The sprockets finally arrived. I am not sure why anyone would want the smaller sprockets with this size track. It is hard to tell when you order this stuff, so this might help someone.
Here is what one set of track looks like on their two different sized sprockets -
This is what a 31" track with the 9 tooth gets you. 31" track is 1.5 21" track lengths, if you make two you will have one segment left over. So you're making two tracks out of three sets of track.
Look at all of the motor clearance you get. It's perfect for the larger diameter gear motors. Now the motor can be inside of the chassis, which is what my plan was from the beginning. Measure twice is the moral of that story.
Here is a picture of the larger edition. This size is pushing the limit of the ePVC and I added a small brace. This is just to get an idea of the largest robot I can build with the parts I have. I don't think it will be this big and will think about it over the upcoming holidays.
That battery is 2200mah to give you an idea of the size.
Those larger sprockets sure look like a better match for those treads.
I like the aluminum L channel. I've been wondering about adding some to my threaded vehicle design.
If the ePVC isn't going to be strong enough you might want to consider the plywood used in model airplanes. It's not as easy to cut as the ePVC but (as I'm sure you already know) it's a lot stronger.
Do you know if you'll raise the center section of the bot to increase the ground clearance? Your last photo sure looks like it could get high centered rather easily. I'm sure you're thinking about these things.
I'm interested in how different the bot behaves with the larger sprockets. It sure looks better.
I'd imagine you've traded some torque for increased speed.
Thanks for the updates. It's fun to follow your progress.
The ePVC is temporary so I can play with different configurations. Once I'm happy with the design I'll finalize the drawings and have the parts made out of aluminum. So what I end up with might not work well as just ePVC.
As far as ground clearance goes, there is an overlying issue there. First off, there is no way this track will ever work in mud. I'm not sure about sand, but I kinda doubt reliability in sand. So the question is are these tracks outdoor worthy, because if they are not, then ground clearance doesn't matter. This is another thing I sort of overlooked when making the initial purchase.
The cool thing is I will build the chassis to accommodate wheels and tracks. As far as the tracked robot goes, I think it will be staying inside. All of this is pending some real testing. I'm going to take it to the beach and do some sand trials.
In the meantime the larger sprockets did add noticeable speed. It now travels at a normal walking speed. I'm not sure you could notice less torque. Even the big design is unstoppable. Last night I was showing people how it can knock down the dog gate in the hallway then drive over it. It's violent, I hit the Christmas tree and almost knocked it over. I do love these motors.
I'm glad you're still with me on the project. It has been a lot of fun for me too and I hope it answers any questions about the tracks. I'll probably do an official write-up on them eventually. I don't know why Lynx doesn't offer up more resources on their site.
Oh yeah, I heard from Roboclaw that my controller was one that had a bad bunch of FETs, so I'm off the hook for smoking it
This thread has a lot of views... If you are an outsider coming here and find any information of use, do yourself a favor and try the Propeller microcontroller.
Back to the tracks. I spent some time running the larger and smaller tracks outdoors in semi-harsh to extreme conditions. I put some tubes over the motors and encoders and put all of the electronics in a tupperware container. I was going to fab some sheet metal covers to help prevent debris and objects from getting into the tracks. Instead I opted for worst case scenario first. I would especially like to see what is going on inside of the track assembly during this test as well.
This is why I need to stop juggling 50 things at once. My attention is too divided when I'm in front of a computer, especially in the middle of the workday. When it comes to building robots, I want to have fun, not do what I do for a living and turn it into a huge production. I saw the tracks, didn't do a lot of research, couple clicks later and here we are. If someone asked me are these outdoor tracks, I'm still not sure which way I would lean.
There are plenty of outdoor environments where these tracks can be used, and they work great. There are also some places where these tracks will fail to the point of having to be removed. Remember to keep in mind that my design is very basic. It is possible to design a chassis that would slow the accumulation of foreign matter, maybe even prevent it entirely. In that case I'm sure that these tracks could qualify as outdoor tracks.
I think wheels (or differently designed tracks) would be a better choice than these tracks for outdoor use in anything mission critical. If you look at the tracks on a bomb disposal robot you'll see they're a lot different.
