So, the 50:1 (7.2v) motors are better than the 30:1 motors, but they are not good enough. I have code that attempts to run the bot in a square by driving forward 10 inches, then doing a 90 degree rotate in place, then repeating to get back to where it started.
I'm using the encoders available from LynxMotion. They are 400 counts/state transitions per motor revolution, which is 1/50th of a wheel revolution, so 20000 counts per wheel revolution. Should be pretty accurate if there is no slipping.
The slowest speed I can reliably run the motors at is 20% duty cycle pwm. Which is about 30-35RPM of the wheel (~1/2 rotation per second). Currently, when it does the square, it's always off by a couple inches when it gets back to near the start point. It seems that the 90 degree rotates are the main problem. Sometimes it appears to turn the right amount, but often it's at least a smidge off and sometimes it's as much as 5 degrees off. Since I am using the encoders, it's got to be that wheels slipping, right? Maybe if I add some weight? My test surface is a wood tabletop. It's not coated/painted or polished, but it's reasonably smooth. The wheels are still the stock wheels. I'll try adding some weight and see if that helps. If I can get it to do the square reliably then I'll share a video and move on to longer distances and more turns.
I've got those alternate wheels on order (two different hardnesses), and I went ahead and ordered some of the pololu motors. Some 75:1 6V motors ( http://www.pololu.com/catalog/product/2275 ) and some 100:1 12V motors ( http://www.pololu.com/catalog/product/1446 ). They both come with encoders and they look decent enough, we'll see if they suck or not. They aren't the little Smile ones that you find on most of pololu's little bots.
If you look at the picture below you can see the original wheels I had chosen (and preferred) for the Stingray and this is where I got my motor ratios from. Unfortunately changing the wheel size requires a metal change, so we'll likely handle this in the motors. It will be interesting to hear what Roy has to say about the 50:1 motors, since they are what was used in the picture below (with the smaller wheels though).
Oh, it's a little hard to see, but the two motor mount plates behind the robots are fitted with heavy duty 12V motors with built-in encoders. These motors were heavy, fast, had lots of torque and of course had built in encoders. They were also pricey and you needed a 12V supply to run them. I can lookup information on these if anyone is interested.
I'm funding it myself. I want my bot to be good. Plus, I can use the other motors on some other projects/bots.
I'm single and have a decent job, so I can afford it to spend a little extra on my hobby.
Besides motor gearing and overshoot, I think that the Stingray's rear omniwheel is part of the repeatability problem when dead reckoning. It looks great, but when it moves sideways, there's a fair amount of friction in those tiny edge rollers. Maybe not the end of the world on a smooth floor or tabletop, but from my experience, a big dumb rubber caster works best in the real world when there are bumps and rugs along the way.
I think I can overcome the drag of the tail wheel for the most part with more weight over the drive wheels. It's all pretty lightweight right now. I'm using the Lion power pack to power it.
I don't need it to be super accurate, but it would be nice if it could do a small square and only be off a small amount (1/2 or 1/4 inch).
After some thought about this I decided to go through my parts boxes and get the Stingray out(I had forgotten about some of the parts I purchased over the summer!), I already had the 3 7/8" orange wheels, and I also have the Polu 50:1 12v dc motors http://www.pololu.com/catalog/product/1444 . I'm wondering if using the 50:1 motors with the smaller wheels would be any better? I have to find the code I was using when I last worked on this, it's been a while.
@Chris, with the smaller wheels, I assume you are talking about needing to change the height on the tailwheel? I'm thinking about either making slots or even mutiple holes for different sizes of drive wheels. (Maybe something that could be done on future versions of the Stingray?)
So far at least the Polu motors bolt right up to the mounts, i'm in the process of replacing the MSR1 with another prop board and a separate motor controller(probably a 10amp controller, the stall current on each of these motors are 5amps). If I can help, let me know.
Yes, when the wheel size was increased the tail wheel had to be lowered. Going back would mean moving the rear axle would need to move upward some or the Stingray will nose down.
As for Roy funding his changes, we're open to feedback, but if there's some hardware we have in stock which will help Roy he can contact me directly and I will see what I can do to help him out. He knows that.
