Pole Lifter - Winch

[nick_sheets]

Hello Chiefdelphi

Our team was looking to add a lifter for offseasons comps. We have one cim available and we were looking for some help on building a 1114 type setup. But we are currently stuck on figuring how to find a ratio needed for lifting.

Thanks in advance.

Nick

[Aren Siekmeier]

You'll need to decide a couple things. First, how will you hold yourself once you've lifted? Friction? The platform? What will keep the robot from back driving the hinge? Will it lean into the tower, or will you use a winch? Second, a single CIM provides about 1.2 N-m or [edit].88[/edit] ftlbs, times whatever gearing you are using. This torque divided by the length of your arm must be at least the weight of your robot, preferably more, or you will not be able to lift. Play around with arm lengths and gear ratios until something works out. Remember, if you're relying on friction to hold you on, the arm length affects the normal force at the grip. For the gripper, mock up some prototypes and try to find a simple way to do it that hopefully doesn't require any extra actuation. Our team didn't build like this, but we were thinking about it for a while. It definitely has its advantages, specifically the flexibility of hanging from any zone (and it can be very fast).

[rahilm]

Originally Posted by nick_sheets

Hello Chiefdelphi

Our team was looking to add a lifter for offseasons comps. We have one cim available and we were looking for some help on building a 1114 type setup. But we are currently stuck on figuring how to find a ratio needed for lifting.

Thanks in advance.

Nick

A great tool to go through this is JVN's Design Calculator ( http://www.chiefdelphi.com/media/papers/2059? ). It has all the motor data and you can input gear ratios and it will calculate the output torque and everything you need to know. Coupled with the math provided above, you can quickly figure out the ratios you're going to want for your lifter. I would suggest when inputting gear ratios to use gears that are available from AndyMark (or whichever supplier you prefer) so you're not having to make your own gears.

Also, if you need more help, feel free to PM me. Our team is in the final stages of our off-season hanging design that uses the vertical pole, so we've gone through the math involved.

Edit: I just realized, you're from Fremont, so we're ~30 miles away from you. If you'd like, you could come down to San Jose, and we could help you through the whole design process

[Chris is me]

Once you use JVN's calculator to size your winch properly, you'll have some other considerations. Mainly, you need to prevent your system from backdriving. Virtually every team used some kind of ratcheting system for this but some teams (notably 1114) used other methods. Basically, your robot probably won't hold itself up with no power, so I would take the time to design an anti-backdrive solution into your system right now rather than find out later.

I assume by "1114 style" you meant lifter design, so it sounds like you have that worked out.

[nick_sheets]

Thanks for all the quick responses.

We were just stuck on the math. It was just hard coming up with ratios. Thank you for all your help.

@rahilm: highly curious, what overall reduction did you guys go with?

Nick

[Chris is me]

Forgot to add: For 2791's winch, we run a CIM through an AM Planetary and AM Toughbox Nano for an approximate reduction of 46:1. Our winch spools string around a 1 inch diameter drum; your mileage will vary based on the size of your winch.

Winches will need a lot less of a reduction than a chained drive would.

[rahilm]

Originally Posted by nick_sheets

Thanks for all the quick responses.

We were just stuck on the math. It was just hard coming up with ratios. Thank you for all your help.

@rahilm: highly curious, what overall reduction did you guys go with?

Nick

We use a PTO system, meaning we're using all 4 of our drivetrain CIMs to power the hanger. The overall reduction we went with is 400:1, which includes the drive transmissions, and additional gearbox joining both sides of the drive, and then a final reduction through chain. We're looking at a lift time of ~2 seconds.

I think I should mention that our hanging system is closer to 254 than 1114 in design in that we are using a chain drive hooked up to the arm doing the lifting, instead of using a winch like 1114 does.

[AdamHeard]

For a system such as this, you really need to model and do the math for YOUR system.

The lever arms vary from team to team, the CG location varies, etc... For everything. To say team X had a ratio of 50:1 so we can use that is not going to work.

You need to look at your rotating joint, and figure out what torque the CG is causign around that. That number will vary throughout the rotation. For conservative gearing, through gearing have that equal 20% of the stall torque, that assures it'll be moving. If you gear it near stall, you'll just sit and pop breakers.

[rahilm]

Originally Posted by AdamHeard

For a system such as this, you really need to model and do the math for YOUR system.

The lever arms vary from team to team, the CG location varies, etc... For everything. To say team X had a ratio of 50:1 so we can use that is not going to work.

You need to look at your rotating joint, and figure out what torque the CG is causign around that. That number will vary throughout the rotation. For conservative gearing, through gearing have that equal 20% of the stall torque, that assures it'll be moving. If you gear it near stall, you'll just sit and pop breakers.

Very much agreed.

For example, we asked around a lot before plunging into design. We asked, IIRC, 33, 148, 254, and 1114, all of which use a wildly large range of reductions. The important part of asking around though, was understanding the underlying concepts involved between the different designs. IIRC, 33 had a fairly small arm they were rotating about, so they were able to use 1 CIM and hang fast. 254 was like us, using a PTO, because they had a much larger arm to rotate about, and so to hang in a short amount of time needed much more power. (Don't take my word for their systems. If you're curious, ask the teams who made them. This is my interpretation of their explanations) Using the information they gave us, we were able to determine what kind of ratios we needed to hang in the time we felt reasonable.

[NickE]

Another very significant consideration is ability to go under the tunnel. If you can go under the tunnel and would like to continue to have this feature, you will need to design some sort of two-stage mechanism so that you can get your lifter above the level of the platform to lift. This could either be a two-stage arm (like 254), an arm and an elevator (like 25), a telescoping arm (like 233) or something else. You will notice that the robots which hang in one motion (like 1114, 33, 27, 148/217, etc.) do so by starting the match with their hanger above the level of the platform and cannot go under the tunnel (Note: there may be an exception to this that I'm not aware of).

[Ether]

Originally Posted by compwiztobe

a single CIM provides about 1.2 N-m or 86 ftlbs,

Double-check that 86 ftlbs number. I don't think it's correct. You might want to edit it so someone doesn't use it.


~

[JamesCH95]

Originally Posted by compwiztobe

, a single CIM provides about 1.2 N-m or 86 ftlbs

0.86ft*lbf Be wary of the order of magnitude!

Edit: Ether beat me to it....

[sgreco]

There's nothing wrong with ratcheting, but my team was originally going to use a ratcheting system, but my team found it much simpler to attach a lag bolt to the arm a gate latch to the robot. Whenever we lifted, it would lock in place. We oriented the gate latch such that gravity would keep it open during the match so we could keep our arm down. Then we would deploy the arm and grab the tower. Once we got to about a 45 degree angle to the ground, gravity would act differently on the gate latch such that it would shut and lock once got ourselves to our 90degree hanging position. It was simpler for us than the ratcheting systems we prototyped.

As for lift speed, you won't get as fast a lift speed as other teams with simply one CIM. Many of the fastest lifts this year used a system in which they would disengage the their drive motors and hook them up to the winch in order to lift quickly. You won't be able to accomplish the speed of 1114's lift with just one CIM, but that doesn't mean you can't accomplish something comparable. My team used a BaneBots planetary gearbox (http://banebots.com/c/P80K-nnnn-0005) geared 12:1, then a custom gear ratio of 3.5:1 in order to optimize a balance of torque and speed. We had some flaws getting properly aligned, but it would only take ~3 seconds to lift given those 12:1 and 3.5:1 ratios.

(It should be noted that this ratio was not pivoting around the axis of rotation. We had a winch attached to the top of the arm that was 24 inches from the pivot point.)