Without modifying the tracks (more on that later) you'd be stuck working with very tight tolerances and maybe some kind of sweeper material to close the gaps. If you didn't, a stick or a pebble could easily be a wrench in the gears. Sand was not as bad as I thought, it leaves an awesome track trail pattern in the wet sand. In dry silty sand doing a turn in place didn't pile sand into the track like I thought it would, but there was some accumulation. I purposely setup the robot to fail in this regard and it held up quite well, which to me means there is a lot of potential. Unfortunately, doing more chassis designing for this kind of thing, with my skill level, could potentially take a lifetime.
So, unfortunately I will not be building an outdoor tracked robot. I'm selling all of my electronics, parts, test equipment, gear, and soldering iron on eBay and taking up needle point. Actually no, that won't work because I'd need all that stuff back to build an automated needle point system. There goes that idea.
Okay, okay, sooooo the new plan is to see if I can modify the tracks without sacrificing integrity. I'm thinking that the sprocket tooth could be hollowed out, or the holes the teeth ride in can be drilled out. I would much rather modify the sprockets than the tracks. In fact I might just make some "better sprockets" for these tracks.
I spent some time with the tracks, and sacrificed a small section to see if they could be modified for reliable outdoor use in harsh environments, mud and sand mostly. It would be easy but time consuming to modify the tracks for full outdoor use.
So I cut away the "problem" part of the track so that the sprocket would "push" anything in the way outside of the track.
On the left is the unmodified track segment, on the right is the segment I used a cutting disc and needle nose pliers on.
One track segment cut like this does reduce its structural integrity. It is reduced in a way that when the track is assembled doesn't matter as much. Essentially it can flex easier, but I don't think it would matter when all of the segments are connected.
Now I really need to decide what I'm going to do with these tracks. My options are -
#1. Switch to a wheeled robot using the same exact everything.
#2. Keep the tracks the way they are, build an indoor/outdoor capable robot and plan on never driving though mud or sand.
#3. Modify them and they work great, or find out it was a bad idea and I should have just used them for an indoor robot.
I'm leaning toward #2. What do you think?
I would get much more use out of an indoor robot. The original outdoor robot deal with GoPro and NBC is taking too long and could be in 2015! I'm sorry this thread has turned into a long process involving just the tracks. I will get past this soon and start having some aluminum pieces machined.
I'm sorry this thread has turned into a long process involving just the tracks.
I'm personally very glad you documented the track issues. I've been very curious about these tracks and I've learned a lot from this thread.
What do the outside of the modified tread look like? One of the appeals of these tracks is how cool they look. I'm afraid your mod, while improving the tracks function, could severely degrade the aesthetics of the threads.
Thanks for taking the time to document this so well.
Thanks Duane! I'm hoping the rest of the community is with you on that. I'm not sure why this thread has so many views. I try to keep the keywords evened out, it must be the tracks though. Google indexes this forum so fast I started proofreading before I hit post.
Above is a picture of the exterior of the modded track. It does bend and twist a little more, but again once assembled this will probably not be a problem, definitely not for this size robot. A real issue I can see is cutting it might weaken the glue that hold the rubber to the plastic.
Something to keep in mind is that these tracks could probably be used with lawn and garden equipment, 200+lbs robots, etc. They are super heavy duty and excuses, excuses that is why I assumed outdoor use.
Above is the way the track pieces interconnect. There is a little nylon pin and two plastic reloadable "rivets". As you can see, the way they attach is very strong and the mod shouldn't make them weaker by much.
I have decided that I will not modify them. Even with the mod, there is a big chance these tracks will stall, bind, etc with dirt introduced.
Duane I haven't been thinking much about aesthetics, I'm not sure you would notice the cuts from more than a couple feet away. I guess you could drill them or cut nicer, but either way I still don't see these as "reliable" for outside.
I have my indoor chassis almost finalized. I will send off the drawings this week. The ePVC is getting upset with me, although it is holding up VERY well, and I never use caps.
Now I really need to decide what I'm going to do with these tracks. My options are -
#1. Switch to a wheeled robot using the same exact everything.
#2. Keep the tracks the way they are, build an indoor/outdoor capable robot and plan on never driving though mud or sand.
#3. Modify them and they work great, or find out it was a bad idea and I should have just used them for an indoor robot.