I tried unsuccessfully to obtain some from the manufacturer. If you have some 100:1 samples, send me some, and I'll put some encoders on a Stingray and show you some maneuvers.
@Chris: My offer still stands... Roy need not go it alone!
Thanks Chris, today so far since I didn't want to lose the ground clearance I ended up using a pair of Polu motor mounts and drilled out some holes to mount them to, and installed the smaller wheels. I just have to get some other hardware out and setup before I can test.
Well the good news is that the motor/wheel combination at least works on the Stingray( the 30:1 12vdc Polu motors and smaller wheels). The bad news is that motors I got from Polu are pretty cheap, one of the encoders basically started to fall apart and there is considerable slop in the motor shafts. To makes things more difficult my hard drive on the pc I program with went south, I am expecting a replacement early next week, so I am at a stand still until I get that done....
erco, the motors used on the Stingray are compatible with many similar motors sold by various vendors. Trust me when I say, 100:1 would be way too slow...probably about BOE-Bot speed. 50:1 is what the original spec was and these motors had good torque and speed. The current motors are 30:1.
I received my Pololu Order Wednesday. The motors seem decent enough to me. The don't have any more "play" in the shaft than the other motors I have. The order also included some wheels to try. I should get the Tossen Robotics order today with the wheels that are similar to the stock ones.
This weekend will include several tests with different wheels and motors.
zappman, those motors look decent enough, but I don't think 1:52 is good enough.
I have used these motors for a few years now and I am quite happy with them. I especially like the integrated wiring harness which works nicely with the Serializer controller board. I initially used the motors on a relatively small robot seen here:
I have been using 8.4V Tenergy NiMH battery packs to power the motors and, as I mentioned above, the RoboticsConnection Serializer controller that includes dual 4-amp H-bridges.
It would be great to see the Stingray kit sold without motors since so many of us already have motors or have a specific preference for a particular model and specs.
So, the 50:1 (7.2v) motors are better than the 30:1 motors, but they are not good enough. I have code that attempts to run the bot in a square by driving forward 10 inches, then doing a 90 degree rotate in place, then repeating to get back to where it started.
I'm using the encoders available from LynxMotion. They are 400 counts/state transitions per motor revolution, which is 1/50th of a wheel revolution, so 20000 counts per wheel revolution. Should be pretty accurate if there is no slipping.
The slowest speed I can reliably run the motors at is 20% duty cycle pwm. Which is about 30-35RPM of the wheel (~1/2 rotation per second). Currently, when it does the square, it's always off by a couple inches when it gets back to near the start point. It seems that the 90 degree rotates are the main problem. Sometimes it appears to turn the right amount, but often it's at least a smidge off and sometimes it's as much as 5 degrees off. Since I am using the encoders, it's got to be that wheels slipping, right? Maybe if I add some weight? My test surface is a wood tabletop. It's not coated/painted or polished, but it's reasonably smooth. The wheels are still the stock wheels. I'll try adding some weight and see if that helps. If I can get it to do the square reliably then I'll share a video and move on to longer distances and more turns.
I've got those alternate wheels on order (two different hardnesses), and I went ahead and ordered some of the pololu motors. Some 75:1 6V motors ( http://www.pololu.com/catalog/product/2275 ) and some 100:1 12V motors ( http://www.pololu.com/catalog/product/1446 ). They both come with encoders and they look decent enough, we'll see if they suck or not. They aren't the little Smile ones that you find on most of pololu's little bots.
Motor upgrade tip for anyone going with the Lynxmotion 50:1, 7.2 v motor - be careful using the mounting screws supplied with the original Stingray motors. The screw length is fine for two of the threaded holes. For the third hole, the length will cause interference with a gear in the gear case. I discovered this after wondering why the motor would not run, and taking apart the gear case. One gear was rubbing on the screw. The motor worked fine after grinding about 1/16 inch off of the screw.
The 50:1 slows the robot to a nice speed at 50% PWM. I'll post the meters/second after I find my metric tape and stopwatch.
Something that may be of interest to some of you are these new 12V Gear Motors that Parallax now carries. With a nominal voltage of 6VDC these should provide a better range of voltage for control as well as a 30:1 gear ration resulting in 200RPM output at full speed.
