PDA

View Full Version : [Question(s)] In Wheel Motor for 1/5 Car



sam
07-13-2009, 10:13 AM
I am making a 1/5 size model car. I need the wheels to be able to turn 360 degrees. So, I need 4 motors to run the wheels and they need to be inside otherwise they will get in the way when turning.

So I am looking for a motor that is kinda slim (5 cm max) and pretty small (10 cm max diameter). I need it to be already geared down and have enough speed and torque.

So my question is, does anyone know where I can buy a motor like this one?

Thanks,

Sam

ScuD
07-13-2009, 11:26 AM
Couldn't you just use bevel gears through the axis of rotation of the wheels, thus enabling you to use any motor?

But I guess that makes everything a lot more complicated...

ROBOTMAN
07-16-2009, 02:31 PM
How much torque do you need? A micro servo modified for continuous rotation might work. Or just a regular servo. Modifying a servo isn't as hard as it sounds you could even rip the circuit board out and solder right to the motor wires! This seems like a simple solution.

sam
07-18-2009, 09:23 PM
Thanks for the replies guys!

I looked into bevel gears, but the way it's positioned I'm not sure it could work. What I have come across on the servocity site is chains and sprockets that could suit my need perfectly I believe.

As for the torque, I think I have chosen my motor and it is this one :

http://www.robotmarketplace.com/products/ML-42-24.html


Specs :

ML-42
24V

Ratio 24:1
RPM 240rpm
Rated Torque 12 in-lb
Rated Current 2.1 Amps
Rated Output 34.7 W
Stall Torque 90 in-lb
Stall Current 13 Amps
Weight 1.15 lb

I was looking high and low to find the most powerful motor I could find that my Sabertooth 2x10 could run. Since it can go up to 24 volts I was looking for a motor in that range. The one I found suits perfectly. I will just add a 10 amp fuse so my Sabertooth won't overheat.

Now I've run into a second problem in my design. While the wheel will be turning 360 degrees, the wires of the motor will get tangled up. I thought maybe I could hook it up like in one of those big flashlight. It has to copper plating parts on the upper part of the flashlight. They make a circle. Then there are the positive and negative leads that both touch one of the copper circles. So there can be power to the light while turning the head.

Thanks a lot

For the moment I am reading up on how I could use an inexpensive Solarbotics motor to use as a kind of servo to turn the wheels. But I would need to hook up a Pot and manage to control the motor to make it hold.

ScuD
07-19-2009, 11:46 AM
If you're using bearings in the design, those can be used as faux slip-rings as well, though they do not always insure a proper contact... Plus, with 2 amps and dodgy connections (i.e. sliprings) you may want to put your flyback diodes directly across the motor leads, i.e. after the slip rings, to prevent burn-in.

Are you going the analog way of creating a servo system, or digital? I was surprised to see how relatively easy you can make a PID loop with opamps.

sam
07-20-2009, 08:23 AM
If you're using bearings in the design, those can be used as faux slip-rings as well, though they do not always insure a proper contact... Plus, with 2 amps and dodgy connections (i.e. sliprings) you may want to put your flyback diodes directly across the motor leads, i.e. after the slip rings, to prevent burn-in.

Are you going the analog way of creating a servo system, or digital? I was surprised to see how relatively easy you can make a PID loop with opamps.

Thanks for the reply!

Thanks a lot for the slip-ring word which alowed me to figure out what I was looking for was called! :veryhappy:

Yep, if the motor stalls at 13 amps... I need that diode.

I started reading on Openservo and other sites. I will have to learn how to build my own or buy one. If I buy a premade one (should I?), I need the wheels to turn more than 360 degrees, but every Pot I ever had doesn't even turn 360 degrees. So how can I get position feedback (special pot or something else) and will the controller be able to control the servo over 360 degrees?

Edit :

Ok I found some pot that can go over 360 degrees (damn their expensive! like 20- bucks for a 5 turn pot from digikey!) , but controlling them is still un-understood. Will it use the same 1 to 2 ms intevals to control the rotation?