I'm leaning toward #2. What do you think?
Thinking out loud - option #4
Suppose that you could integrate 2 wheel wells, perhaps the width of your crawler, one fore and one aft. Contained within each wheel well youd have an encapsulated foam filled cylinder that would fit between the tracks. The exterior surface of the cylinder would be slightly toothed to resemble that of an old steamboat paddlewheel. The cylinders would pivot such that they could be lowered some distance below the bottom track. Two servos one fore and one aft would lower these cylinders (wheels) and these wheels would be driven by the crawler sprockets. Now your tracks would ride above the mud, swamp, and sand and would be kept clean.
Then, youd have an indoor machine, an outdoor machine, a sand and mud machine, AND given enough buoyancy from the foam filled cylinders you could even traverse a river with the same machine! If youre interested in accuracy of distance traveled your encoders might work for the indoor adventures but you might have to switch over to GPS mode for swamp and lake adventures. Perhaps youd need to split the cylinders if you wanted steering actually, youd probably have to split each cylinder into two separate sections anyway due to the way the tracks are driven.
For convenience, some type of moisture sensor could sense swamp and lake conditions to automatically lower the cylinders, and Im sure with your creativity itd be a slam dunk for you to design a sensor to detect sandy conditions too.
Food for NON-SERIOUS thought At least now you know why I spent a few hours today helping a gal install some porcelain tile in her kitchen instead of spending those hours as an industrial design engineer.
Why so many views? We love and learn from (your) projects here. And, we cant let a little mud stand in the way of a cool project.
I was thinking about using something like you mention on the idlers to take them out of the equation. If it were just the drive sprocket I could put that in the middle on top because I have spares. Then at least gravity would help keep the track clean, which means sand would probably be fine.
I'm not sure about adding more mechanical actions, what you said would be awesome but I think that I will just make this an indoor robot with semi-outdoor capabilities.
I'm working on the final design with the larger sprockets today and will post more soon.
Thinking about a sensor for sandy conditions, that is interesting.
Here is the final design using the larger sprockets and all three sections of track. I'm having it fabricated so it will be a month before I get them back. I might take them to get powder coated after, not sure. Until then the ePVC is holding up well but I can't add much weight. I got a smaller 5v 3a UBEC for the 5v stuff. This should be large enough to mount lots of sensors to and carry a lot of batteries. I still have to decide if I'm using PWM and encoders or traditional RC style PWM and encoders. Either way this is pretty sweet because it's the first chassis I've had built out of something other than hand tools
The chassis has shipped! I will have it Thursday of next week.
I'm sticking with the PWM based motor controller, and will be reading the encoders with the Propeller. I will attempt this using the code Duane sent me, in SPIN and PASM.
One thing I realized was I got way ahead of myself. So I'm setting up a very clean environment with only one motor, and no robot. I made a motor stand out of ePVC, and put a little piece of red tape on the motor shaft as an indexer. My plan is to hit this hard until it is done, and all the other side projects will need to wait. I never got the job this project was intended for, but I also think it would be a much better indoor robot anyway.
Duane D, if you have any further work on the code I've been using please let me know. I have also set the radio controller aside for now, I do not think that I will add it back.
I re-tested the encoders and setup a mock environment similar to the power input on the robot. Right now I am looking at Duane's code, and also looking at two different objects, one for PWM motor control, and the other for quadrature encoders. Between the three I don't think there will be a problem, other than I should have started coding instead of procr... err waiting for the chassis.
The ePVC made it really easy to prototype. Everything lined up beautifully. The T6 aluminum is somewhat of a next step because it is easy to file, drill, etc. Here are some pics:
Thanks, I have it all assembled now and it is very smooth. I wish I had the means to build stuff like this myself. I could care less about 3D printers, I'd kill for a decent CNC mill though.
It's 0.125" 6061 T6 aluminum the total cost for all four pieces was $105, I asked for unfinished and was shocked to see it is somewhat polished. I could have doubled the order for only a couple bucks more but I'm still prototyping. It has a long way to go and things are going to change. Once everything is finalized I will give away the plans for free or sell low cost kits.