Something that may be of interest to some of you are these new 12V Gear Motors that Parallax now carries. With a nominal voltage of 6VDC these should provide a better range of voltage for control as well as a 30:1 gear ration resulting in 200RPM output at full speed.
What about an exposed armature shaft for encoders? The one in the picture doesn't appear to have that feature. If not then we may not be able to swap out the Stingray motors for those of us using encoders.
Something that may be of interest to some of you are these new 12V Gear Motors that Parallax now carries. With a nominal voltage of 6VDC these should provide a better range of voltage for control as well as a 30:1 gear ration resulting in 200RPM output at full speed.
This powerful 12 V DC Geared Motor with 200 RPM output could be the muscle in your next robot. Manufactured by Hsiang Neng these motors are solid and a direct replacement for the 7.2 V motors used in our other compatible kits. Features:
Gear ratio doesn't mean anything without knowing the RPM and Torque of the motor itself. While these may have the same gear ratio on the output shaft, the 7.2V motors produced 300 RPM whereas the 12V motors only provide 200 RPM. That is a big difference and something that you should take into account.
What about an exposed armature shaft for encoders? The one in the picture doesn't appear to have that feature. If not then we may not be able to swap out the Stingray motors for those of us using encoders.
Robert
Robert, yeah these do not have the motor shaft exposed, so they would not be compatible with the U.S. Digital Encoders.
It's certainly true that this is a different motor winding for higher voltage and lower current: more windings of finer wire. Yielding slower no-load speed, although you haven't listed stall torques for comparison. Nonetheless it's likely quite similar to the stock 30:1 motor in terms of coasting and dynamic braking, which we know it needs more of. We won't know until you put them in a Stingray for a side-by-side comparison to a stock Stingray on Youtube!
FWIW, why not offer 100:1 motors as well and let people choose for themselves?
Comments
I'm using the encoders available from LynxMotion. They are 400 counts/state transitions per motor revolution, which is 1/50th of a wheel revolution, so 20000 counts per wheel revolution. Should be pretty accurate if there is no slipping.
The slowest speed I can reliably run the motors at is 20% duty cycle pwm. Which is about 30-35RPM of the wheel (~1/2 rotation per second). Currently, when it does the square, it's always off by a couple inches when it gets back to near the start point. It seems that the 90 degree rotates are the main problem. Sometimes it appears to turn the right amount, but often it's at least a smidge off and sometimes it's as much as 5 degrees off. Since I am using the encoders, it's got to be that wheels slipping, right? Maybe if I add some weight? My test surface is a wood tabletop. It's not coated/painted or polished, but it's reasonably smooth. The wheels are still the stock wheels. I'll try adding some weight and see if that helps. If I can get it to do the square reliably then I'll share a video and move on to longer distances and more turns.
I've got those alternate wheels on order (two different hardnesses), and I went ahead and ordered some of the pololu motors. Some 75:1 6V motors ( http://www.pololu.com/catalog/product/2275 ) and some 100:1 12V motors ( http://www.pololu.com/catalog/product/1446 ). They both come with encoders and they look decent enough, we'll see if they suck or not. They aren't the little Smile ones that you find on most of pololu's little bots.
Oh, it's a little hard to see, but the two motor mount plates behind the robots are fitted with heavy duty 12V motors with built-in encoders. These motors were heavy, fast, had lots of torque and of course had built in encoders. They were also pricey and you needed a 12V supply to run them. I can lookup information on these if anyone is interested.
Just outta the goodness of your heart, Roy?
I'm single and have a decent job, so I can afford it to spend a little extra on my hobby.
Besides motor gearing and overshoot, I think that the Stingray's rear omniwheel is part of the repeatability problem when dead reckoning. It looks great, but when it moves sideways, there's a fair amount of friction in those tiny edge rollers. Maybe not the end of the world on a smooth floor or tabletop, but from my experience, a big dumb rubber caster works best in the real world when there are bumps and rugs along the way.
-Phil
I don't need it to be super accurate, but it would be nice if it could do a small square and only be off a small amount (1/2 or 1/4 inch).