This is the haziest part in my head. Any help would be awesome!

Thanks

ScuD
07-20-2009, 08:51 AM
There's two options really: build a servo from scratch, including electronics, or taking apart a working one and adapting it to suit your needs.

If you're going to adapt one, you'll indeed need one that turns over 360°. You can indeed use those 5 or 10-turn pots, but that's still "only" 3600°, after that you'll still be stuck.

There's relative encoders with an index mark, these are like the encoders in old computer mice, but they have an extra hole in the circumference indicating a start/stop of a full rotation. Still, those are digital signals, so you couldn't hook that up to a standard servo.

You might be able to hack a pot to turn a complete 360°, but it won't last long, plus I don't think the electronics will like it. Complete overshoot resulting in a PID loop going haywire and all that...

I'd seriously consider using a simple geared motor, and attach some kind of optical encoder (or even a magnetic one), and just keep counting pulses until it's in the desired position. Do you truly need full PID ?

sam
07-20-2009, 11:13 AM
I'd seriously consider using a simple geared motor, and attach some kind of optical encoder (or even a magnetic one), and just keep counting pulses until it's in the desired position. Do you truly need full PID ?

I don't think so. I was thinking of using some VEX components but now I'm now sure they would be strong enough after visiting the forum... But it's so useful with all the gears made to fit with each other!

Something like a worm gear would prevent the piece from rotating, and at the same time, I wouldn't use the motor to make it keep the position so it would save the precious energy in the batteries.


More Importantly, do you know where I could buy slip-rings?

Oh, and the schematic on this PDF is in mm right? (I hope so:tongue:) I hate it when they don't put it. It causes a little dought inside of me.

ScuD
07-20-2009, 11:58 AM
If you use a motor with an ordinary or planetary gearbox (i.e. no worm gear) you could short the motor leads to act as a brake. That can usually take quite a load, the higher the gear ratio, the more load it can handle.

Not really sure where you can buy slip-rings, i tend to scavenge vacuum cleaners or (little harder to find) AC motors for them.

PDF does seem like mm's, although it's not clearly marked.

Adrenalynn
07-20-2009, 12:09 PM
Or use an AX-12? They're 360 degree out of the box with feedback, aren't they?

lnxfergy
07-20-2009, 12:57 PM
Or use an AX-12? They're 360 degree out of the box with feedback, aren't they?

They're 360 or feedback really. You set a speed in 360 mode, I don't think you can actually pick the endpoint exactly (you may be able to empirically determine what it takes to rotate say 360 degrees, get it into the region you want and then set the endpoint, but I'm not sure if that would work)

-Fergs

Adrenalynn
07-20-2009, 01:33 PM
Yeah, that's kinda the way I was thinking. But really, if you're trying to get exact positioning - use a wheel encoder and stop trying to... wait for it.... wait for it...

reinvent the wheel.

sam
07-20-2009, 05:04 PM
reinvent the wheel.

I'm not trying to reinvent the wheel! :tongue: That would be stupid!

I was just looking into the pot thing because I saw this at ServoCity.com

http://www.servocity.com/html/spg425a-mr_multi-rotation.html

Looked interesting enough to think about my options.

Now I took a look at the encoders, should I go with optical, magnetic or mechanical? It seems that mechanical is the cheapest and since I need at least 8, it could be practical.

What do you think?

Thanks a lot for the replies,

much appreciated.

Adrenalynn
07-20-2009, 05:42 PM
I would, being the responsible adult, advise you against giving in to the peer pressure to try pot... :tongue:

How much precision do you need?

If you're feeling adventurous, old computer mouses can be gutted to make pretty decent home-rolled encoders. That's all they are anyway.

sam
07-20-2009, 07:12 PM
I would, being the responsible adult, advise you against giving in to the peer pressure to try pot... :tongue:

Advised taken and respected! :tongue:



How much precision do you need?

If you're feeling adventurous, old computer mouses can be gutted to make pretty decent home-rolled encoders. That's all they are anyway.

That's what I was reading on Instructables, but then I saw some selling on digikey for 1.85$, and since I need like 8 encoders, I don't have enough mice anyway.