There are very few robots this size on the market that are affordable. I'd like to fill that gap however I can. Small robots like the BOE Bot are awesome until you want to put a small computer on it. The next affordable option for me was to build something large with wheelchair motors and that turned out to be far to big and could just barely fit through a doorway. This size should be enough to carry enough payload to put whatever you want on there.
I have only designed the track part, now I'm using L channel, sheet metal and more ePVC to develop a base. From the base I will use a large diameter PVC pipe to go up. That will have a "T" piece on top, the side of the T will be for arms, and the top will be for some kind of head. I'll post a drawing soon.
Another update. Looks like it's back to the PWM driver. Either that or I end up with a third MCU. I would love to use HB25's but they are only RC pulse. I was hoping to graduate to PWM at some point when I understood the code a little better.
Okay so back to the PWM driver... I used Duane's code and turned it into a child object for my main program and it works great. I think I'm finally grasping the SPIN structure and cogs.
Someone asked if I had XCTU on my iPhone. No, I do not. I'm using remote desktop to my computer and XBees. None of that matters, this is all for testing. I thought it would be easier to show what is going on in the video by using my phone, but the screen doesn't show up well so I may just use a screen recorder.
I have to take a step back again and fix the tracks. In the last video above you can see when the tracks go fast there is some kind of alignment issue. I'm not going any further until that is resolved. I need to cut down all of the standoffs the nylon idlers run on. I don't know why they are .12" longer than the wheel hub axle but I can tell that is adding to an alignment issue. I'd like to get that right to cut down on current consumption.
A standoff standoff. Adding to some of the confusion behind the Lynx tracks is the length of the standoffs. Armed with nothing but basic hand tools I'll need to shave all 10 of them down. I'll probably build a little block to allow me to do so accurately. Here are some pics of what is going on with it.
The red circle of doom is showing the gap. The idler needs this outside screw in place or there is slop. Originally I had that screw in place but it bends the frame and I still think it's a little off due to the bend.
Okay it's better, but still not quite where I want to be. Elevating the tracks from the ground is the worst case scenario, you'll notice more noise and alignment issues than with the tracks on the ground. So working in this area is great because it increases the small issues allowing us to make adjustments where needed.
[video=youtube_share;i77_O12RduU]
I think the next thing to do is reduce the friction. I need some kind of lubrication on the track pins and between the track segments, and I think I'll be a lot happier with the results.
White lithium grease comes to mind, but taking the tracks apart doesn't appeal to me as much as some kind of poly safe spray. Even if the poly safe spray lube was to wear off quicker, it would still be easier to reapply.
Are you sure you need to cut the standoffs? Is there something you're doing different than the way things are done in the Johnny 5 chassis to require the standoffs to be modified?
Do your bearings have flanges? If so what if you switched the flange from the outside to the inside? Would the width of two flanges make up the difference?
It just seems really strange the parts wouldn't fit together correctly.
That is where the confusion came into play. Having the flange on the inside increased the friction of the idler sprocket when the screws were tight. It also made it so that the bearing could pull through the chassis should one side come loose. With the flange on the outside the pulling force of the opposite screw holds it in place much better. Still not sure why the J5 chassis is like that.
It wasn't a problem to shave some length from the standoffs at all, I used a little jig I made and a flat file and did 5 at a time in a couple minutes.
First issue, the tracks, noise and alignment. I post on another forum (gasp) a video of my setup and have been reassured that everything is normal. It's still a little hard for me to swallow the current drain of these tracks, but as of now I'm going to stop complaining about it since I know that it is as good as it will get.
Second, the EMIC 2 is cool and everything but without the visual feedback, well it's just not the same. Everything was great on a single breadboard, I even used a level shifter which I still do not feel is necessary for a short strip of WS LEDs. The problem came down to a noisy 5v rail. So now I have two 5v rails. One switching and one linear. The linear drives the WS LED strip. The LEDs vary in intensity in this video but that is because of my code which is refreshing them when it doesn't need to. I could fix that but it doesn't make sense right now since it will run an entirely different routine, as well as be called upon from another piece of code.
You may notice the speech is delayed after the movement of the tracks, this is by design since I want the movement of the robot to be a higher priority than the speech feedback. I also have to mention using the EMIC 2 is so much better to debug code than the PST or XCTU given the proper environment.
