Ball screw reduction ratio

This seems *awefully* fast for a ballscrew mechanism. These are
generally made for precision, not speed, though speed is a relative
thing.

Sources of ballscrew components like SKF provide engineering white
papers with design calculations. Sounds like you're doing something for
a steering mechanism. You may also want to look into rack-and-pinion
(easier mechanically), hydraulic, and pneumatic. For the latter two you
can provide positional feedback using a pot or encoder, and a PWM valve.
These might be more durable, and you'd definitely get the speeds and
torque you're looking for. The positional accuracy will depend on your
control system, and the quality/type of proportional controlled valves
you use.

-- Gordon

"Gordon McComb"

I'm monitoring Ebay hoping to get a cheap ball screw set. I have the DC
motor with the caracteristics I mentioned before, and now I have to
translate it to linear motion. This is not for the steering, but for
actuating the breaking mechanism. On my ATV, breaking is mechanical (steel
cable and lever, as opposed to hydraulic), so it needs considerable force to
get actuated. I will probably add a reduction lever or even a gearhead to
the motor to achieve greater reduction. I probably don't need a very precise
positional accuracy, a linear pot or a bunch of switches will do.

I thought about replacing the whole breaking system by a hydraulic one, but
I think I'd be spending too much time on the mechanics, which I'm not very
experienced with.

Cheers

Padu

Since you're not looking for accuracy or smoothness in motion, a simpler
leadscrew may work better, and would probably last longer. It works on
the same basic principle as a ballscrew, but doesn't rely on the
reciprocating ball bearings. These aren't all that expensive new; Reid
Tool and others sell these in standard lengths, with various pitches and
diameters to suit your needs. You'll also need some bearings and blocks
to set the shaft into. These are actually more expensive than a 12" lead
and matching nut!

My concern would be the "hammer effect" you'd get powering that motor to
3000 rpm for the under one second you need for full travel.

Are you sure a 200-300 rpm 12 volt gearmotor, attached to a long enough
lever, wouldn't be enough to actuate the brakes? They're spring return
already, I'd imagine. You can test the torque needs with a fish scale.
Try it along different lengths of a 6-12 inch lever.

-- Gordon

"Gordon McComb"

I've thought about lead screw, but I didn't like their efficiency. Since the
solution goes on a mobile platform, I was thinking about the energy savings.
Now you expose a new variable that I hadn't thought before (longevity). I'll
have to rethink it, but since it will be used for breaking, I believe the
duty cycle is very forgiving.


Yes, if no gear reduction is in place, I will have to carefully design the
acceleration curve in order to avoid jerking. The thing is that we are out
of [official] budget, and a new gearhead for this motor costs over $600
bucks. I'm monitoring ebay for something compatible with my motor, but so
far no luck.



I have a 12V 190RPM laying around, but I need to perform a few tests to
check its torque (ebay stuff, no datasheet). If I had the linear motion
device (either a ball screw or a lead screw), then I could go one step
further and test how it performs actuating the brakes.

Another thing that I haven't decided yet is if I will actuate brakes (front
and rear) individually or have only one actuator for both. The first has a
technical advantage for off-road riding, but increases the complexity of the
system in all aspects (hardware, electronics and software)

Cheers

Padu

Padu, have you checked our fabulous
http://www.industrialliquidators.com/locations.html
here in San Diego?


(front

"Wayne Lundberg"

Yes, the other day I made the mistake of going there with my wife, and
needless to say, after 1 hour browsing the nice stuff they have, she was
pretty upset.
Unfortunately they don't have a good selection of gearheads (they have lots
of stepper motors though), and the only ball screw I could find they didn't
have the nut.

I'm still working on the steering mechanism of my rover, so I have plenty of
time to lurk on ebay.

Cheers

Padu

I have a ball screw, nut and some motors I'm not using. Stop by and take a
look. Make an offer... money or talent later on in my new project.
It's up on the rack and I'm too lazy to get to it now. But it's about 18
inches long, about 5/8 dia and the nut is circulating balls for high
accuracy.

