I am building a throttle by wire conversion for an older carbureted engine. I am wondering which would be most effective at pulling a cable with a rotational motion, a servo or a stepper motor, I need to make sure that the response time is as quick as possible, or if I use a large gear to drive a small gear I could just use a high torque servo or motor to achieve the speed.
How fast must the butterfly valve rotate?
How many degrees must it rotate?
Are you going to connect the motor direct to the butterfly spindle?
A high quality servo seems like a simple solution.
What is the worst that can happen if (when?) your system fails? Almost all of the expense in commercial products like this is due to the need to deal with risk.
...R
there is a return spring that the motor or servo fights against, in theory if the servo fails it will return shut. I will have to calculate the rotations necessary although that would also depend on weather or not it is geared to increase the speed of the motion of the carbs. The idea at this point is essentially just to have a mechanism inside of a box pull a short cable the other end being connected to the normal throttle shaft on the bank of carbs.
Ashdricky:
there is a return spring that the motor or servo fights against, in theory if the servo fails it will return shut.
DON'T rely on that assumption for at least two reasons.
If the code crashes it may leave the servo with a proper signal so that it maintains its position.
If the power fails the internal mechanical resistance of the servo may be enough to prevent the throttle closing.
...R
do you know a way to have a failsafe that will work? I do not know how production fly by wire systems work in respect to failsafe
Ashdricky:
do you know a way to have a failsafe that will work? I do not know how production fly by wire systems work in respect to failsafe
I have no idea how they work either. One possibility is that they have two (or three) separate microprocessors monitoring each other.
I suspect it is this sort of problem that has led Toyota and others to recall millions of cars.
...R
Ashdricky:
do you know a way to have a failsafe that will work? I do not know how production fly by wire systems work in respect to failsafe
How safety critical is this BTW?
I would suggest ensuring the system is mechnically fail-safe if the electric power is
absent, and provide a big red kill-switch to isolate the supply in emergencies. This
implies a motor with minimal friction when unpowered, used with closed-loop control.
Kill the power and the return spring wins. Something like a high torque BLDC would
be plausible in the absence of concrete numbers. Think something like:
(and probably an absolute encoder)
I don't believe a kill-switch is the appropriate solution for preventing a car crash. The system must be safe enough without requiring human intervention in the case of a fault.
...R
We don't know if this engine is on a vehicle yet do we?
MarkT:
We don't know if this engine is on a vehicle yet do we?
I freely admit I jumped to that conclusion. I don't immediately see the point of "souping up" a stationary engine. But it certainly won't be the first time I was wrong (nor the last, I suspect).
...R