Banana
Is there any way we can get our mottoes that turn our wheels to move faster?
Quote 0 0
gyoung
1. make sure you are setting the motor to maximum PWM value: 127
2. make gears or use belt/pulley to give a better ratio. 

The motors have a fixed RPM. You cannot modify them, nor are you allowed to modify them.
The motor specs are found in the BRI File Manager:

MAIN / Public Resources & Training / Kits

Quote 0 0
jgraber
If you use bigger wheels with the same motors, the motors will move faster because the robot is moving faster.
If you use the small fast motors for wheels, rather than the large strong motors, the robot/motors will move faster.
Quote 0 0
dmgauntt
It may seem counterintuitive that the robot will move faster with the small motors; intuition here happens to be correct.

It is often not true that the robot will move faster with the smaller motors.  They have a higher top speed, but that is only when they are not turning anything.  When you put a load on a DC motor, it will slow down; if you put a large enough load on the motor, it will stop completely.  With the smaller motors, this stall torque is 9.5 in-lbs, and with the larger motors it is 23 in-lbs.  That means that if you put a load torque of 10 in-lbs on each motor, the smaller motor will have stopped entirely, but the large one will have slowed to about half of its top speed (more precisely (23-10)/23=56% of top speed).

How does this apply to a BEST robot?  Suppose that you have a very light using only 3 wheels; the two drive wheels and the small skate wheel.  As the robot moves, there is very little friction, and so it doesn't take a lot of torque to move the robot.  In that case, the smaller motors may be able to drive the robot faster than the large motors.

However, if you have a heavy robot that uses skids rather than wheels, there is a lot more friction to overcome to get the robot to move, so it will take a of torque from the motors to move it.  If the robot is heavy enough,  the small motors will move the robot much more slowly than the large motors, and quite possibly the robot won't budge with the small motors.

One final consideration: my experience is that a faster robot will be more difficult to control.  This is not a speed competition, it is a celerity competition.*  If you reach the copper bucket in half the time of everyone else, but can't pick up anything because you don't have fine control of your robot, you will not do well.

* Speed is inversely proportional to the time it takes to move a given distance.  Celerity is inversely proportional to the time it takes to accomplish a given task.



Quote 0 0
jgraber
Intuition varies based on experience.

If the back torque on wheel motors is less than 5in-lbs,
 then using the small faster motors for wheels will always be faster.

Control/celerity for a given setup are usually improved by some combination of practice and software assistance.

My hypothesis is this:
  For the same top speed, it will be easier to control
1  faster motors (less geared down) with smaller wheels (less back torque)
 than
2 slower motors (more reduction gears) with wheels large enough (more back torque) to give same speed.
But "easier to control" is a qualitative preference, rather than quantitative. 

I'll see about adding survey questions about wheel diameter and motor selection to the compliance forms
for Texas BEST Regional.   That will be a larger sample of state of the practice than our two experiences,
and some quantitative backup data.
  Maybe there will be a common approach for wheel motor assignment among all the semi-finalists or finalists.

It would be interesting to consider a theoretical game where it is clear that mechanisms
need one fast and one strong motor,  leaving a mixed motor pair to use for wheels of different diameters.
It may be possible to make a very narrow (5" wheel track) robot this way.

Quote 0 0
EvansPhysics

Just a summary of my teams experience this year:

 

Early calculations showed that we would achieve higher speeds with the smaller motors and wheels that would fit the dimensions the team members desired.  Physical tests confirmed that this was the case and that the small motors were very ineffective with sliders.  This would free up the larger motors for beefier task.

However, as the robot evolved and the team added and reinforced features, the weight of the robot increased – more than anticipated. With the weight of a loaded mine cart, the small motors were no longer able to reliably navigate the tunnel lips.  The drivers had to rock the robot back and forth to eventually get through.  It drastically impaired our performance. 

This is leading the team to consider modifications before the regional.  Specifically, they are considering switching the motors and adjusting any corresponding gear ratios as well as reducing weight.

 

One additional complication was that the repeated sudden direction changes led to the gears stripping in one of the motors.  The bright side of this is that the team now knows how to repair the gearbox and the importance of programming a damper to reduce the impact of sudden direction changes.