"Wayne Lundberg"

Thanks for the offer. Right now I'm monitoring a mini ball screw on ebay
that would be perfect in both dimensions (price and length) for my project.
Right now price is $1, but I foresee buying it for about $20-$30.
If I don't win the auction I'll stop by. We should really get together
sometime to chat over some beers about our projects, seems like there are
lots of overlaps between them.

Cheers

Padu

Sounds good to me!

I have not ridden ATC's, but old motorcycles used lever and cable front
brakes and pedal and linkage rear brakes. The braking action on either
the front or rear brake is going to be proportional to the force applied
to the cable more then the position, although it's probably very
non-linear. It is also not likely to be a constant relationship, as
brake heating and wear will change it for you.

Almost certainly, you don't want to use the same actuator on the front
and rear brakes. On motorcycles and I assume ATC's, the weight shifts
forward very heavily in braking operation. The result of this is it's
very easy to lock up he rear brake which can cause unpleasant results!
If I had the choice of a single actuator driving both brakes or only
driving the front brake, I would take the front brake only. ATC people
may have a different opinion, but this is something to watch for.

Good Luck,
Bob

Let's see if I can remember any mechanics...

Force, rather than torque (as the output is linear rather than
rotational). You need to know the diameter of the screw as well as the
pitch (screw lead). That'll let you work out the angle of the threads
to the axis of the screw, and convert the torque of the motor into a
force at the edge of the screw. You already know the input force is
normal (at right angles) to the axis of the screw. The reaction (as in
Newton's 3rd law) of the screw to the input force is normal to the
threads of the screw. The output force is along the axis of the screw.
Now you can draw a diagram like this:


     F  _
 |<-----/|
 |     /
 | I  /
 |   / R
 |  /
 | /
 V

I (down) is your input force. R (up and to the right) is the reaction,
normal to the screw threads. F (left) is the output force, along the
axis of the screw. (I hope the diagram makes sense, it's a bit tricky
with ASCII).

Plug in I (calculated from your motor torque and the radius of the
screw), the angle of R to I (from your screw diameter and pitch), and
do a bit of trigonometry to work out F, your output force. This
neglects friction of course, but I expect you can find a typical
efficiency value for ball screws on the web somewhere, and multiply
your answer for F by that.


Tim
--
Did I really still have that sig?

You don't need to know the diameter/radius; that term cancels out.
The truth is, diameter is not even very well defined in this situation.

tangential force = torque / radius

effective slope = pitch / circumference
= pitch / (2*pi*radius)

axial force = tangential force / effective slope
= (torque / radius) / (pitch / (2*pi*radius))
= torque / (pitch / 2*pi)
= 2*pi*torque/pitch

At high force, ballscrews are quite efficient.  If you hold them on
end,
ballnuts will spin and slide down a screw on their own.  Try that with
a
leadscrew!

chris

Let me see if I understood, so axial force in my case would be

2*pi*(1900oz-in)/(6mm)
converting to SI:
2*pi*13Nm/0.006m081N !!

Considering efficiency of ball screw close to 80%, net force would be about
11000N? That would be enough to support 10 times my own weight. Is this
correct?


Cheers


Padu

That is correct.  If you have trouble believing that, think of it this
way:

A 1 meter radius circle has a circumference of 6.28 meters.
The energy required to push 13N around that circle once is:
6.28m * 13N = 84J

This will cause a 0.006m linear motion with 14081N force:
0.006m * 14081N = 84J

Keep in mind that 14081N is the stall torque; it won't actually be able
to push that hard while moving.

chris

Well, that made me very happy. It means that I won't need to use the
1900oz-in motor to push the load. In fact, maybe even a strong stepper could
do it, since the load is static.
I'll do some experiments.

Cheers

Padu