On Saturday 27 August 2005 07:16 pm, Denny Esterline scribbled:
> Ok, I'm thinking about hand controls for the clutch on a stick shift car.
> Sure, you could just get an automatic and not need a clutch - but what fun
> is that?

How about Tiptronic (a.k.a. many other things based on each manufacturer)?  
Also, many newer cars already provide "paddle shifters" on the steering 
wheel.  The advantage to this system is quicker shifts, compared to any 
manual system, even a powered system as you are designing.

BTW, are you also planning to power the shifter and automate the whole system? 
:-)


> Obviously a motor, a couple limit stops and a toggle switch would engage
> and disengage the clutch, but I'm thinking of something a little more
> refined; at least proportional control, possibly a force feedback system.

Why force?  I would think position would be more direct.


> At the moment I'm thinking of using two pots, one for the control and one
> for the position sense.

Check out BEI and Novotechnik for pedal position sensors -- basically pots 
with a lower angular range which might be more appropriate here.


> The electronics could be all analog, just a couple
> high power op-amps driving the motor so the feedback tracks the hand
> control (I'm thinking a squeeze grip on the shifter lever). Although I'd
> probably reach for a PIC instead (A/D both pots and PID the motor).

Will any reasonable-sized motor be fast enough for this?  Especially if you're 
using a manual (non-hydraulic) clutch.  If you don't need to feather the 
clutch, perhaps you might want to use a hydraulic cylinder/actuator.  The 
necessary pieces (plumbing, valves, etc) should be available easily from the 
low-rider crowd.


> I have a couple thoughts about sensing the force applied to the clutch
> linkage. The mechanics could be designed with a strain gage to actually
> measure the force, but measuring the actuator current could be effective
> also. It might also be possible to add some software that takes the PID
> values and calculates the approximate load. That load data would be
> translated back to a feedback actuator on the hand controls. Though I'm not
> entirely sold on the usefulness of force feedback on a clutch.

I still think position woule be more useful for this.  But I can't see 
position feedback/PID making much of a difference -- once setup properly, a 
pot-directly controlling the clutch position should be all you need.  But 
position feedback could be good for a safety mechanism.  Without the cable, 
etc, a clutch is normally engaged, so any failure with the motor/system and 
it will default to the "unsafe" position.


> Opinions, ideas and criticisms sought.
>
> -Denny


-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist