pic: 3CIM Ball Shifter

[AdamHeard]

Originally Posted by Andrew Schreiber

33 Ran an autoshift in '04 with their 4 speed. The the whitepaper is http://www.chiefdelphi.com/media/papers/1580 but it doesn't go into much detail on the shift algorithm. I'm sure Jim has/would describe it somewhere.

They're not out here

[apalrd]

Originally Posted by Andrew Schreiber

33 Ran an autoshift in '04 with their 4 speed. The the whitepaper is http://www.chiefdelphi.com/media/papers/1580 but it doesn't go into much detail on the shift algorithm. I'm sure Jim has/would describe it somewhere.

Our 11 code has the full implementation of it (for a 2-speed), but the code is a bit messy and I would implement it differently now. It's drivetrain/autotrans.vi. Upshift was enabled (and the driver feedback was all good), both downshifts were disabled due to driver feedback (Specifically he would rely on coasting into the peg with a tube, and the coast down would throw off his maneuvering as he was slowing down). Better calibration could have likely improved this.

The logic is fairly simple. We do three types of shifts:
Upshifts
Coast down shifts
Kick down shifts

Acceleration should be multiplied by the sign of velocity (or the abs of velocity should be used to calculate acceleration) to normalize for changes in direction.

Upshifts are based on thresholds for speed (greater than), abs of avg of throttles (for skid steer - this is pre-halo and culver drives) (greater than), vehicle acceleration (greater than), abs of diff of throttles (less than) (not turning)

Coast down shifts are based on absolute low speed (~2fps).

Kick downs are based on speed (we used 8 fps, which is higher than redline in low) (less than), abs of avg of throttles (greater than), not turning, vehicle acceleration negative and less than calibration (large negative number).

Upshift handles normal upshift driving.
Coast down shifts back when the vehicle speed is close to zero so it can upshift again at he next launch.
Kick down shifts down when you hit something and need to push.

This algorithm worked well enough for FRC. The corner cases (when turning) are simply ignored by the autoshifter, which was 'good enough'. We also impose a minimum time between shifts to prevent gear hunting, most automatic shift implementations see this and our solution is a 500ms inhibit timer.

[Ian Curtis]

Originally Posted by apalrd

I guess this differs from the 'west coast' opinion that 'high gear is where we operate, and low gear is just in case we need to push'. We operate in both gears.

These are design philosophy issues. Sometimes your design philosophy leaves every day performance on the table. Airplanes could be a lot lighter (and more efficient) if they weren't good for a very very very very bad day.

Something else I think is important for many teams to consider is that there is a driver in the loop. Even if your slow gear is Killer Bees fast, if you're driver isn't up to snuff you will spend lots of time monkeying around when you could have been scoring. I still think one of the best things FIRST could do to improve game play is make the default speed of the kitbot slower. You see lots of operators that really just don't have the practice to be efficient at top speed.

[DampRobot]

Originally Posted by AdamHeard

We use it in the same way, and then also use it for thins like balancing on the bridge last year.

For how our drivers operate, I view low gear as insurance that lets you gear as fast as you desire for other objectives, but still have a drive mode where you won't get destroyed by defense (tripping breakers, etc...)

Because of this, we love a real slow low gear.

Agreed. Maybe this is just a west coast perspective, but in my mind a slow gear gives you the ability to have a super fast top speed (15+ fps as some west coast teams like to have). A low gear is only used when you're in a pushing match, or positioning accurately such as in autonomous, which isn't all that often. I personally don't see much point at having a low gear being lower than traction limited, which is about 6fps for a 4 CIM drive.

Part of this may (and I'm speaking with a complete lack of emperical evidence here) be because of a slight efficiency advantage of WCDs over "east coast drives." Most non-WCDs I've seen have more places where power can be lost (gears, more chain reductions, 35 chain vs 25 chain, etc). West coast drives have as little as two reductions down to a wheel, and typically have either 25 chain or belts to transmit power to outside wheels. Non-WCDs on the other hand are usually based off AM shifters, which require at least three reductions for the power to release the wheels to reach the wheels. Perhaps this small added efficiency lets west coast accelerate slightly more quickly in high gear.

I know I just threw out a bunch of stuff which I really can't prove and don't 100% believe is true, stuff that I would usually not post. But, given the apparent difference between east coast and west coast perspectives here, I wanted to propose a theory that might explain some of the differences.

In terms of 6 CIM vs 4 CIM drives, I'd say 6 CIMs is nice but not really necessary. You'll accelerate better, be faster and more powerful, but there are real drawbacks. One is current draw, which detracts from other mechanisms and can trip the main breaker. The other, as Karthik pointed out, is that it takes away from your ability to have CIMs power other mechanisms.

[evanperryg]

Originally Posted by Andrew Lawrence

I love the 3-CIM shifter, but the ratio is terrible. 17 ft/s is a great high gear for 3 CIMs, and 5 ft/s is a great low gear for 3 CIMs, but speeds of 9-12 fps low or high gear don't benefit much from the extra CIM.

The problem with having such an insanely high gear is that the robot will be very hard to control. Another issue with such a huge difference in gearing is momentum. This year, my team used AM supershifters at 5/12fps. If we we were in high gear then switched to low at full speed, we would tip over 75-80% of the time. Imagine how bad it would be going between 17 and 5.

