Thanks for the interest in the design, I am pretty happy so far with how it is turning out. Still have some things to figure out, but the goal is to have it up and running in a couple weeks. At that time, I may make a few extra chassis components from a sheet of the material for those interested. Long term goal is to have it setup with a small H-Bridge and motors due to the cost of the Naro servo and that may alter some of the design. For now, I really like the Naro servos embedded into the lower deck.
I am planning on putting one into the raffle at OPC, but it will not have the PropBSC which means I have to determine what other setup to use. I am thinking of a Quickstart or a M44D40+ and PowerTwig combo. Either way, the LiPo can't be the power supply since it needs to power the servos at 4.8v+ and those options don't have a boost circuit. The PropBSC has that built in, so that's easy that way, but they are not in production (yet). The 2 front holes for the standoffs are spaced to also be used to mount a Quickstart, but the power source selection is still an issue.
Thanks Matt! Yes, actually, I already verified that Rocka conveniently fits through one of the rear side panel on my Stingray. Rocka's final width will be about 4" and less than 3" tall.
On a side note, I ran some numbers and it looks like a complete chassis could easily be sold for $10. So, I may keep the format progressing.
Got my hardware order in today and did some test fits so I can finish my drawing updates. Using 1" screws, 1/2" spacers, and plain nuts, the configuration with my PropBSC setup is pretty clean. The wheels in the picture were cut from some 2" wide piece of scrap foam from another project, so it was beeter to use it to make the 1.7" diameter wheel samples than to throw it away. The hub of the finished MDF wheel will have an inset for a servo horn so that the wheel can be attached to the servo's gear head with mating teeth. The tail design isn't finished yet, so I just placed a 1/4" standoff under the back end to make it sit where it should with the tail installed. It is surprisingly strong and rigid when assembled. Will definitely take a beating.
Switching over to 1.5" screws and 5/8" standoffs provides the mounting for a Quickstart and the 5 AA battery holder. (The PropBSC setup uses a LiPo, so it hides between the decks rather well.) I still need to setup a version with my M44D40+ and PowerTwig. I think that will be a cleaner setup than the Quickstart. Again, the issue with such a small platform is power, but then again, one key feature of the PropBSC is to address these types of scenarios.
Next up will be to finish the drawing for the tail (looking at a simple curved bar tail and the BOE-Bot's tail wheel) and finalize the wheels. Also, I have created a "hardware free" concept that does not need screws or standoffs. More details on that next week.
It's fun to see it packing a QuickStart! That gives an excellent idea of scale along with seeing the battery pack snuggled in there. This is shaping up into a very nice package!
First, my wheel design. After looking for ways to make the wheel to servo attachment clean, I began looking at creative ways to utilize a standard servo horn. I could just glue it to the wheel, but of course wanted to ensure it's alignment was proper and would not overstress the servo hub. My first concept includes a 4 way servo horn drawn in the center of the wheel and then setup for a slow and powerful raster on the laser to etch down the MDF. This created a "nest" for the servo horn. A hole was cut into the center to allow the horn's hub socket to mate properly with the servo hub.
I lucked out with the dimensioning and the servo horn snapped into it's nest perfectly. This turned out well, but takes 2.5 minutes to cut a pair of wheels and the nest needs to be a bit deeper which would put cut time closer to 3 minutes. It needs to be deeper so that the horn's hub socket could fully mate to the hub without the side of the wheel hitting the servo hub encasement. It did look good on the Rocka though:
So I cut a few more with some variations, one which added a larger hole in the center to clear the top of the servo hub's encasement. I then cut a second wheel with a simple hole in the middle for accessing the servo horn screw and added this to the first wheel. The two wheels are then held together by a SumoBot "tire" (black rubber band). Since the two wheels are only held together by the band, I can attach the first with the screw and then add a solid second wheel which would give a really clean look to the wheel.
By adding the larger hole in the center, that reduced the amount to raster (shown in blue), so the cut time was reduced to 100 seconds for a pair. (will have to add about 25 seconds to cut two solid wheels though) Red lines are vector cut lines:
With the new parts cut out from my most recent round of prototyping:
The lower deck houses the servos. I had to predrill the MDF for the servo screws as MDF loves to split when screwed into it sideways with screws that are greater than 25% of the width of the MDF. Since this MDF is o.2" thick, that doesn't leave much room for a screw that won't require predrilling. Mounting them in and routed the wires for them as well as the LiPo pack.
