Orange button (glaring white in the pic) is the ON/OFF, we probably only need think about that one and the joystick really for the target project, the other controls are for umanned remote control purposes auxiliary really, but i can run you thru them:
Orientation - imagine facing the joypod as if you were sat in the driving seat, so the scuffed side is LHS (West) and the two screws you see are bottom (South)
blue button (left) for head tail lights,
a push button horn (just out of shot RHS),
a toggle switch for indicator lights(top/RHS),
the battery meter strip (top)…
Oh and the twist knob dial thingy to the left, adjusts the speed range limits (MPH).
↑ Here is the joypod’s coupling socket, which plugs into the control brain thingy.
Oh - this is geeting too complex in terms of info, this. I do not have time but for assisting in more than the overall lines, sorry.
I am not sure; Do you say that you have acces (one way or the other) to some switches that we can hack into, so that we (at low power, not the cables to the motors) can control the chair by simple on/off of these?
If so, what you can do is:
Get a Picaxe starter-pack, and one extra 28 board+chip (look at “start here” in top menu)
Get a Bluetooth - set (I’d go for “Easy Radio”- look in components)
Insert relays where the switches are hacked
Get another joystick (can still be found in toy-stores, now they just have a complete TV-game in them) - take out the TV-game, and insert one Picaxe 28-board and an emitter (Bluetooth or EasyRadio).
Program both chips, you are go
I would strongly recommend starting by making a robot, such as the “Start here” in top-menu. And then advancing with wireless, before looking into trying to hack the chair.
These things works best if a person contrates on one task / layer at a time, makes it work, gets to know all about it… and then add another layer…
The omni directional joystick is proportional rather than simple on/off, the photo of the jack plug i highlighted has a removable sheath that can expose the back of the connector so it’s individual pins can be probed for voltage values and functionality.
Hmm… Well then - with or without relays, You will have to hack in yourself. I would need to get it in my hands to be able to say anything more clever
But I hope you make it. And also - if you do,why not just add 2 sensors on the chair? When remote-controlled at a distance, would it not be nice if it automatically steered free of obstacles?
I wish you luck - I have a similiar project, except the idea with mine is to make it completely autonomous. Here is a link to my “scratchpad” of work.
The least invasive plan would be to “strap on” a radio control rig which can push the joinstick as a human rider would. You could do this with 3 servos:
servo 1 - push button on or off servo 2 - push joystick left or right servo 3 - push joystick forwards or backwards.
Then you could drive the thing around with a toy airplane controller.
I would recommend this, because interfacing with the controller circuitry can be complex - at least with my project it definately was. My project consists of having an on-board computer drive the chair autonomously. I did not like the idea of the computer controlling the chair through servos. The computer is digital, the controller circuitry is digital, so I thought “Why should the computer control the chair through an analog servo mechanism”.
Long story short, I tried interfacing the computer directly to the control circuitry. It was frustrating but educational. I never realised at least with this model of chair - a large amount of the ciruitry was dedicated to SAFETY! Everything attempt I made would lock up the brakes and freeze the chair. Instead of interfacing the micro-controller controlling the chair, I decided at that point to try to interface with the MOSFETs controlling the power. I had bad luck there to, the safety circuitry was inteferring with driving the MOSFETs. Eventually, I burned one MOSFET out during experimentation - now I have to build all the driver ciruitry back up. Again this is not a trivial task. To sum up here are a list of possible solutions:
1. Build a “strap on” radio control harness which can push a button and move the joystick in an X-Y direction 2. Integrate with whatever the joystick is connected to - this “might” be something equivalent to a potentiometer - if so it might be easy to take one of the servos and attach it directly to the pot. 3. Integrate directly with the control wires - this might be complex if your system has a variety of safety features and no accessable diagrams of specifications. 4. Integrate with the motors directly - at some point its either a relay or a mosfet which is controlling the motors. My bet is that it will be mosfets that are being pulse width modulated for variable speed control.
Those are the 4 options which I could think of, I would still recommend #1.
Good Luck!
Hope my rambleings have been helpful in some form.
I’m familiar with this chair. Oooohhhh - You have the walnut effect joystick! Smart!
The joystick is a dual-axis potentiometer and both tags and wiper of each come out on the screw connector. (And a couple of switches for power, etc.)