 

A different item of concern with the small motors is simply their ruggedness.  While not the best practice, it is common to have wheels mounted directly on the motor shafts.  (We may move away from this.)  This puts a lot of lateral force on the motor shaft.  Previously, we had never noticed an issue with this, but with our bulky robot and continued impacts on the tunnel lips, our drive shafts have started to have some lateral play.  We even had the thin screws that mount the gearbox onto the motor break.  This is definitely not the way a motor should be treated and we have learned our lessons.  The large motors have a stronger construction, but we are still going to treat them more respectfully

 

In summary… unless you have a very light robot, large motors are probably the best choice for drive motors.

Quote 0 0
jgraber
It looks like you are at PennStateDubois hub, which competed last weekend, and moving on to the North Plains regional Dec 3,4.  
Thats lots of time to rebuild and practice compared to us: Dallas BEST game day is Nov7 followed by Nov14 Texas BEST regional.

Thanks for that interesting history, highlighting many of the BEST Engineering practices of evaluate/refine/iterate.
If I understand correctly,
  your team did use small motors on wheels during competition,  
  but had trouble going over the thresholds,
  so now you are considering swapping to large motors on wheels for regional, but haven't yet?
Did I get that right?

re Striped motor gearbox:
  This was more common on previous small motors with 3/16" shafts,
but we replaced all those gearboxes with sturdier ones during the 1/4" shaft upgrade,
after checking that the new design could handle eqv of 24 lb robot start/stop.
Yes, Program slew rate control is a big help to reduce these stresses, as well as reduce peak demand through electrical system.
 What size motor stripped the gear box?    
 What is your wheel size?
 Current weight?
Feel free to PrivateMessage or email me a picture.

re wheel direct mount vs motor axle forces:
  You might try building cup shaped wheels such that the floor contact point is directly under the gearbox axle.
That should reduce the vertical torque on the gearbox axle, right?

|
|-Motor  Ascii art of usual direct mount, flat wheel
|
^ floor contact point

vs cup wheel
/||
|-Motor
\||
  ^ floor contact point

Quote 0 0
EvansPhysics
one week turnaround for regionals is really fast... I don't think we could do it in much less than the six weeks we have.  But that is mainly because we travel from PA to ND for regionals.  We try to take around the 20 most active team members and expect our travel and lodging costs to approach $15000.  (We fly, but a charter bus is about the same cost.)  We exert a LOT of effort fundraising since we receive virtually no financial assistance from our school (budget issues).  Hopefully BEST takes off a bit more in PA (Penn State might try expanding it through other branch campuses in the future) and we can eventually have a Northeast regional.

Your summary is essentially correct...

Small drive motors
8" drive wheels
8" non drive omni-wheels
a hefty 23+ lbs (Really increased as we modified) - add about 6 lbs for the mine cart and PVC... almost 30 lbs during the game

We stripped a new small motor.. and broke screws that hold the gear box on the motor - even though the motor is attached to the mounting plate by the gear box.  other than the forces from tensioning these screws, there should not be any significant forces other than torque loads.  (but without looking inside the gear box, I may not be correct.)

Small Motor.JPG 

I'll send a pick of the robot later when I have a chance.
Quote 0 0
David K
If you are using the small motors to drive the wheels, make sure to check that the screws that hold the gearbox housing to the motor can are tight on a regular basis.  If they get loose, the torque path changes from clamp-up through the stack-up, to fastener bending of these very small screws (and the screws will break).

The large reversing loads from using these motors as drive motors will cause these screws to loosen up.
Quote 0 0
widerbestrobot
His team has programming against speed change gear stripping.
Quote 0 0
jgraber
One theory of bug-killing is "kill the head, kill the heart",  meaning use both head (programming) solutions and heart (physical) solutions.
Tightening motor screws is the physical fix.  
Programming slew rate control is the programming fix.
  Think of what a great notebook entry it would make if you show how you drilled holes in the motor mount, and in the wheel and hub,
such that you can do this motor screw tightening maintenance without removing the wheels.
Quote 0 0
EvansPhysics
Hey widersbestbot,

Looking forward to seeing your team at regionals.  Hopefully I'll have time to look over your code this time.  

