Recent Posts

Pages: 1 2 3 4 5 6 ... 10
31
I designed a multistage belt drive a little over a year ago. I used GT2 belts for their anti backlash properties. In hindsight it probably wasn't the best approach to the problem. My opinion is that by the time you add tensioners to the design it creates too many components to be ideal compared to other technologies. When you have multiple stages it is difficult to tension the belts by moving one of the shafts which means tensioning using an idler pulley.

Fair enough.

Harmonic drives are an awesome mechanism; low backlash, high gear ratios, and low weight. they are available  however i think they are still covered by patent so the price is high and the sources limited. I used some steppers with harmonic drives before i think they ran about 1.5K each with encoder. also i think 1:32 is actually a bit low for harmonic drives.

Yup. Freakin patents, man. I see whole servos going on ebay for like $150, but it's all piecemeal. Ratios are weird, no idea how to control them, etc.

if you really need low backlash I recommend looking into cycloidal drives. they are supposed to have low backlash and high reduction ratios. however i have not found a good source for them, i did run across an industrial supplier but i'm sure price is prohibitive there. however they may be a little easier to fabricate than a harmonic drive.

Ugh. Ok. I'll take a look.

I don't see why you need low backlash for a skateboard but to be fair i dont ride one. if your system really only experiences load/torque in one direction this isn't really an issue. if you get rid of this requirement it opens up your drivetrain options considerably.

Hey man, I'm not on the robot forum because I'm building skateboards. The skateboard was just my inspiration. It's pretty impressive watching a skateboard propel a 200lb man up a 25% grade. I just saw that and thought, man, that would be an awesome actuator. All these skateboards come with a belt on each motor and a 1:2 reduction ratio. I just think I probably need more, unless my math is wrong.
32

Haven't made them, but have used them.

Not so sure about the no backlash. You will need a tensioner (or the weight of the motor on a pivot) if it isn't a cogged belt and that has to have some give.

[...]

Belt drive is fairly efficient so I see no real problems with adding an idler to make your reduction in two steps.

As far as calculating torque requirements, remember that torque for DC motors is maximum at stall or zero RPM. So, unless you want to  run very very very slow you need more torque than just the load. You'll get the most power at half the rated torque and the most efficiency at somewhat less than that.

Jeff

Good info, thanks. Yes, these are cogged belts. Here's the parts list so far, including links to the belt and pulley parts: https://docs.google.com/spreadsheets/d/1k9Cnwwm4x43ejwWfrssY_DFm7a6TMI1JPrzzTkUxPp4/edit?usp=sharing
33
if you want to go into industry i would suggest the mechanical degree. I think this would provide more job opportunities where a robotics degree might be more restrictive. If you know that you only want to the type of robotics work that aligns with the subject matter you referenced then by all means go for the robotics degree. however robotics is a wide field and requires all sorts of engineers. also the industry is not that big so robotics jobs are limited compared to other industries/mechanical engineering. If you want the academic life and to do research then the robotics degree might make more sense.

Disclaimer: these are just my opinions, i am a recent masters of mechanical engineering graduate myself



 
34
Mechanics and Construction / Re: new here... howdy. Also, multi-stage belt drives?
« Last post by bdeuell on August 26, 2015, 04:38:41 PM »
I designed a multistage belt drive a little over a year ago. I used GT2 belts for their anti backlash properties. In hindsight it probably wasn't the best approach to the problem. My opinion is that by the time you add tensioners to the design it creates too many components to be ideal compared to other technologies. When you have multiple stages it is difficult to tension the belts by moving one of the shafts which means tensioning using an idler pulley.

Harmonic drives are an awesome mechanism; low backlash, high gear ratios, and low weight. they are available  however i think they are still covered by patent so the price is high and the sources limited. I used some steppers with harmonic drives before i think they ran about 1.5K each with encoder. also i think 1:32 is actually a bit low for harmonic drives.

if you really need low backlash I recommend looking into cycloidal drives. they are supposed to have low backlash and high reduction ratios. however i have not found a good source for them, i did run across an industrial supplier but i'm sure price is prohibitive there. however they may be a little easier to fabricate than a harmonic drive.

Also there are many low backlash planetary gearboxes used in industry. However precision is key to these gearboxes. for the lowest backlash I believe they actually measure components (using a CMM) find the best fitting ones and then assemble them. even after that they might not get it right in which case they put that one aside build a new one.