It MIGHT be possible to splice in a couple of electronic variable resistors in place of the stick. You know those serially addressable 256-position pots you get? I dunno how much current goes through the stick. Prolly not that much.
The prosthetics dept at your local hospital should be able to get a technical spec for the controller. If I’m entirely truthful, I’d consider this to be a “safety-critical” application and,as such, it frightens me a little.
Where are you, geographically? If you’re in the UK, try a web search for “Remap.”
24V wheelchair motors Do you know how to drive / brake / get feedback from one of the motors from the chair? I have the motor in my garage and it has about 15 wires hanging off it and no clues. It would sure be better than these windscreen wiper motors.
Thanks guys for your input so far, am in London UK, and today had a chat with the chap that runs http://www.roamingrobots.co.uk which is the knock on sport based on the tv series Robot Wars. Figured as it was r/c rather than autonomous that I need it would be worth tracking down people who love using bits of old electric wheelchairs in their ‘battle bot’ r/c’s…
Looks like he may be able to help me do the work as they offer robot building services, he says 2.4ghz bluetooth type technology will be far safer which I figured. Apparently battle bots are all moving away from, 27mhz / 40hz bands. You’re right BaseOverApex safety is important, if it went runaway it could snap someone’s leg like a twig or dent a car big time.
Cheers GroG, your project is very interesting, deffo want a big STOP button - and yeah I need to invest in a 2nd joypod rig to fit servos on and strap on the back as you say, would like to have both the r/c pod and the stock pod hooked up to the contol brain either with a splitter or a switchable splitter.
The strap on pack doesnt need to be exposed to the elements nor have to be used directly by hand so it can be boxed away once complete.
The walnut shifter will be going - too top heavy for unmanned roaming, it will send the F50 off course.
Yes at least from my motors - I’ll be driving them with mosfets and a PWM circuit. Brakes, I dont need no stinking brakes … when the top of my H-bridge is driven high - it will brake the motors, I have run into problems with this before - when running at top speed (i found out that top speed is pretty wicked fast - something the original ciruitry would never allow) and then braking generates a huge voltage spike - enough to fry mosfets even with protection diodes. For feedback I’ll be rolling my own encoders - which i’m starting to work on now.
I only had 5 wires to worry about - 3 were an electric brake assemly. Once I removed that I was down to 2 - I managed to figure out the 2
15 wow? I assume you probably have some other components in there like a tach or an encoder or something nice. I would start by turning the shaft and the 2 wires where you get the biggest voltage output are probably the powerlines.
Hey GroG, curioius - how would yours stop in the event of a power failure, or dead battery on a slope with no ‘stinking’ brakes? Used to be a mobility scooter out there in the market that would just run out of juice and thenn free wheel the old deer into oblivion
Oh and when you talk of yours allowing wicked fast speeds this interests me for manned use :OP
at some point i’m going to put the electric brakes back in - but for now they were interferring with my progress of hacking the chair…
Yeah i was supprised - turns out i got an extra-heavy duty chair, and the controller circuitry kept the top end of the speed fairly low. Its packing 2 deep cycle marine batteries. I keep posting info as i work.
Definitely give Remap a shout. http://www.remap.org.uk/ We make free stuff for folk with disabilities and some of them are pretty technical… We have engineers in Ealing, Hammersmith & Hounslow, Brent & Harrow, Barnet, Croydon & Romford. That’s just the London area.
Hmm. So on this page you say that at work you work with jet engines, and in this post you say that you build devices for disabled people. I’m trying to picture a gadget that involves both jet engines and the disabled, and it sounds awesome
Yes. During the day, I build and program test rigs which pretend to be jet engines. That way, the folk who write teh software to control the jet engines don’t have to have a big, smelly, noisy engine intheir office. This in a necesary evil in order to subsidise robot-building.
In spare time, I help build stuff out of crap I have lying around to help disabled people do stuff more easily. It gives me an outlet for my strange super powers and abilities. I think I would probably, otherwise be an evil genius and end up in prison.
I assume that you want this “unmanned power-chair” to move, upon remote request, WITHOUT ANYONE ABOARD, from a secure place, to a place that the disabled person could get on board, and then he/she would take direct control of the power-chair.