Our slew rate control came about after we first had an issue with the gears in the motor.  We should have done it first, but... The loose gearbox screws showed up with the slew rate control.  All of this happened after our notebook was due, so we have a few changes to document before regionals.
Quote 0 0
jgraber
I have data to generate statistics for Dallas BEST motor choice vs wheelsize, and I hope to collect data at Texas BEST also, and then do a report of some kind.
A team I advise to use small motors, based on a demo bot testing, could not get over the threshold with small motors.  
They changed motors before round 4 and had more success. 

I should retract or qualify my comments on small motors for wheels:
   There is a lot more safety margin when using large motors for the wheels.

Quote 0 0
dmgauntt
I just ran some calculations based on the motor specs, and found that ultimately the difference between the large and small motors is not as large as I had thought.

The ratings on the motors are:
MotorLargeSmall 
Free speed4390RPM
Stall torque23.539.49in*lbs
Max power252.95213.53in*lbs/min


The maximum power is half of the free speed times half of the stall torque.  Note that the large motor has only 18% more maximum power than the small motor.

The robot speed is  the motor speed times the wheel diameter, and the drive force (per wheel) is the motor torque divided by the wheel diameter.

Suppose that you put 8 inch wheels on the drive motors.    You can then generate a speed vs. force plot for the drive wheel for each of the two motor sizes.

[application]

If there is 2 lb of friction on the loaded robot (1 lb per drive wheel), then using a small motor you would have a maximum speed of about 10 in/sec, and using the large motor the maximum speed would be about 20 in/sec.

Suppose that you change the wheel size for the small motor to 3.8 inches.  The plot then looks like this:

[application]

The robot speed at 1 lb friction per wheel is still higher for the large motor, but not by much.

Of course, this is based on an ideal model of the motor performance, and ignores the maximum torque on the gearbox; that would be the torque that damages the gear head.  My team (currently in the 9th year of competition) has always used the large motors for the drive wheels, has never used slew rate control in the software, and has never damaged the gear head.

I just wish I could say that about servos!

Quote 0 0
jgraber
Your graphs just show broken link [application]  for me in both Chrome and IE.  
Can you link a spreadsheet?   I'll post mine here on the technical file manager shortly.

The robot speed is  the motor speed times the wheel diameter,
I thought the robot speed in ips (inches per second)  is motor speed in rps * circumference of the wheel in inches= RPS * PI * 2 * Radius

and the drive force (per wheel) is the motor torque divided by the wheel diameter.
I thought the drive force is the motor torque divided by the wheel radius.(not diameter) 

Then adding (per wheel) adds another 2x somewhere, which just confuses me;
I'd rather pretend a one drive motor robot of half the weight.

 Maybe if I type out the way I think of it, we can see if my way makes sense to anyone.

Assume that drive wheels are like a winch drum radius R inches lifting a weight M lbs.
Torque on the motor is R*M = inch-pounds.
For a R=4 inch Radius (8" diameter) winch drum, lifting M=1 lb weight,
the torque is 4 in-lbs.
  The Large motor speed is 43*(1-4/23)  = 35.5 rpm ;  rim speed = PI*D*rpm/60 = 14.9 in/sec
  The Small motor speed is 90*(1-4/9.5) = 52.1 rpm ;  rim speed = PI*D*rpm/60 = 21.8 in/sec

For the same size winch drum (or wheel), the small motor is faster until >6 in-lb torque required.
If you use smaller wheels (6.3" diameter) on the smaller motor, then the small motor is faster until > 7in-lbs.  
Does my formula of rpm vs torque make sense?  MaxRPM * (100% - loadtorque/maxtorque)

For Paydirt,  a faster max speed doesn't help if you can step over the threshold, which was my team's problem with small motors on 8" wheels, and the mine cart weight mostly over the skids instead of over the wheels (as I had envisioned). 

What's the torque-required formula for M pounds, D wheel diameter, H step size?
This used to be in an old BEST handbook, but I can't find it now.
H is probably 1" given a 3/4" threshold floating on top of carpet that the robot wheels sink into by 1/4".

Something like T=R*weight*sin(theta),  where theta is the ramp angle, but then how do we calculate the ramp angle of a step, given H and R?  Maybe atan(H/R)?  The Angle Theta goes down as Radius goes up, but increasing radius increases the torque linearly.   I added this to rows 39+ of the spreadsheet.   Its probably not right though, since a wheel radius below step size maxes out at theta 90 of which sin(90) =1 which is just back to T=R * lbs definition.

Quote 0 0