Originally Posted by Andrew Schreiber

33 Ran an autoshift in '04 with their 4 speed.

I'm pretty sure 111 did a 3-speed this year. Also, how could 4 speeds possibly be beneficial, even as an autoshift?

[Andrew Schreiber]

Originally Posted by evanperryg

The problem with having such an insanely high gear is that the robot will be very hard to control. Another issue with such a huge difference in gearing is momentum. This year, my team used AM supershifters at 5/12fps. If we we were in high gear then switched to low at full speed, we would tip over 75-80% of the time. Imagine how bad it would be going between 17 and 5.

I'm pretty sure 111 did a 3-speed this year. Also, how could 4 speeds possibly be beneficial, even as an autoshift?

Back then we had a lot less power, to get reasonable acceleration and still have a high top speed was hard.

[apalrd]

We didn't want a 4 speed, we wanted a 3 speed. We wanted a fast and very fast gear, like our current design goals for high and low, but we had less power, so we needed a low low for pushing (remember this is back before bumpers) and couldn't get away with the fast-low we like now.

But, with the resources we had, it was easier to build a 4 speed than a 3 speed (since we end up with 2 2-speeds in series). The shift patterns are then low/low,low/high,high/low,high/high, with a different ratio spread of each of the 2-speeds in series. But the 2-3 shift (going from low/high to high/low) requires a double-swap shift which is just plain nasty to execute synchronously between two shifters while under power. A possible solution is to run it like a 3-speed with two options for 2nd gear, and essentially jump-shift past the 2-3 shift. This would add complexity to the shift scheduler, although more than 2 speeds already adds complexity to our shift scheduler.

A ball shifter design can do any number of speeds just by adding more gear pairs to the shafts, so it's possible to do a 3-speed instead of a 4-speed, but it's much harder to manufacture than the dual 2-speed with andymark parts or dual 2-speed mesh shifter design (or even the dual 2-speed ball shifter).

So 4 speeds was easier to build, but the 2-3 shift is bad.

[Lil' Lavery]

Originally Posted by apalrd

A ball shifter design can do any number of speeds just by adding more gear pairs to the shafts, so it's possible to do a 3-speed instead of a 4-speed, but it's much harder to manufacture than the dual 2-speed with andymark parts or dual 2-speed mesh shifter design (or even the dual 2-speed ball shifter).

For reference, 222 has been running a 3-speed ball shifter since 2004.

http://www.chiefdelphi.com/media/papers/1539

[Andrew Schreiber]

Originally Posted by Lil' Lavery

For reference, 222 has been running a 3-speed ball shifter since 2004.

http://www.chiefdelphi.com/media/papers/1539

I thought they'd gone to just running 2 speeds as of late?

[camtunkpa]

Originally Posted by Andrew Schreiber

I thought they'd gone to just running 2 speeds as of late?

This is correct. We only ran 3 speeds in 2004. We didn't find the added complexity and weight to be worth the added gear ratio.

2013 was the first season since 2004 that we haven't made our own transmissions. This season we went with the vexpro ball shifter and will likely use them again in the future FWIW.

[Lil' Lavery]

Originally Posted by evanperryg

I'm pretty sure 111 did a 3-speed this year. Also, how could 4 speeds possibly be beneficial, even as an autoshift?

Their initial design goal was 3(+) speeds, but the method they chose allowed for 4 speeds just as easy as 3. Essentially, each shifting mechanism they incorporated doubled the potential ratios. One mechanism would be two speeds, two mechanisms allows for four potential speeds. If they added a third, it would be 8 speeds. Rather than try to explain it in greater detail myself, I'll simply use the wonderful resource they posted on the gearbox (with nice pictures that explain it better than any amount of paragraphs ever could).
http://www.chiefdelphi.com/media/papers/1580

[AlecMataloni]

Originally Posted by evanperryg

I'm pretty sure 111 did a 3-speed this year.

We used the COTS two-speed ball shifters from Vex this year. Nothing too fancy.

[Aren_Hill]

Originally Posted by cad321

loading photo ...

Quick bit of advice, it is probably good form (on both the ball shifter and dog gear setups), to put your third reduction after the shifting shaft.

This method can be seen on the Ball Shifter with optional 3rd stage, and the AndyMark Sonic/Super Shifters.

This lowers the overall torque being transferred by the shifting components, which results in less chance of failure and lower required shifting forces.
See this thread for once instance of failure: http://www.chiefdelphi.com/forums/sh...ght=20+carnage

Doing this would also allow you to make your gearbox overall shorter as you can offset the last output shaft in whichever direction you like.

[Lil' Lavery]

I have another question. Why bother with a 6-CIM, 2-speed drivetrain at all? It seems like a lot of resources to invest in only marginal gains over either a 4-CIM, 2-speed or 6-CIM single speed.

[Karthik]

Originally Posted by Lil' Lavery

I have another question. Why bother with a 6-CIM, 2-speed drivetrain at all? It seems like a lot of resources to invest in only marginal gains over either a 4-CIM, 2-speed or 6-CIM single speed.

Something very important that also needs to be considered is what is the opportunity cost of those two extra CIMs. The concept of opportunity cost is invaluable in making effective strategic designs decisions, especially when it comes to resource allocation.

Here's a Kahn Academy lesson that does a great job of explaining opportunity cost. A bit more detailed than is needed for these purposes, especially in terms of the PPF, but hey at worst you're going to learn some extra stuff.

https://www.khanacademy.org/science/...?v=pkEiHZAtoro