I don't see why you need low backlash for a skateboard but to be fair i dont ride one. if your system really only experiences load/torque in one direction this isn't really an issue. if you get rid of this requirement it opens up your drivetrain options considerably.
35
Mechanics and Construction / Re: new here... howdy. Also, multi-stage belt drives?
« Last post by mklrobo on August 26, 2015, 03:09:30 PM »
 ;D Hello!
Your request;
 Has anyone ever make a multi-stage belt drive? What are the advantages and disadvantages? Anyone have tips for construction? I can't find much information on line. Seems like nobody ever makes more than a single stage belt drive. Why is that?
The only thing I can think of at this time, is a belt drive called an autodrive belt system.
This is basically 2 triangular cones side by side, one tip is near the wide end of the other cone, and
visa - versa. The belt is linking the two cones. The purpose is to provide a gear ratio from one cone to
the other; The width of the cone(s) ratio from one to the other. I do not know if that is what you are
looking for?. An academic problem example can be found in the REA's Problem Solver's book, Mechanics or Statics.    Good Luck!  ;D
36
Mechanics and Construction / Re: new here... howdy. Also, multi-stage belt drives?
« Last post by cyberjeff on August 26, 2015, 02:59:30 PM »
Howdy,

I'm new here. I'm a software engineer by trade. I've dabbled in basic electronics theory. I've got a small woodworking shop because I'm interested in material properties (it just kinda grew out of the need to know more about wood). I've got a crappy 3d printer for prototyping, and a website I made because it seemed like something cool to do: http://www.createthis.com

I aspire to get more metalworking tools in the near future, like a cnc mill, or maybe just a regular mill. I currently have a big floor mounted drill press and a TIG welder, neither of which I'm particularly good with.

A couple of years ago, I was pretty interested in servos, and wanted a larger format "smart" servo, so I made this: http://www.diyservo.com

I learned I don't want backlash in my servo, and therefore I couldn't use planetary gears, so I abandoned the project.

I've since picked up an interest in quad copters, which led to brushless motors, which led to me picking up a boosted board ( if you don't know what that is: https://youtu.be/tyNJYbwBWnQ ), which led to me ordering a bunch of parts for a custom e-skateboard drivetrain to use as a large format actuator.

I just finished my torque calculation this morning. Apparently my actuator will make about 292 oz-in torque: http://www.rcgroups.com/forums/showthread.php?t=2486645

But I really want about 9600 oz-in. (50 ft-lbs), so I'm going to need a 1:32 reduction ratio. I really want to use a harmonic drive gearbox, but I can't find them for sale anywhere other than ebay, and I can't really find what I want on ebay, so I'm investigating alternatives. ( Finding someone willing to sell me various harmonic drive gearboxes might solve my whole problem, I don't know )

I'm currently toying with the idea of using belts for this, since they're zero backlash. So... getting to the point finally... anyone ever make a multi-stage belt drive?



Haven't made them, but have used them.

Not so sure about the no backlash. You will need a tensioner (or the weight of the motor on a pivot) if it isn't a cogged belt and that has to have some give.

Your torque requirements are probably above what you can get here:

https://www.servocity.com/html/smooth_hub_pulleys.html#.Vd4ijflViko

And garden and automotive stuff is probably overkill:

http://www.northerntool.com/shop/tools/category_lawn-garden+lawn-mowers+belts-pulleys+4294903686

Belt drive is fairly efficient so I see no real problems with adding an idler to make your reduction in two steps.

As far as calculating torque requirements, remember that torque for DC motors is maximum at stall or zero RPM. So, unless you want to  run very very very slow you need more torque than just the load. You'll get the most power at half the rated torque and the most efficiency at somewhat less than that.

Jeff

Quote

 What are the advantages and disadvantages? Anyone have tips for construction? I can't find much information on line. Seems like nobody ever makes more than a single stage belt drive. Why is that?

Thanks,

--
Jesse
CreateThis.com
37
Misc / Re: Can hobby roboticists save A Country's industry?
« Last post by mklrobo on August 26, 2015, 02:56:49 PM »
 8) Awesome input!
I am exploring my own leg system, and it is quite energy demanding.  :'(
I was hoping some robotic tech could create a humanoid robot (like Chappie) to
do the manual labor.(Farming, small business labor)  ;)
38
Mechanics and Construction / new here... howdy. Also, multi-stage belt drives?
« Last post by createthis on August 26, 2015, 11:47:14 AM »
Howdy,

I'm new here. I'm a software engineer by trade. I've dabbled in basic electronics theory. I've got a small woodworking shop because I'm interested in material properties (it just kinda grew out of the need to know more about wood). I've got a crappy 3d printer for prototyping, and a website I made because it seemed like something cool to do: http://www.createthis.com

I aspire to get more metalworking tools in the near future, like a cnc mill, or maybe just a regular mill. I currently have a big floor mounted drill press and a TIG welder, neither of which I'm particularly good with.

