Please help confirm my motor selection

Spend a little bit more for a better motor:

http://www.lynxmotion.com/Product.aspx?productIDA8&CategoryIDq

Though it's rated at 9.6 volts, it'll run a little slower at 6 volts and
a little faster at 12, with associated torque decrease/increase. As
you'll likely run the motor through an H-bridge anyway, and most entail
a voltage drop, from a 12 volt source you'll get about 10-10.5 volts,
close to the spec rating.

This motor has better torque and planetary gearing tends to last longer.
It's only $8 more, and it gives you wiggle room. This is the better
approach when creating prototypes, especially if the price is about
equal. The only disadvantage, much may not be for your project, is that
the motor and gearbox is about 4" long, so if you mount them end to end,
that sets your minimum vehicle body width to roughly 8". Sounds about
right for a medium-sized tracked vehicle, though...

-- Gordon

---snip---

I'd start by buying a couple of battery powered screw drivers, take them
apart, and mount them on your robot. You can buy them for less than ten
bucks at Harbor Freight. Then, when you get into requiring more power, buy
some variable speed 3/8" drills and connect the trigger mechanism to RC
servos for full speed control and a lot of power.

Wayne

Well, for reference, there's what the VEX motors are rated at:

http://www.vexlabs.com/vex-robotics-motor-kit.shtml

Free Speed  100 rpm @ 7.5 volts
Stall Torque 6.5 in-lbs

They are strong enough to drive a VEX bot with the VEX tread kit with two
motors but it's under powered.  The treads have a lot of friction in them.
I don't think it can reach your 1 ft per second speed - maybe half that?
Mine is sitting next to me but the it's got the wrong software loaded in it
so I can't test it's speed.  I tried to make it go faster using about 3
different gear configurations but the speed improvement was minimal and the
power loss was unacceptable.  So I left it in the direct drive
configuration.

The bot is probably about 5 lbs and I doubt the vex motors could make it
climb straight up a wall (but maybe it could).

So, I would say the motor above should work (since it's slightly stronger
than the VEX motors, but to hit the speeds you were looking for, I suspect
you are going to need something larger to overcome the friction loss in the
treads.

Also, the RPM rating of a motor is it's no load rating.  It will always be
less under load (just spinning the treads with the bot held off the ground
will be a substantial load).  The full specs on your motor give the
performance graphs so you can see the speed vs torque curve.  The spec also
lists the rated load as 1.14 kg-cm and the rpm at rated load as 95 (which
is under your 100-200 target range.  Max efficiency happens at around 85
rpm and max power output happens at around 55 rpm.

It also says, "Do not exceed rated load. Damage may occur!".  So I take
that to mean you shouldn't drive it at over 1.14 kg-cm (.99 in-lbs) which
is odd because that cuts out most of the performance graph they give.  But
that's what it seems to say.  It might be an indication of how much load
the gears are designed to take.

So, like with the vex motors, I think that motor will probably be running
quite a bit under the 100 rpm minimum you were looking for under the load
caused by the friction of the tracks, and it looks like the rated load of
the gears might not hold up very well under your intended application
(rough outdoor terrain).

The motor Gordon suggested looks like it would do a better job at meeting
your needs. But it's a lot more powerful (and uses more energy) with a
stall torque of 7.46 ft-lbs (89.52 in-lbs).  But if you want to actually
reach your 200 rpm speed, you might need something with that high of a
rating.  Since speed is generally linear with torque, you can estimate the
torque delivered at 200 rpm.

The speed is 271 rpm with no load, and 0 rpm at 89.52 in-lb, so the
equation is:

   speed = 271 - 271t/89.52
         = 271 - 3.027 t

so at 200 rpm:

   200 = 271 - 3.027 t
   t = (271 - 200) / 3.027
   t = 23.5 in-lbs

which is about 4 times what you calculated was needed for the full 10 lb
bot to climb straight up.  But your calculations didn't include the load
due to the friction of just turning the treads (which I have no idea how to
calculate).

Or, we can calculate how fast it will go straight up using your 5.9 in-lbs
number:

  rpm = 271 - 3.027 * 5.9 = 253 rpm

or with only one motor and one tread driving it:

  rpm = 271 - 3.027 * 11.9 = 234 rpm

But again, this doesn't include the friction of the treads (which may only
be on the order of 1 or 2 in-lbs for vex-style friction bearings and less
if you use real bearings).

