Drive by wire electric Go Kart

[Vrod2003]

New here as of a few hours ago. I have done some searches but have not came up with what I am looking for.

I getting ready to put together a frame to hold a car racing seat (cheap light $125 seat) for a DIY racing simulator using a ThrustMaster feedback wheel & pedals. I planning on putting some wheels on it so I can move it around and out of the way when not gaming. The intent is to park it in from of my 140" projected screen for some Forza 5 action rather than bolted to a small desk. So something just popped into my head say how cool it would be to be able to dual purpose this by dropping in some battery(s)/motor/big wheels and then use the existing controls to simply unplug the thing from the wall and into the on board battery (assuming a car battery w/ inverter) along with a laptop... Hopefully you more experienced folks know where I am going with this.

The obvious hurdles would be get the wheel and pedal signals passed to the computer and then back out to a control panel to handle the steering conversion and power to the motor.

So would my efforts be futile? Has anyone even attempted this before?

I can find the motor, axle, wheels. Need to find some sort of servo I suppose to control the front steering and a control box to feed some info from the pc into it. Then the other obstacles would be how to get the wheel/pedals to talk to the whole thing.

Advise, comments, or flames welcome.

 

[landuse]

Welcome to the forum. It sure sounds as if you have a unique project that you want to attempt. I am not an electric guy, so whatever I say would be useless, but someone will give you some answers soon

So let me get this straight....You want to be able to use the steering and the peddals for the game, as well as proper go kart riding when you are done? The steering would probably be the hardest to sort out, because for a computer game there is no actual moving steering parts like tierods etc. that would physically move wheels. You are going to have to build that all in. Same with the pedals and brakes. You are going to have to run cables etc to get them to actually work in the real world

Hopefully I am on the right track with what you are asking....

 

[Vrod2003]

Yes you have nailed it. Trust me, its way over my head and am seeking some expert advise for sure. I am thinking though I will likely have a manual brake on the system but all the rest as you pointed it will be drive by wire essentially. I have no idea if its been done before on a small scale like this. Though I highly doubt I would be the actual "first" to do it, so I am sure there are folks out there with some helpful advise. I just need to find them to help educate me.

 

[itsid]

Welcome!

Pedals should be fine those can easily be altered..
but the steering wheel will be a problem, not actually the wheel controls,
most computer hardware is durable and precise enough for that task, but the actual steering mechanism will be quite a task.

You'll need a servo that can handle the forces (do not underestimate them)
and a feedback loop,
I'd say a redundant system with two control loops and two feedback loops.
so in case one set fails, you still have one to control the kart.
At best if one fails it triggers a killswitch for the motor, so you don't actually drive anywhere with just one working but just can safely stop the kart.

I think you're best bet is an industrial servo motor (no RC stuff please!)
a linear actuator would be nice, and to complete the redundancy on the steering, a second one (both up for the task) working on the opposite ends of a pitman arm.

Oh with the feedback loop you could actually use a force feedback steering wheel, that'd be a much nicer experience I think.

That's a ton of programming though, quite a handfull of controllers to think of and synchronise...
Can you do that?

'sid

 

[Vrod2003]

Well, I can solder and weld and write some code. I'm a IT Tech guy by nature for 25 years now but that by no means makes me any sort of electrician. What I can't do is have an unlimited budget and I hope to scavenge as much as possible for the project. If I can plan to the point where the obvious obstacles can be overcome then I hope to actually attempt it. This thing does not have to scream, 15-20mph would be the max I would like to get out of it a 3-5 mile range. As far usage, I would say less than 20 miles in a year. More of a novelty project with the idea of it being a conversational piece when completed.

Currently I have the wheel w/ pedals, spare laptop, seat and I am going to start the frame to mount it all up. I need to think it a bit through so that it can be built and be compatible or retrofitted if I decide to actually make it into something driveable.

For reference here is the wheel I already have:
http://www.thrustmaster.com/en_US/products/tx-racing-wheel-ferrari-458-italia-edition

 

[itsid]

Well, I can solder and weld and write some code. I'm a IT Tech guy by nature for 25 years now but that by no means makes me any sort of electrician.

perfect, no electrician needed, just an electronics guy
knowing about controllers, sensors, actuators and such and how to wire them in the end is enough.

