We don’t plan on using pneumatics on our robot so I was looking for ideas that used something else to control/move a mechanism that would brake/stop elevator.
I was wondering if servos can be used to control brakes or some other kind of anti-backdrive component. Poking around andymark I saw shifting gearboxes controlled by servos for example. I’ve seen people on CD discuss bike brakes, ratchets, etc…
But I haven’t seen anyone discuss how you can implement these ideas. Anyone done this before and have details/pics they can share?
(note, I’m not a mechanical type… i do software, just trying to get ideas to share with my team).
Look at HOT’s tech notes and see how they used a motor, I believe it was a window motor, to actuate their shifting. You can do a linkage that actuates the bike disc break on something like a cam.
I don’t know what your gearbox ratio is to do your lifting or how much you wan’t to lift but I would strongly suggest looking at your motors used, gear ratios, and power consumption to see if you can hold is strictly with programming which would be stalling the motor. For something like a cim, it wouldn’t be as big of a deal to stall under the right loads compared to something like a 550 or 775 that relies on the turning of the shaft to cool it. There are some pretty good papers on Chief already talking about thermal mass and such.
-Ronnie
In addition to worm drive boxes, there are some drill gearboxes that have locking chucks which essentially prevent backdriving. The most popular one is the Dewalt XRP and there is a cheaper one from Harbor Freight.
For our climber for Ultimate Ascent we used a ratchet wheel and pawl obtained from McMaster-Carr and used a solenoid to release it. With a ratchet mechanism it will wedge in place so you’ll have to drive it up before a legal solenoid will have the power to release the pawl. http://www.mcmaster.com/#gear-pawls/=vkbtc5 They aren’t cheap and you’ll need to fabricate something to connect it to your driven mechanism as well as fabricate a way to connect the pawl to your solenoid. Stainless is not as easy to machine as mild steel or aluminum. I do not know what solenoid we used or where it was obtained.
It may just be simpler to use a transmission which cannot backdrive, like a worm gear. They provide large amounts of reduction in a small amount of space, and do reasonably well with large amount of torque. We plan on using a 10:1 worm reduction in our elevator for exactly these reasons.
If you’d like more details on the worm drive we’re using, let me know and I can send you some specifics.
Provided your torque requirements aren’t too high, and you’re using a motor which can handle it, you can implement a brake in software. Short version, you can use a PID to maintain the position of the motor using a holding current. We’ve done this before using CIMs and Bag motors. This does require a sensor.
We used a disc brake on our 2014 shooter and found it very easy to work with. You can check out our CAD for details of how we implemented it. We used pneumatic to pull the cable, but you could just as easily use a spool or leaver arm on a very large versa planetary reduction to pull (I think you could do an over-center linkage thing with a hard stop kinda like the cho-cho linkage to have no holding current and make the software pretty easy). You would have to do the math on those details.
We’re using a servo with a cam to actuate a pin, which will interlock with holes drilled in one of our gearbox gears. It’s based off drivetrain braking systems by team 25. A cam is a nice mechanism to use for this application, since you really don’t need much travel at all, and it can be made completely backdrive resistant through strategic positioning of dwells. For a pure elevator, I agree that software solutions are probably the best, but our robot has an interesting feature which requires it to hold its position while disabled.
In 2014, we used a dog gear which interlocked with a custom made ratchet and pawl mechanism, using WCP dog shift components. It could have just as easily been done with the andymark servo shifting setup. It worked, but we weren’t happy with how much force disengagement took while under load (the eventual solution was to pulse our winch forwards ever so slightly when disengaging, taking the side load off the dog gear).
In general look at catapult pull down mechanisms from last year… Many of them had to have the ability to lock in place without stalling the motor, and could be adapted to work for this as well.
We are also using a worm gear drive - AndyMark RAW. Many right angle drives are worm gears and can hold position without requiring the motor to be energized. Bevel gears are also right angle, but don’t exhibit the anti-backdrive feature. Other legal worm drive motors include the window motors and the tailgate motor.
Just a warning that the window motor also has a pin that engages when not energized that causes issues with speed controllers - there are CD threads on how to remove the pin.
Also keep in mind that the ratcheted gearboxes from last year can not drive in both directions - by design. Depending on how you unload your elevator, you may need to drive the motor backwards.
We chose the RAW drive due to its simplicity to control. The PID loop does not have to do anything special (like engage a break) and can drive to a desired position in either direction. Given the relatively large load of up to 60lb, we were concerned with the holding current for traditional gearboxes.
AndyMark shows the RAW box as out of stock, and has removed the estimated restock time of Jan 26th. We did not want to wait till Jan 26th and ordered the individual part to build our own - which turn out to be more difficult than imagined. They do seem to have restocked the “worm box” and perhaps it is sturdy enough for your elevator. Other wise, your best bet may be the tailgate motor if you want a worm drive.
I hope our software does not add 60lb of friction to the system :yikes: . Although that might explain why the robot does not drive as fast as we would like
I’m drawing a blank…probably because I can’t see what it is we’re trying to figure out how to stop moving. Can you post some pictures or drawings or renderings of the robot or design?
we had success with a “ratchet/pawl” where the pawl was a heavy enough piece of scrap steel on a hinge that gravity lowered it to engage the ratchet, and a window motor pulled a piece of cable attached to the pawl to release it. a window motor hub can be easily altered to actuate a simple switch which would then (after adding some simple code) make this window motor into a sort of binary servo motor.
while mechanically this worked fine for us, [STRIKE]cheap[/STRIKE]free and simple were two of our main constraints, which may or may not be the case for your team…