So Gordon's motor selection will be overkill in terms of the speeds and
power you wanted to reach, but the motor you selected looks like it's not
going to reach the speeds you were aiming for and may not hold up well
because you are going past the rated load which it says not to do.

You can always reduce the power of a motor by driving it at less than 100%
(or with less voltage) but if the motor doesn't have enough power, there's
not much you can do.

--
Curt Welch                                            http://CurtWelch.Com/
curt@kcwc.com                                        http://NewsReader.Com/

Thank you so much everyone for taking the time to lend me some advice.
I am going to look a little bit more now that I have some fresh
thoughts but those planetary gear motors are negligibly more than what
I had budgeted and as Curt said, I can run them below 100% and
definately meet my specs.  Besides, I can always cannibalize them for
the inevitable "next bot"!  My only initial concern with them is that
the locked shaft current went from 1.5A in the ones I was looking at to
9.1A so I am going to need to look into some better controller$ than I
had planned.  I was initially considering Parallax Motor Mind B
(http://www.parallax.com/detail.asp?product_id '961) which is only 6A
peak, but I am only really starting to look at controllers so I wasn't
married to those ones (although I liked the built in tachometer).

Thanks,
Michael



Curt Welch wrote:

Pololu has a dual high-current motor driver carrier that may be of
interest -- and the page includes some nice detail on handling
high-current motors:  <http://www.pololu.com/products/pololu/0707/>

HTH,
- Joe

(sorry if this is a double post, my newsreader is complaining)

That motor controller does look pretty good.  I also came across this
one (http://www.dimensionengineering.com/Sabertooth2X10.htm ) which
costs a little more.  It seems roughly equivalent in terms of power but
provides for regenerative braking... how cool is that?!  But in the end
that is actually more of just a novelty for my particular application
so I am going to really consider the ones you mentioned.

I was still *hoping* that I could find motors requiring less current to
save me some cost on the controllers and also to provide better battery
life (and I am looking to see if I can trade-off some of my speed or
weight requirements to make it happen)

Along those lines, I came across these motors: (
http://www.lynxmotion.com/Product.aspx?productID%0&CategoryID ).  I
didn't find them at first because they appear to be out of stock or
maybe even discontinued but it seems I can find them elsewhere online.
They look to provde ~twice the torque of the planetary motors that we
discussed above
(http://www.lynxmotion.com/Product.aspx?productIDA8&CategoryIDq )
using far less current.  This is from being 12v instead of 9.6v and
also that they are geared down further.

Planetary Gear Motor - 9.6vdc 51:1 271rpm (LS-550S) (6mm shaft)
   http://www.lynxmotion.com/Product.aspx?productIDA8&CategoryIDq
   Stall Torque = 7.46 ft. lbs <--- 8595 g*cm
   Current (at 9.6v no load) = 1.8A
   Current (at 9.6v locked shaft) = 9.1A

Gear Head Motor - 12vdc 50:1 152rpm (6mm shaft)
   http://www.lynxmotion.com/Product.aspx?productID%0&CategoryID 
   Stall Torque = 231.5 oz-in <--- 16634 g*cm
   Current (at 12v no load) = 145mA
   Current (at 12v locked shaft) = 3.8A


(I think).  And just guess-timating I think running at no-load 152rpm
may be able to still meet my 100rpm under-load requirement once it
accelerates up to speed (but maybe not).  Double the torque is nice
too!

The other thing is that I believe the Gear Head one can make use of
these motor-integrated quadrature encoder kits
(http://www.lynxmotion.com/Product.aspx?productIDD8&CategoryID )
should I want to add them in the future (I am starting out with
non-quadrature encoders for now).

As of right now I am still planning to get the Planetary ones and a
good chance the Pololu controller.

Any thoughts?   (my wife is getting sick of me talking about motors and
controllers and told me to just buy the f*$&#n things already!)

-Michael

Joe Strout wrote:

HAH!... oops read that planetary at 7.46 In*Lbs not Ft*Lbs!  Now that
9.1 amps makes more sense!