If I can plan to the point where the obvious obstacles can be overcome then I hope to actually attempt it.

obvious would be the steering mechanism, as I said earlier..
that's not too easy to achieve I think, and it's critical;
if that fails you'll be hurt.
Back to that in a minute...

This thing does not have to scream, 15-20mph would be the max I would like to get out of it a 3-5 mile range.

That sounds easy.. maybe 1000W will do already (my1020 motor for example ~100$)
a simple matching controller is in the 30-50$ range
So that part of your build isn't expensive nor complicated.

Currently I have the wheel w/ pedals, spare laptop, seat and I am going to start the frame to mount it all up. I need to think it a bit through so that it can be built and be compatible or retrofitted if I decide to actually make it into something driveable.

For reference here is the wheel I already have:
http://www.thrustmaster.com/en_US/products/tx-racing-wheel-ferrari-458-italia-edition

Well, the thrustmaster is already overcomplicated with all those buttons that end up without being used in the end,
so maybe you want to find an older much simpler FFB steering wheel.
Keep in mind that it will be ruined for PC/PS or Xbox use in the end! so buy cheap.

And back to the obvious obstacles.
DO NOT use a laptop other than for programming and uploading the firmware to the controllers.
Again, the more parts you have the bigger the lag from sensor to actuator.
And 50ms lag in a steering setup can be really annoying already.

So no laptop!!

Tap directly into the wheels internal controller, attach a small microprocessor (an arduino should do)
you'd need servo shields (or external servo controllers)
and of course good code.

Keep it short.. use the fastest tiniest code possible,
you want to reduce lagging as much as possible.
steering: no more than 20 lines of code I'd say.
( IIRC a segway uses 14 lines to balance.. that's why it's quick enough to do it properly )

The pedals.. the more I think of it the more I'm sure you should not even use them.
three pedals and you only use ONE?
and there are pedals that work perfectly with the motor controller, small single throttle pedal.. your good to go.
PLUS: that eliminates additional controller code on the arduino

First
get to know each and every cable inside the steering wheel,
what it does, what signal it uses etc.
start from there.
get an arduino (they're cheap) and try to connect it.
read all the variables the steering wheel tells you,
maybe even skip the internal controller and tap directly into the sensors, to see if that's easier to work with.

you can already try to figure out if and how you want to control the force feedback.

Now, as soon as you know HOW the steering will be working electronically,
get some RC servos for each signal
and make some bench tests, just add an indexer to the servo horn and mount it to a board with a dial and set reset and retest the code over and over,
if you're happy with the result remember to incorporate redundancy on the steering.
stay safe

And now the hardest part:
find a good servo for the task,
I'd say a linear actuator with about 2" travel is what you want,
keep in mind that it should be capable of at least 200 pds and is precise on the amount of travel;
the higher the precision the better the steering.
Oh and it needs to be fast too, set speed is critical.
And that's where its going to be expensive.
My best guess is, the steering servos are somewhere in the 150-200$ range each at least.
maybe used ones can be had cheaper.

You can of course fab something with a std industrial servo (360° rotation motor) but I don't know if that's much cheaper in the end, but I'm almost sure it'll be more complicated.

'sid

PS really cool idea.. I like to see that in the end

 

[Vrod2003]

All great input.

Hear was my thoughts, though they differ from your suggestions.

1. Keeping the existing wheel *is* a requirement as that is the primary use of the rig so if it can't be done with it then powerizing the thing is a no go. If I have to use completely different wheel/pedals and etc. it would would defeat my goal. The wheel input/output is a single USB cable so it really needs to go to some device that can read the data coming from it.

More overview:
The hardware I have is currently hooked up to an XBOX One. The device is also compatible to the PC. This would mean that there should be software libraries available to read the inputs and outputs of the wheel and pedals. That gives me control over the tension on the wheel etc. The pedal setup is only a 2 pedal system, throttle and brake. My thoughts would be to have the PC read the wheel and then via serial or some other output send the signal to an arduino device or directly to the controller if possible. Am would believe that there is also software libraries available to talk from the PC to the arduino device already available as well.

I have some spare shocks off of my motorcycle though they may be too stiff. I also have a swingarm, axle, belt pully, and shaft pully from the same motorcycle as well should I decide to go with a single rear tire in a trike style. I do have the rear rim and tire as well but its on the large and heavy side I would believe for this project. I have some A-Arms off of an old quad I would not mind using for the front end. Multiple head lights (not required). Bits and pieces here and there, lol.

Awesome news on the motor info and controller info, very much appreciated. To be honest all info so far is good info. I would rather know ahead of time if the project is futile before dumping in too much time and money. Your knowledge is appreciated.

If the the wheels steering column was more sturdy, I had even given thought to placing a gear around it and manually controlling the steering like an old rat rod. However it is not up to the task for that much stress and would surely fail.

Maybe I should also state that "Electric Go Kart" for this project is just simply because when doing some research for what I was dreaming about, I found myself at this forum as the most probable for getting some knowledge base.

 

[itsid]

1. Keeping the existing wheel *is* a requirement as that is the primary use of the rig so if it can't be done with it then powerizing the thing is a no go. If I have to use completely different wheel/pedals and etc. it would would defeat my goal. The wheel input/output is a single USB cable so it really needs to go to some device that can read the data coming from it.

Okay, keep the wheel no big deal;
but don't use the USB cable.. just don't, you'll need to decode the whole set of data and you only need 1/8th of the information, thus every step starts with filtering and you loose time!
That is not a big deal on a computer, all game action lags by the same amount, so you won't have any problems, but real life is not a game you cannot pause it to wait for your data to be computed.

It's mandatory to read directly from the controller OR sensor to avoid any other input messing your steering up.

Of course you can use the pedals of the steering wheel, but it doesn't make sense..
they're detachable anyways (just a small cable to unplug) and getting a small pedal throttle is cheaper than building the controls for your existing setup.

If you plan is getting a usb driven kart..
(not actually what I would recommend or would have thought reading your topic.)
then yes well
a 5h177y idea I must say, but get a microcontroller with USB port and read the signals raw and try to process them,
that'll increase code by a huge amount, and you might fail due to the lag you created with the additional code.
And no, you Win PC with it's CPU is not faster than 15-20 lines on a microprozessor

More overview:
The hardware I have is currently hooked up to an XBOX One. The device is also compatible to the PC. This would mean that there should be software libraries available to read the inputs and outputs of the wheel and pedals. That gives me control over the tension on the wheel etc. The pedal setup is only a 2 pedal system, throttle and brake. My thoughts would be to have the PC read the wheel and then via serial or some other output send the signal to an arduino device or directly to the controller if possible. Am would believe that there is also software libraries available to talk from the PC to the arduino device already available as well.

Really???
Add another 5-6THOUSAND lines of code to read a signal?
You don't want me to comment that, do you

I have some spare shocks off of my motorcycle though they may be too stiff. I also have a swingarm, axle, belt pully, and shaft pully from the same motorcycle as well should I decide to go with a single rear tire in a trike style. I do have the rear rim and tire as well but its on the large and heavy side I would believe for this project. I have some A-Arms off of an old quad I would not mind using for the front end. Multiple head lights (not required). Bits and pieces here and there, lol.

No.. no... noooo.. and no!
You want to build a buggy, no kart then?
That's somewhat different (in weight mostly)
thus you'll need about twice the amount of power to get it moving,
with a motorcycle wheel on the back (assuming something like 24" tire diameter) even more than that.

well..
I think you should rethink the whole project.
Adding "Windows" to your electronics is -I'm sorry- stupid;
you need realtime processing, and windows isn't even a real time operating system, that just doesn't match up.
DOS might work Windows will not.
And even that I wouldn't want to try to be honest.

'sid

 

[Vrod2003]

Hmm, all good points. If the latency is a show stopper on my approach then destroying a $500 wheel is indeed a show stopper unfortunately.

I wonder if I could do some testing on the route I mentioned small scale like with a simple RC front end to check out the latency though.... Hell, the effort and learning might even lead to some other cool project stuff I could start on or build from. Coding is not a big deal for me in any modern language and is sort of like a good refresher on skills for me, so no loss there time-wise either.