A couple of years ago, I was pretty interested in servos, and wanted a larger format "smart" servo, so I made this: http://www.diyservo.com

I learned I don't want backlash in my servo, and therefore I couldn't use planetary gears, so I abandoned the project.

I've since picked up an interest in quad copters, which led to brushless motors, which led to me picking up a boosted board ( if you don't know what that is: https://youtu.be/tyNJYbwBWnQ ), which led to me ordering a bunch of parts for a custom e-skateboard drivetrain to use as a large format actuator.

I just finished my torque calculation this morning. Apparently my actuator will make about 292 oz-in torque: http://www.rcgroups.com/forums/showthread.php?t=2486645

But I really want about 9600 oz-in. (50 ft-lbs), so I'm going to need a 1:32 reduction ratio. I really want to use a harmonic drive gearbox, but I can't find them for sale anywhere other than ebay, and I can't really find what I want on ebay, so I'm investigating alternatives. ( Finding someone willing to sell me various harmonic drive gearboxes might solve my whole problem, I don't know )

I'm currently toying with the idea of using belts for this, since they're zero backlash. So... getting to the point finally... anyone ever make a multi-stage belt drive? What are the advantages and disadvantages? Anyone have tips for construction? I can't find much information on line. Seems like nobody ever makes more than a single stage belt drive. Why is that?

Thanks,

--
Jesse
CreateThis.com
39
Misc / Re: Can hobby roboticists save A Country's industry?
« Last post by cyberjeff on August 26, 2015, 10:06:50 AM »
;D Hello!
I was listening to Glenn Beck, and he brought up the cost of labor in industry.
Some countries reduce their labor cost to almost nothing, (bad for the individual)
and make them very competitive to make products.
What if hobby roboticists started an open source robot to be used in industry for manual
labor? Could it be done? This would reduce the labor cost to the cost/maintenance of the robot.
With the new batteries (Powerplus - 13.2 ampere hour) and the CPU power available, it may be possible! :o
this would create jobs for techs, provide cheap manual labor, and help build up the country that is using the Robots. What do you think? ;)

Manufacturing is full of industrial robots. These are powerful machines and although expensive, they are less  than the cost of human labor.

I live in the US, here the economy is driven more by services than manufacturing. And the service business that is growing fastest is healthcare (note Japanese heavy involvement in robotics as well as an aging population). Where I think hobbyist have an opening is not in the mechanics but in the software end and the development of artificial intelligence. Building software is cheap, building rugged strong hardware is very expensive.

At any rate doing something that has already been done does not advance anything. What more can be done with wheels that isn't already done? Which is why I have been studying legged robots and what I have discovered is a fertile but largely untended field. The designs out there are deeply flawed, for example, the vast majority of the quadruped robots are almost square in shape and that is an architecture almost doomed to fail. Take a square card table and fold up one leg and it will fall. This why there are no squarish animals, except ones like tortoises that spend their lives already on the ground.

So, where hobbyist can make a difference is in exploring new ground, developing new ideas that may not have a short term return on investment. I don't see how any of us are going to make a better industrial robot. But it is possible that we may find techniques and architecture that will break open new fields.
40
Misc / Can hobby roboticists save A Country's industry?
« Last post by mklrobo on August 26, 2015, 08:59:25 AM »
 ;D Hello!
I was listening to Glenn Beck, and he brought up the cost of labor in industry.
Some countries reduce their labor cost to almost nothing, (bad for the individual)
and make them very competitive to make products.
What if hobby roboticists started an open source robot to be used in industry for manual
labor? Could it be done? This would reduce the labor cost to the cost/maintenance of the robot.
With the new batteries (Powerplus - 13.2 ampere hour) and the CPU power available, it may be possible! :o
this would create jobs for techs, provide cheap manual labor, and help build up the country that is using the Robots. What do you think? ;)
Pages: 1 2 3 4 5 6 ... 10

Get Your Ad Here