Anyway, here it is, if you're feelin real crazy watch it in 1080:
Comments
Here is what one set of track looks like on their two different sized sprockets -
Lynxmotion 6 tooth 2.75" SPRK-01 (8" axle to axle) (interior height 1.75")
Lynxmotion 9 tooth 3.75" SPRK-02 (6.5" axle to axle) (interior height 2.75")
Look at all of the motor clearance you get. It's perfect for the larger diameter gear motors. Now the motor can be inside of the chassis, which is what my plan was from the beginning. Measure twice is the moral of that story.
That battery is 2200mah to give you an idea of the size.
I like the aluminum L channel. I've been wondering about adding some to my threaded vehicle design.
If the ePVC isn't going to be strong enough you might want to consider the plywood used in model airplanes. It's not as easy to cut as the ePVC but (as I'm sure you already know) it's a lot stronger.
Do you know if you'll raise the center section of the bot to increase the ground clearance? Your last photo sure looks like it could get high centered rather easily. I'm sure you're thinking about these things.
I'm interested in how different the bot behaves with the larger sprockets. It sure looks better.
I'd imagine you've traded some torque for increased speed.
Thanks for the updates. It's fun to follow your progress.
The ePVC is temporary so I can play with different configurations. Once I'm happy with the design I'll finalize the drawings and have the parts made out of aluminum. So what I end up with might not work well as just ePVC.
As far as ground clearance goes, there is an overlying issue there. First off, there is no way this track will ever work in mud. I'm not sure about sand, but I kinda doubt reliability in sand. So the question is are these tracks outdoor worthy, because if they are not, then ground clearance doesn't matter. This is another thing I sort of overlooked when making the initial purchase.
The cool thing is I will build the chassis to accommodate wheels and tracks. As far as the tracked robot goes, I think it will be staying inside. All of this is pending some real testing. I'm going to take it to the beach and do some sand trials.
In the meantime the larger sprockets did add noticeable speed. It now travels at a normal walking speed. I'm not sure you could notice less torque. Even the big design is unstoppable. Last night I was showing people how it can knock down the dog gate in the hallway then drive over it. It's violent, I hit the Christmas tree and almost knocked it over. I do love these motors.
I'm glad you're still with me on the project. It has been a lot of fun for me too and I hope it answers any questions about the tracks. I'll probably do an official write-up on them eventually. I don't know why Lynx doesn't offer up more resources on their site.
Oh yeah, I heard from Roboclaw that my controller was one that had a bad bunch of FETs, so I'm off the hook for smoking it
Back to the tracks. I spent some time running the larger and smaller tracks outdoors in semi-harsh to extreme conditions. I put some tubes over the motors and encoders and put all of the electronics in a tupperware container. I was going to fab some sheet metal covers to help prevent debris and objects from getting into the tracks. Instead I opted for worst case scenario first. I would especially like to see what is going on inside of the track assembly during this test as well.
This is why I need to stop juggling 50 things at once. My attention is too divided when I'm in front of a computer, especially in the middle of the workday. When it comes to building robots, I want to have fun, not do what I do for a living and turn it into a huge production. I saw the tracks, didn't do a lot of research, couple clicks later and here we are. If someone asked me are these outdoor tracks, I'm still not sure which way I would lean.
There are plenty of outdoor environments where these tracks can be used, and they work great. There are also some places where these tracks will fail to the point of having to be removed. Remember to keep in mind that my design is very basic. It is possible to design a chassis that would slow the accumulation of foreign matter, maybe even prevent it entirely. In that case I'm sure that these tracks could qualify as outdoor tracks.
I think wheels (or differently designed tracks) would be a better choice than these tracks for outdoor use in anything mission critical. If you look at the tracks on a bomb disposal robot you'll see they're a lot different.
Without modifying the tracks (more on that later) you'd be stuck working with very tight tolerances and maybe some kind of sweeper material to close the gaps. If you didn't, a stick or a pebble could easily be a wrench in the gears. Sand was not as bad as I thought, it leaves an awesome track trail pattern in the wet sand. In dry silty sand doing a turn in place didn't pile sand into the track like I thought it would, but there was some accumulation. I purposely setup the robot to fail in this regard and it held up quite well, which to me means there is a lot of potential. Unfortunately, doing more chassis designing for this kind of thing, with my skill level, could potentially take a lifetime.