Michael wrote:

Note that the 9.1 amp draw is from a locked shaft, which on a tracked
vehicle will almost never happen, or if it does, you'll want a stall
sensor to prevent it from happening. Most of the electronics have a
overcurrent sense that will turn them off when the output current is
exceeded, and this is something you do want to take advantage of if you
don't want to be buying new treads all the time. IOW, this is a
worst-case scenario that you want to avoid at all costs.

I would guess the nominal load current consumption of the motor to be
closer to 2.5-3.5 amps, which is still reasonable for a number of the
drivers out there. If you'll be using 12 volts, a DMOS controller is
well suited.

It's a fair bet the motor will also run well at less than the rates 9.6
volts, which means you can get excellent torque from just 6 volts -- and
the current draws will be about 30% less, too. That means a lighter
robot because the batteries won't be so large.

FWIW, I like avoiding pound-force inch expressions to avoid this very
problem! Sticking with just imperial measurements, you can convert to
oz-in, or lb-ft. Here's a Web page that makes it pretty easy:
http://www.convert-me.com/en/convert/torque

According to the calcs: 7.46 pound-force foot = 89.52 pound-force inch =
1432 ounce-force inch.


-- Gordon

(sorry if this is a double post, my newsreader is complaining)

That motor controller does look pretty good.  I also came across this
one (http://www.dimensionengineering.com/Sabertooth2X10.htm ) which
costs a little more.  It seems roughly equivalent in terms of power but
provides for regenerative braking... how cool is that?!  But in the end
that is actually more of just a novelty for my particular application
so I am going to really consider the ones you mentioned.

I was still *hoping* that I could find motors requiring less current to
save me some cost on the controllers and also to provide better battery
life (and I am looking to see if I can trade-off some of my speed or
weight requirements to make it happen)

Along those lines, I came across these motors: (
http://www.lynxmotion.com/Product.aspx?productID%0&CategoryID ).  I
didn't find them at first because they appear to be out of stock or
maybe even discontinued but it seems I can find them elsewhere online.
They look to provde ~twice the torque of the planetary motors that we
discussed above
(http://www.lynxmotion.com/Product.aspx?productIDA8&CategoryIDq )
using far less current.  This is from being 12v instead of 9.6v and
also that they are geared down further.

Planetary Gear Motor - 9.6vdc 51:1 271rpm (LS-550S) (6mm shaft)
   http://www.lynxmotion.com/Product.aspx?productIDA8&CategoryIDq
   Stall Torque = 7.46 ft. lbs <--- 8595 g*cm
   Current (at 9.6v no load) = 1.8A
   Current (at 9.6v locked shaft) = 9.1A

Gear Head Motor - 12vdc 50:1 152rpm (6mm shaft)
   http://www.lynxmotion.com/Product.aspx?productID%0&CategoryID 
   Stall Torque = 231.5 oz-in <--- 16634 g*cm
   Current (at 12v no load) = 145mA
   Current (at 12v locked shaft) = 3.8A


(I think).  And just guess-timating I think running at no-load 152rpm
may be able to still meet my 100rpm under-load requirement once it
accelerates up to speed (but maybe not).  Double the torque is nice
too!

The other thing is that I believe the Gear Head one can make use of
these motor-integrated quadrature encoder kits
(http://www.lynxmotion.com/Product.aspx?productIDD8&CategoryID )
should I want to add them in the future (I am starting out with
non-quadrature encoders for now).

As of right now I am still planning to get the Planetary ones and a
good chance the Pololu controller.

Any thoughts?   (my wife is getting sick of me talking about motors and
controllers and told me to just buy the f*$&#n things already!)

-Michael

Joe Strout wrote:

Thank you so much everyone for taking the time to lend me some advice.
I am going to look a little bit more now that I have some fresh
thoughts but those planetary gear motors are negligibly more than what
I had budgeted and as Curt said, I can run them below 100% and
definately meet my specs.  Besides, I can always cannibalize them for
the inevitable "next bot"!  My only initial concern with them is that
the locked shaft current went from 1.5A in the ones I was looking at to
9.1A so I am going to need to look into some better controller$ than I
had planned.  I was initially considering Parallax Motor Mind B
(http://www.parallax.com/detail.asp?product_id '961) which is only 6A
peak, but I am only really starting to look at controllers so I wasn't
married to those ones (although I liked the built in tachometer).

Thanks,
Michael



Curt Welch wrote: