same ratio gearwise And called I= with a number .
We have a program wich calculate the highest for seen loads and give a advised ratio .
When we sell spudpolewinches that have to pull out poles on waterway`s or grounds we use frequenty inverters that counts the speed of the mass being pulled,and will ajust the motor speed and therefore the speed of the wormwheelgear.
greetz Henk

[wannabeaflyer]

Hi Guys thanks for the info and links about wormgears .. i did not mean to detract attention away from the great work Roland has done, it was just over the years the balancing act for DIY motion simmulator Builders seems to me, to be the comprimise required regarding Weight , Inertia,and The real bugbear cheap easyish to make actuators, using common non expensive but accessable parts ( Utopia diy sim actuator wise :-)) i suppose what little i have managed to pick up is, their is not at present a simple fix to this and its guys like Roland, Ian and Jim over at Cheap motion cockpits NZ that have really lead the way in giving hope for those of us out there that still have that Dream making ( finishing ) our own motion platforms . so Keep on doing what your doing Roland and the rest of the innovators, i for one always eagerly await to see what you guys develop cos it is always done with the DIY Guy and Common basic Skills in mind and that in itself is great news Cheers :-)