You can also see the tail dragger mounted. I intended to add an extension to make it look like a stingray tail, but failed in that this extends from the lower deck which makes it look more like a deformity than anything natural. Here is the stackup as viewed from the front showing the IR sensor pairs and it's profile:
I think the lower deck should be lower to get it closer to it's big cousin, the Stingray. And the stackup from the rear-side emphasizing the LiPo and XBee The XBee adapter board is on the underside of the top deck. When you plug in the XBee into the sockets, it sandwiches the deck and holds perfectly:
Up next, Rocka is ready to get some finishing touches......
So, here is the first "official" prototype built up to about 80% complete. It still needs a few tweaks and some wiring to connect a few things properly. And of course code, of which I haven't even thought of yet......
I really need to start a thread specifically for Rocka since this thread is actually about object detection and drive motors..........
Have you tried making wheels with servo splines cut out? It works really well and is quick to laser. All you need to add is a tiny washer between the screw head and the wheel to hold it on.
I have thought about direct attach by cutting the splines, but figured the accuracy of the dimensions would be critical enough to make it a pain. I may try it though if I make a version upscaled to use standard size CR servos. Right now my plans are to make a handful of this version, then redesign to use dc motors.
That is definitely my plan. There will be one in the raffle that will have either a Quickstart or a M44D40+/PowerTwig combo running it and possibly the XBee. My only concern now is being able to have the code for it done in time. I have not done any coding for over 2 months, plus I have no experience with the IR Tx/Rx pair on a propeller, so will either have to find some code to cannibalize or bribe a few gurus to help me out Friday night at the reception.
Had some delays with progress, so while one will be assembled at OPC, I don't think I will have it wired/coded. Maybe I can get some help Friday night or Saturday, hint, hint.... There will be a kit of parts for a Rocka for the raffle of course.
Anyhow, I made some changes to the wheel setup and have that down. I also played with my idea of using vertical beams that lock into the upper and lower decks to mount the servos rather than the cutout I currently have. My firs tcuts tonight proved to be considered for the final design. I will still leave the holes for standard hardware as using something other than a Naro Servo is on my list. The long term plan is DC motors and a small daughterboard for the mainboard. Anyhow, on to pics!
Look ma, no hardware, just wood! (where did I save that "erco approved" logo?......)
Upper and lower decks sandwiching one "beam"
Upper and lower decks with both beams and servo. Each beam will have a hole sized for an appropriate screw to secure the servo.
On a side note, I did master the Naro Servo conversion to continuous rotation. I can modify one in about 5 minutes with little effort. The only difficulty is securing the pot in it's center position. The pots used on these tiny servos are pretty sloppy when not stabilized by the servo housing, so simply touching the shaft during re-assembly will knock it off true center of 90°. For example, from my early trials, I have two servos that work perfectly, but center at 88° and 85°. Now that I can get them to center at 90°, I'll be working on pictures and probably an instructable after OPC.
Comments
I am planning on putting one into the raffle at OPC, but it will not have the PropBSC which means I have to determine what other setup to use. I am thinking of a Quickstart or a M44D40+ and PowerTwig combo. Either way, the LiPo can't be the power supply since it needs to power the servos at 4.8v+ and those options don't have a boost circuit. The PropBSC has that built in, so that's easy that way, but they are not in production (yet). The 2 front holes for the standoffs are spaced to also be used to mount a Quickstart, but the power source selection is still an issue.
-MattG
On a side note, I ran some numbers and it looks like a complete chassis could easily be sold for $10. So, I may keep the format progressing.
Switching over to 1.5" screws and 5/8" standoffs provides the mounting for a Quickstart and the 5 AA battery holder. (The PropBSC setup uses a LiPo, so it hides between the decks rather well.) I still need to setup a version with my M44D40+ and PowerTwig. I think that will be a cleaner setup than the Quickstart. Again, the issue with such a small platform is power, but then again, one key feature of the PropBSC is to address these types of scenarios.