The use of the motorcycle wheel and rim would not have likely have been the case as I mentioned due to size and weight. More on the lines of the rear end parts with a small diameter wheel spaced in swingarm was what I was thinking. I really was just mentioning it in the fact I have essentially and entire Harley V-Rod rear end (swingarm pivot back) including calipers/rotors and master cylinder w/ hand brake from a conversion project I had done in the past. IE, parts collecting dust in my garage and readily available for reuse.
http://www.1130cc.com/gallery/files/4/2/img_1811_original.jpg
http://www.1130cc.com/gallery/files/4/2/img_1812_original.jpg

Your comment on a buggy might be at the 50% in between mark. So somewhere between a buggy and a go kart dimension wise.

 

[itsid]

Hmm, all good points. If the latency is a show stopper on my approach then destroying a $500 wheel is indeed a show stopper unfortunately.

I can imagine
That's why I suggested getting a cheap old less fancy ffb wheel
Basically you do not need the actual wheel anyways, just it's base (that's where the magic happens )
And older (I think it was) thrustmaster wheels allow attachments of 'real' racing steering wheels as long as they're momo steering wheels (because that'd be the hub that fits)

Anyways, you want this exact wheel for an unknown reason,
and that's where you will have to make a decision.
A) find a cheap 'broken' one on ebay (chewed cables are nice, broken pedals maybe...)
B) find another cheap ffb wheel
C) make your peace with rendering that one useless for your xbox.

I wonder if I could do some testing on the route I mentioned small scale like with a simple RC front end to check out the latency though.... Hell, the effort and learning might even lead to some other cool project stuff I could start on or build from.

Of course you can get a small rc and try it with some software on a laptop,
and I believe it'll work reasonably well..
BUT that's far from good enough for a real time application.

Keep in mind, you'll have to expect much much greater forces on the servo, and even fast industrial servos are due to their higher power, most of the times much slower than fast RC servos. (but have a better resolution / precision)

But sure, go ahead, try it if you have all the parts.

Coding is not a big deal for me in any modern language and is sort of like a good refresher on skills for me, so no loss there time-wise either.

great, but again I recommend microcontrollers, because of size, power draw and speed.
So modern languages are more or less useless.. it's Basic or Assembler most of the times

The use of the motorcycle wheel and rim would not have likely have been the case as I mentioned due to size and weight. More on the lines of the rear end parts with a small diameter wheel spaced in swingarm was what I was thinking.

I wouldn't recommend making a the chassis at all at this point,
mainly because it's easy to mess up the steering geometry,
and since you want to get fancy on that, that needs to be rock solid mechanically at least,
so that you KNOW it would work and all errors are code based.

if you feel comfortable with Camber, Caster and Ackermann
go ahead build it yourself.
But:
Build a manual steering, and test the steering that way find a hill roll the kart down and see how it works,
you'll then be able to tweak it if necessary.

The thing is: if you make a frame it'll be most likely much heavier than a bought racing kart frame (those weigh next to nothing nowadays)
and sometimes you can find a naked frame for less than 50$ on CL.. Some even got them for free around here.
(so that's cheap, and reliable!)

You want to make a lightweight kart, that way you can stick to the 1kW motor I mentioned.
if you add suspension and build the frame yourself, you increase the inertia, and you need a bigger motor to overcome it
(then we soon talk about golf cart or fork lift or industrial cart motors with about 4kW)
And it starts getting expensive again.

But that's your choice and your's alone.

The first few things I would get IIWM is an arduino to play around with and learn about it (really they're cheap and a good replacement for the laptop)
Again, there are usb ports available on such tiny devices,
I'm not sure that you can access the usb of the arduino from the code (never tried that) or if it's just for updating the codes.
But hackaday lists tons of micocontroller boards incl a review

Plus a small servo controller .. maybe the micro maestro (pololu)
it's cheap and works great as far as I have been told.
That way you get to know the components you need in the end without spending a fortune on trying
(we talk about ~50$ for now)

Again, you should sit down and make a list of the things you really want, you want,
the things that you cannot do no matter what.
And then do some research on the net..
what parts you might need and what it'll cost you to get them.

Then multiply that amount by 2.4 and you know what it'll cost to build it.
yes.. you will forget something in the list and that's going to add up really fast and if not, you will feel very confident because you saved all that money

Good luck!
And keep us posted.

'sid