As for the motor, I just realized that they sell high current motor drivers on digikey. They sell this :

http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=BTS555E3146IN-ND

Driving 1 motor to 165 amps. This isn't an H-Bridge, so what do I do to control the motors speed? What else would I need? Additional heat sinks/fans? Can this be done ? I guess I am completely wrong...

Thanks a lot for the help

Sam

Edit :

Found a few tutorials on making high current motor controllers. I'm going to read more before.

ScuD
07-21-2009, 12:45 PM
Those are actually high-priced "supervised" mosfets. I say supervised because of all the protection built-in.

You could build an H-bridge with those but it'd prove costly..

sam
07-21-2009, 08:01 PM
It looks like it will be indeed expensive to make myself one.

That is why I started looking at this motor controller (OSMC) :

http://www.robotpower.com/products/osmc_info.html

It can handle 160 amps continuously and peaks of 400 amps. It need PWM to control the motors (which won't be too complex since I have many AtMega).

I thought I would connect 2 motors on each side (4 motors total, 2 controllers). A motor controller like this one would permit me using bigger/better motors. So I started looking and found these two motors :

DeWalt 24V Hammerdrill Motor

http://www.robotmarketplace.com/products/BP389010-00.html

and

FIRST CIM Motor

http://www.trossenrobotics.com/store/p/5142-FIRST-CIM-Motor.aspx

The DeWalt one packs a lot more punch and is a tad smaller (I think). It seems also more efficient.

The motor controller can handle two of these motors at nominal amps (around 70 amps for the DeWalt). Worst comes to worst, I will put a fuse to make sure the controller is safe.

The motor controller will handle the load, but can the batteries? I have a couple NiMH batteries, looking on the robot market place site, they say it there packs can handle 70 amps continuously. That probably won't be enough. I was also looking at the Nano-phosphate batteries. (the very last one on this page to power DeWalt motors http://www.duralitebatteries.com/flight/batteries-a123.php?location=CA ) It's a 20/30c Norm/Max Discharge, 19.8v Nominal, 2300 mah

They say 30C max discharge, so it would give a 69 max amp usage. If I use three packs and connect them in parallel, they could discharge at 207 amps? Would this work or would I have to connect one battery to each motor controller separately to achieve the fastest discharge?

Thanks a lot for the information!

Sam

sam
07-23-2009, 10:29 AM
I have decided to dial it down a bit because it would be too expensive and too heavy.

I want to go with brushless DC motors for maximum efficiency.

I have found this motor : http://www.subsonicplanes.com/uploads/DM2820-950__page.pdf

Specs:

FOR PLANES
3 TO 6.75 lbs.

WEIGHT
152 GRAMS

Kv
966

WATTS
800 TO 1200

LIPO CELLS
3 TO 5 CELL

14X7 TO 15X8 PROP ON 3S
12X6 TO 13X6.5 PROP ON 4S


And this motor controller :

FM60A -----> http://www.subsonicplanes.com/uploads/ESC.pdf_dot_dash_sequence.pdf

ESC

MAX CURRENT
60 AMPS

NO BEC use
U-BEC and or Reciever Pack

NiXX BATTERIES
8 TO 18 CELLS

LiXX BATTERIES
3 TO 6 CELLS

WEIGHT
59 GRAMS

In the motor information page, it says that one side of the shaft is 6mm and the other is 5mm with threads. Can I attach things on either side, or just the threaded part?

I read in the ESC page I linked that you can change the motor rotation by switching two wires. So I assume I can't make the motor go in both direction with the ESC?

Also, they say the ESC uses the throttle signal to control the speed of the motor. The wire looks like the wire used in radio-control and servos. So I assume it work with PWM? If this is correct, I could use a AtMega 168 chip to send PWM to control the motor speed?

Thanks for your help

Sam

Adrenalynn
07-23-2009, 11:40 AM
Too many "assumes" in that post. I wouldn't make any assumptions, I'd be finding a datasheet, calling the manufacturer for a datasheet, or finding some other product.