I also assume that you would want the disabled person to be able to get off of the power-chair, and send it back to the secure place.
In one case you’d want the power-chair to move from point “A” to point “B” upon remote request.
And then. in the other case, you’d want it to move from point “B” back to point “A”.
Are my assumptions correct?
I do not know how to solve your problem. I do NOT have the talent.
But I bet that there ARE people on this forum that DO have the talent.
Let’s get this thread back on topic, O.K. guys?
Someone asked for help, and then the thread went off on tangents. I’m not trying to be pissy, but as I see it, the thread has been hi-jacked, and the original question ( which I think is very valid ) has not been addressed.
GroG I have been wanting to intervene a power chair at the joystick signal with a computer just as you talked about above. I think you said you were successful after figuring out all the safety crap that accomidated with the chair. Do you have an pictures or a link?
I am very interested-I’m into the arduino microprocessor board and I know a little programming, I also work at a power chair rehab company, lots of parts. I have been mounting servos on my joysticks for the last 3yrs and wants to try something new. I would like to get your email address if possible.
Its been a long road of experimentation, trial, and a helluva a lot of error. But, my latest revision is based on the now famous BOA Bridge (catchy aint it BOA?).
Some quick highlights of the history :
1. Ended up ripping all of the “brains” and safety controls out of the chair, because they did not agree with my plot of enslaving humankind.
2. Tried a large variety of homebrew electronic circuits to control the very hefty wheelchair motors from a digital source - put them all through a variety of tests - all went POOF in flames and mushroom clouds of silca
3. Tried a commercial (robot power) simple H bridge - POOF (that was expensive - not doing that again)
4. Talked with BOA at length about the mystries and benefits of manly MOSFETs - helped a little describing why a gnd on the gate is needed to turn off the MOSFET
5. BOA comes up with the BOA Bridge - a fabuloso simple electronic design
6. Robologist suggests a gate driver - turns out in my case he was right
7. So after much frustration and dead MOSFETS here is the latest revision of the controller circuit. It has not gone through ALL of the tests, however, the ones it has gone through it has not heated up at all. So I beleive this circuit can drive a 200 lbs heavy duty electric wheelchair without a problem - (more to come)
* the schematic is missing the IGBT gate driver
*Ugly free form circuit - I love wire nuts !
On the digital control front :
1. had computer PWM control wheelchair motors through parallel port - found this jerky an not reliaible due to task scheduling and interrupting cpu with a variety of "other tasks" 2. got a cheap DIO card from ebay - to try instead of the parallel port - card did not support pwm - or at least I could not figure out a way to do it 3. got a cheaper Arduino - hooked it up to the computer - Works Brilliantly 4. working on software for the computer & Arduino so that the Computer does heavy number crunching of video image processing from webcam and Arduino is kindof a dumb slave - but a very good dumb slave. I think of it like the dinosaur Stegosaurus - it had two not very big brains - or like the coretex and the cerebellum - a big brain and something that is responsible for reflexes and "staying alive" 5. Setup is Computer -> Arduino -> BOA Bridge -> Wheels . The nice thing about a regular computer it that it is extremely powerful and easy to do things we already expect. For example wireless. So I can do this “My Computer” --(wirless)–> Computer -> Arduino -> BOA Bridge -> Wheels. My computer can use a keyboard or joystick or game paddle etc. so this is pretty easy “Joystick” -> “My Computer” --(wirless miles away)-> Computer -> Arduino -> BOA Bridge -> Wheels… get my drift?
Here is some of the gui portion of the code I have been working on …
The program is a “myrobotlab” framework. Its a modular service based Java application. I am developing it with eclipse, but at some point I should be able to give it away so others can “program/configure” it to their desire - using a graphics - “much like matlab” although I’ve never used matlab (heh)
(see the tabs in the picutre) The individual Services running for the above example are :
Arduino Service ParallelPort service GUI Service Video Service Servo Service (2 of them) Joystick Service
The Arduion runs Decimilia - but really there would be no reason why it could not run other Arduino varieties if I got the specs on them. Nor for that matter any serial connected uProcessor. When I get ready to release 0.9 version, I was hoping for donations of different microcontrollers so I could create services for them… always more stuff …
- too top heavy for unmanned roaming, it will send the F50 off course. 