So, unfortunately I will not be building an outdoor tracked robot. I'm selling all of my electronics, parts, test equipment, gear, and soldering iron on eBay and taking up needle point. Actually no, that won't work because I'd need all that stuff back to build an automated needle point system. There goes that idea.
Okay, okay, sooooo the new plan is to see if I can modify the tracks without sacrificing integrity. I'm thinking that the sprocket tooth could be hollowed out, or the holes the teeth ride in can be drilled out. I would much rather modify the sprockets than the tracks. In fact I might just make some "better sprockets" for these tracks.
So I cut away the "problem" part of the track so that the sprocket would "push" anything in the way outside of the track.
On the left is the unmodified track segment, on the right is the segment I used a cutting disc and needle nose pliers on.
One track segment cut like this does reduce its structural integrity. It is reduced in a way that when the track is assembled doesn't matter as much. Essentially it can flex easier, but I don't think it would matter when all of the segments are connected.
Now I really need to decide what I'm going to do with these tracks. My options are -
#1. Switch to a wheeled robot using the same exact everything.
#2. Keep the tracks the way they are, build an indoor/outdoor capable robot and plan on never driving though mud or sand.
#3. Modify them and they work great, or find out it was a bad idea and I should have just used them for an indoor robot.
I'm leaning toward #2. What do you think?
I would get much more use out of an indoor robot. The original outdoor robot deal with GoPro and NBC is taking too long and could be in 2015! I'm sorry this thread has turned into a long process involving just the tracks. I will get past this soon and start having some aluminum pieces machined.
I'm personally very glad you documented the track issues. I've been very curious about these tracks and I've learned a lot from this thread.
What do the outside of the modified tread look like? One of the appeals of these tracks is how cool they look. I'm afraid your mod, while improving the tracks function, could severely degrade the aesthetics of the threads.
Thanks for taking the time to document this so well.
Above is a picture of the exterior of the modded track. It does bend and twist a little more, but again once assembled this will probably not be a problem, definitely not for this size robot. A real issue I can see is cutting it might weaken the glue that hold the rubber to the plastic.
Something to keep in mind is that these tracks could probably be used with lawn and garden equipment, 200+lbs robots, etc. They are super heavy duty and excuses, excuses that is why I assumed outdoor use.
Above is the way the track pieces interconnect. There is a little nylon pin and two plastic reloadable "rivets". As you can see, the way they attach is very strong and the mod shouldn't make them weaker by much.
I have decided that I will not modify them. Even with the mod, there is a big chance these tracks will stall, bind, etc with dirt introduced.
Duane I haven't been thinking much about aesthetics, I'm not sure you would notice the cuts from more than a couple feet away. I guess you could drill them or cut nicer, but either way I still don't see these as "reliable" for outside.
I have my indoor chassis almost finalized. I will send off the drawings this week. The ePVC is getting upset with me, although it is holding up VERY well, and I never use caps.
Thinking out loud - option #4
Suppose that you could integrate 2 wheel wells, perhaps the width of your crawler, one fore and one aft. Contained within each wheel well youd have an encapsulated foam filled cylinder that would fit between the tracks. The exterior surface of the cylinder would be slightly toothed to resemble that of an old steamboat paddlewheel. The cylinders would pivot such that they could be lowered some distance below the bottom track. Two servos one fore and one aft would lower these cylinders (wheels) and these wheels would be driven by the crawler sprockets. Now your tracks would ride above the mud, swamp, and sand and would be kept clean.
Then, youd have an indoor machine, an outdoor machine, a sand and mud machine, AND given enough buoyancy from the foam filled cylinders you could even traverse a river with the same machine! If youre interested in accuracy of distance traveled your encoders might work for the indoor adventures but you might have to switch over to GPS mode for swamp and lake adventures. Perhaps youd need to split the cylinders if you wanted steering actually, youd probably have to split each cylinder into two separate sections anyway due to the way the tracks are driven.
For convenience, some type of moisture sensor could sense swamp and lake conditions to automatically lower the cylinders, and Im sure with your creativity itd be a slam dunk for you to design a sensor to detect sandy conditions too.