After some thought about this I decided to go through my parts boxes and get the Stingray out(I had forgotten about some of the parts I purchased over the summer!), I already had the 3 7/8" orange wheels, and I also have the Polu 50:1 12v dc motors http://www.pololu.com/catalog/product/1444 . I'm wondering if using the 50:1 motors with the smaller wheels would be any better? I have to find the code I was using when I last worked on this, it's been a while.
@Chris, with the smaller wheels, I assume you are talking about needing to change the height on the tailwheel? I'm thinking about either making slots or even mutiple holes for different sizes of drive wheels. (Maybe something that could be done on future versions of the Stingray?)
So far at least the Polu motors bolt right up to the mounts, i'm in the process of replacing the MSR1 with another prop board and a separate motor controller(probably a 10amp controller, the stall current on each of these motors are 5amps). If I can help, let me know.
As for Roy funding his changes, we're open to feedback, but if there's some hardware we have in stock which will help Roy he can contact me directly and I will see what I can do to help him out. He knows that.
@Chris: My offer still stands... Roy need not go it alone!
What does everyone think about these motors?
Key Specs: 7.2Vdc, 1:52, 160 RPM, 6mm Shaft Diameter, Built in quadrature encoders
http://www.active-robots.com/shop-by-brand/robotics-connection/7-2v-100oz-in-dc-gearhead-motor.html ?
DC Gearhead Motor:
- Rated Input Voltage 7.2Vdc
- No Load Speed 160 RPM
- 1:52 Gear Reduction
- Stall Torque: 100 oz-in (7.2 kg-cm)
- Drive Shaft End Hole Thread Size: 2.5 x 0.45 x 8mm
- No Load Current Draw: 300mA
- Max Current Draw (Shaft Lockup): 4.0A
Built in Quadrature Encoder:This weekend will include several tests with different wheels and motors.
zappman, those motors look decent enough, but I don't think 1:52 is good enough.
http://www.pirobot.org/videos/0010
and here
http://www.pirobot.org/videos/0008
I then used them on a larger robot but used an additional 36/60 external gear reduction to get more torque:
http://www.youtube.com/watch?v=6oDrO5EKU5w
I have been using 8.4V Tenergy NiMH battery packs to power the motors and, as I mentioned above, the RoboticsConnection Serializer controller that includes dual 4-amp H-bridges.
It would be great to see the Stingray kit sold without motors since so many of us already have motors or have a specific preference for a particular model and specs.
--patrick
Roy,
How are your motors doing?
I am thinking about getting a pair of Pollolu #1447 131:1 Metal Gearmotor 37Dx57L mm with 64 CPR Encoder 12 Volt motors for use on my Tracked Humanoid Torso Robot
seen at http://www.savagecircuits.com/forums/showthread.php?993-ZappBot-3-0-Tracked-Humanoid-Torso-Robot. I am trying to pick between these 131:1 motors and the 100:1 motors you have.
The motors that came with the Lynxmotion Robot base, that I want to replace are these 12 Volt Gear Head Motor - 12vdc 50:1 120rpm (6mm shaft) motors.
I want to replace the motors, so the robot will have more torque / load carrying capability and to add the encoders.
Thanks,
zappman
Your cat seems just about totally disinterested
-Matt
The 50:1 slows the robot to a nice speed at 50% PWM. I'll post the meters/second after I find my metric tape and stopwatch.
Sigh...
What about an exposed armature shaft for encoders? The one in the picture doesn't appear to have that feature. If not then we may not be able to swap out the Stingray motors for those of us using encoders.
Robert
There is a typo on the data sheet http://www.parallax.com/StoreSearchResults/tabid/768/txtSearch/motor/List/0/SortField/4/ProductID/788/Default.aspx. I am sure it should not say 44.5 V, see the quote below:
Once I receive and install the motors, I am going to try running them at 9.6V.
Erco,
Gear ratio doesn't mean anything without knowing the RPM and Torque of the motor itself. While these may have the same gear ratio on the output shaft, the 7.2V motors produced 300 RPM whereas the 12V motors only provide 200 RPM. That is a big difference and something that you should take into account.
Robert, yeah these do not have the motor shaft exposed, so they would not be compatible with the U.S. Digital Encoders.
FWIW, why not offer 100:1 motors as well and let people choose for themselves?
I will try to get a side-by-side comparison video up this week and link it in this thread. I should have done that before the motors went up.