Next up will be to finish the drawing for the tail (looking at a simple curved bar tail and the BOE-Bot's tail wheel) and finalize the wheels. Also, I have created a "hardware free" concept that does not need screws or standoffs. More details on that next week.
First, my wheel design. After looking for ways to make the wheel to servo attachment clean, I began looking at creative ways to utilize a standard servo horn. I could just glue it to the wheel, but of course wanted to ensure it's alignment was proper and would not overstress the servo hub. My first concept includes a 4 way servo horn drawn in the center of the wheel and then setup for a slow and powerful raster on the laser to etch down the MDF. This created a "nest" for the servo horn. A hole was cut into the center to allow the horn's hub socket to mate properly with the servo hub.
I lucked out with the dimensioning and the servo horn snapped into it's nest perfectly. This turned out well, but takes 2.5 minutes to cut a pair of wheels and the nest needs to be a bit deeper which would put cut time closer to 3 minutes. It needs to be deeper so that the horn's hub socket could fully mate to the hub without the side of the wheel hitting the servo hub encasement. It did look good on the Rocka though:
So I cut a few more with some variations, one which added a larger hole in the center to clear the top of the servo hub's encasement. I then cut a second wheel with a simple hole in the middle for accessing the servo horn screw and added this to the first wheel. The two wheels are then held together by a SumoBot "tire" (black rubber band). Since the two wheels are only held together by the band, I can attach the first with the screw and then add a solid second wheel which would give a really clean look to the wheel.
By adding the larger hole in the center, that reduced the amount to raster (shown in blue), so the cut time was reduced to 100 seconds for a pair. (will have to add about 25 seconds to cut two solid wheels though) Red lines are vector cut lines:
Up next, assembly with all the new parts!
The lower deck houses the servos. I had to predrill the MDF for the servo screws as MDF loves to split when screwed into it sideways with screws that are greater than 25% of the width of the MDF. Since this MDF is o.2" thick, that doesn't leave much room for a screw that won't require predrilling. Mounting them in and routed the wires for them as well as the LiPo pack.
You can also see the tail dragger mounted. I intended to add an extension to make it look like a stingray tail, but failed in that this extends from the lower deck which makes it look more like a deformity than anything natural. Here is the stackup as viewed from the front showing the IR sensor pairs and it's profile:
I think the lower deck should be lower to get it closer to it's big cousin, the Stingray. And the stackup from the rear-side emphasizing the LiPo and XBee The XBee adapter board is on the underside of the top deck. When you plug in the XBee into the sockets, it sandwiches the deck and holds perfectly:
Up next, Rocka is ready to get some finishing touches......
I really need to start a thread specifically for Rocka since this thread is actually about object detection and drive motors..........
Have you tried making wheels with servo splines cut out? It works really well and is quick to laser. All you need to add is a tiny washer between the screw head and the wheel to hold it on.
Would love to get a close up looksee.
Had some delays with progress, so while one will be assembled at OPC, I don't think I will have it wired/coded. Maybe I can get some help Friday night or Saturday, hint, hint.... There will be a kit of parts for a Rocka for the raffle of course.
Anyhow, I made some changes to the wheel setup and have that down. I also played with my idea of using vertical beams that lock into the upper and lower decks to mount the servos rather than the cutout I currently have. My firs tcuts tonight proved to be considered for the final design. I will still leave the holes for standard hardware as using something other than a Naro Servo is on my list. The long term plan is DC motors and a small daughterboard for the mainboard. Anyhow, on to pics!
Look ma, no hardware, just wood! (where did I save that "erco approved" logo?......)
Upper and lower decks sandwiching one "beam"
Upper and lower decks with both beams and servo. Each beam will have a hole sized for an appropriate screw to secure the servo.
On a side note, I did master the Naro Servo conversion to continuous rotation. I can modify one in about 5 minutes with little effort. The only difficulty is securing the pot in it's center position. The pots used on these tiny servos are pretty sloppy when not stabilized by the servo housing, so simply touching the shaft during re-assembly will knock it off true center of 90°. For example, from my early trials, I have two servos that work perfectly, but center at 88° and 85°. Now that I can get them to center at 90°, I'll be working on pictures and probably an instructable after OPC.