Food for NON-SERIOUS thought At least now you know why I spent a few hours today helping a gal install some porcelain tile in her kitchen instead of spending those hours as an industrial design engineer.
Why so many views? We love and learn from (your) projects here. And, we cant let a little mud stand in the way of a cool project.
I was thinking about using something like you mention on the idlers to take them out of the equation. If it were just the drive sprocket I could put that in the middle on top because I have spares. Then at least gravity would help keep the track clean, which means sand would probably be fine.
I'm not sure about adding more mechanical actions, what you said would be awesome but I think that I will just make this an indoor robot with semi-outdoor capabilities.
I'm working on the final design with the larger sprockets today and will post more soon.
Thinking about a sensor for sandy conditions, that is interesting.
Here is the final design using the larger sprockets and all three sections of track. I'm having it fabricated so it will be a month before I get them back. I might take them to get powder coated after, not sure. Until then the ePVC is holding up well but I can't add much weight. I got a smaller 5v 3a UBEC for the 5v stuff. This should be large enough to mount lots of sensors to and carry a lot of batteries. I still have to decide if I'm using PWM and encoders or traditional RC style PWM and encoders. Either way this is pretty sweet because it's the first chassis I've had built out of something other than hand tools
I'm sticking with the PWM based motor controller, and will be reading the encoders with the Propeller. I will attempt this using the code Duane sent me, in SPIN and PASM.
One thing I realized was I got way ahead of myself. So I'm setting up a very clean environment with only one motor, and no robot. I made a motor stand out of ePVC, and put a little piece of red tape on the motor shaft as an indexer. My plan is to hit this hard until it is done, and all the other side projects will need to wait. I never got the job this project was intended for, but I also think it would be a much better indoor robot anyway.
Duane D, if you have any further work on the code I've been using please let me know. I have also set the radio controller aside for now, I do not think that I will add it back.
I re-tested the encoders and setup a mock environment similar to the power input on the robot. Right now I am looking at Duane's code, and also looking at two different objects, one for PWM motor control, and the other for quadrature encoders. Between the three I don't think there will be a problem, other than I should have started coding instead of procr... err waiting for the chassis.
XBee, ADC, EMIC II, RoboClaw MCU
[video=youtube_share;3Svks4S8GS4]
The ePVC made it really easy to prototype. Everything lined up beautifully. The T6 aluminum is somewhat of a next step because it is easy to file, drill, etc. Here are some pics:
[IMG]https://dl.dropboxusercontent.com/u/79058769/pics/robots/AR1/tracks final/IMG_0501.JPG[/IMG]
[IMG]https://dl.dropboxusercontent.com/u/79058769/pics/robots/AR1/tracks final/IMG_0502.JPG[/IMG]
[IMG]https://dl.dropboxusercontent.com/u/79058769/pics/robots/AR1/tracks final/IMG_0505.JPG[/IMG]
[IMG]https://dl.dropboxusercontent.com/u/79058769/pics/robots/AR1/tracks final/IMG_0508.JPG[/IMG]
[IMG]https://dl.dropboxusercontent.com/u/79058769/pics/robots/AR1/tracks final/IMG_0509.JPG[/IMG]
[IMG]https://dl.dropboxusercontent.com/u/79058769/pics/robots/AR1/tracks final/IMG_0512.JPG[/IMG]
[IMG]https://dl.dropboxusercontent.com/u/79058769/pics/robots/AR1/tracks final/IMG_0513.JPG[/IMG]
I always build my own parts and - well - mostly a couple of times since I am not good at machining.
What does it cost to get parts manufactured?
Enjoy!
Mike
It's 0.125" 6061 T6 aluminum the total cost for all four pieces was $105, I asked for unfinished and was shocked to see it is somewhat polished. I could have doubled the order for only a couple bucks more but I'm still prototyping. It has a long way to go and things are going to change. Once everything is finalized I will give away the plans for free or sell low cost kits.
There are very few robots this size on the market that are affordable. I'd like to fill that gap however I can. Small robots like the BOE Bot are awesome until you want to put a small computer on it. The next affordable option for me was to build something large with wheelchair motors and that turned out to be far to big and could just barely fit through a doorway. This size should be enough to carry enough payload to put whatever you want on there.
I have only designed the track part, now I'm using L channel, sheet metal and more ePVC to develop a base. From the base I will use a large diameter PVC pipe to go up. That will have a "T" piece on top, the side of the T will be for arms, and the top will be for some kind of head. I'll post a drawing soon.
Here is the most recent pic:
[video=youtube_share;gqLkV6VKeRQ]
[video=youtube_share;aBAnf_ImpeE]
Someone asked if I had XCTU on my iPhone. No, I do not. I'm using remote desktop to my computer and XBees. None of that matters, this is all for testing. I thought it would be easier to show what is going on in the video by using my phone, but the screen doesn't show up well so I may just use a screen recorder.
[video=youtube_share;WgwezBfhljQ]
Sorry, I've kind of been behind on forum stuff lately. I've been busier than usual with job and life.
I haven't watch the videos yet. I'll do so soon.
Here's the main program if you want to check it out - https://dl.dropboxusercontent.com/u/79058769/pics/robots/AR1/code/Main.spin
I have to take a step back again and fix the tracks. In the last video above you can see when the tracks go fast there is some kind of alignment issue. I'm not going any further until that is resolved. I need to cut down all of the standoffs the nylon idlers run on. I don't know why they are .12" longer than the wheel hub axle but I can tell that is adding to an alignment issue. I'd like to get that right to cut down on current consumption.
I'll post some pics of the issue.
The red circle of doom is showing the gap. The idler needs this outside screw in place or there is slop. Originally I had that screw in place but it bends the frame and I still think it's a little off due to the bend.
[IMG]https://dl.dropboxusercontent.com/u/79058769/pics/robots/AR1/tracks final/standoff1.JPG[/IMG]
Here you can see the idler hub next to the standoff:
[IMG]https://dl.dropboxusercontent.com/u/79058769/pics/robots/AR1/tracks final/standoff2.JPG[/IMG]
A spacer won't do the trick because of the bearings.
[video=youtube_share;i77_O12RduU]
I think the next thing to do is reduce the friction. I need some kind of lubrication on the track pins and between the track segments, and I think I'll be a lot happier with the results.
White lithium grease comes to mind, but taking the tracks apart doesn't appeal to me as much as some kind of poly safe spray. Even if the poly safe spray lube was to wear off quicker, it would still be easier to reapply.
Any ideas?
Do your bearings have flanges? If so what if you switched the flange from the outside to the inside? Would the width of two flanges make up the difference?
It just seems really strange the parts wouldn't fit together correctly.
You're correct, on the Johnny 5 the bearings are opposite how I have them.
[IMG]https://dl.dropboxusercontent.com/u/79058769/pics/robots/AR1/tracks final/j5bearings.jpg[/IMG]
That is where the confusion came into play. Having the flange on the inside increased the friction of the idler sprocket when the screws were tight. It also made it so that the bearing could pull through the chassis should one side come loose. With the flange on the outside the pulling force of the opposite screw holds it in place much better. Still not sure why the J5 chassis is like that.
It wasn't a problem to shave some length from the standoffs at all, I used a little jig I made and a flat file and did 5 at a time in a couple minutes.
First issue, the tracks, noise and alignment. I post on another forum (gasp) a video of my setup and have been reassured that everything is normal. It's still a little hard for me to swallow the current drain of these tracks, but as of now I'm going to stop complaining about it since I know that it is as good as it will get.
Second, the EMIC 2 is cool and everything but without the visual feedback, well it's just not the same. Everything was great on a single breadboard, I even used a level shifter which I still do not feel is necessary for a short strip of WS LEDs. The problem came down to a noisy 5v rail. So now I have two 5v rails. One switching and one linear. The linear drives the WS LED strip. The LEDs vary in intensity in this video but that is because of my code which is refreshing them when it doesn't need to. I could fix that but it doesn't make sense right now since it will run an entirely different routine, as well as be called upon from another piece of code.
You may notice the speech is delayed after the movement of the tracks, this is by design since I want the movement of the robot to be a higher priority than the speech feedback. I also have to mention using the EMIC 2 is so much better to debug code than the PST or XCTU given the proper environment.
Anyway, here it is, if you're feelin real crazy watch it in 1080:
[video=youtube_share;1OMrpWUbUXg]
[video=youtube_share;JuYaI70Y4Ak]