Specs: 5 ft/sec and 15 ft/sec
The gear ratio is as follows. 14 tooth gear on the motor geared to a 50 tooth gear.
Then, the two gear ratios to the dog shaft are 25 tooth to a 50 tooth, and a 40 tooth to a 35 tooth gear. The final reduction is a 14 tooth gear and a 50 tooth gear.
I designed the transmission to use the 50 tooth and 14 tooth gears out of the KOP transmission to save money.
I would love to hear any comments about what is most likely to fail, and how to improve it.
Looks good! You should include some sort of description or explaination though!
I’d try to eliminate the third shaft by going with bigger gears on the second shaft, a bigger gear on the primary reduction (the middle one), and smaller gears on the first shaft. What is your target speed?
I am shooting for 15 ft/sec and 5 ft/sec. I chose 3 stages because I could then use the free 14 tooth and 50 tooth gears from the KOP transmission and machine them how I want.
The gear ratio is as follows. 14 tooth gear on the motor geared to a 50 tooth gear. Then, the two gear ratios to the dog shaft are 25 tooth to a 50 tooth, and a 40 tooth to a 35 tooth gear. The final reduction is a 14 tooth gear and a 50 tooth gear.
As you can see, there are a lot of 14 and 50 tooth gears in the gear ratios that I chose. I only have to buy 25 tooth gears, 40 tooth gears, and 35 tooth gears.
15 fps is probably too fast for FIRST purposes. Most teams and drivers agree you lose the ability to easily control your robot after about 11-12 fps. Additionally, you can easily outrun most other teams at around the same speed.
If you wanted to keep the 3:1 ratio, I’d aim 12 fps and 4 fps (with inefficiencies slowing it slightly from those values).
I agree that 15fps is too fast, if that’s the final speed. But it looks like he’s gearing the robot to move at 15fps, which means it’ll probably end up in the 12-13fps range, which is good.
I think the ideal low gear is in the 5-6 fps range, and the ideal high gear is in the 11-12fps range, after inefficiencies, but those numbers (especially low gear) may not be very feasible, depending on the weight of the robot, and the tread being used (that is, how strong the robot needsto/will be able to push). I have always thought that robots moving in the 3-4fps range were too slow.
Basically, I agree with the OP’s speeds… I would even say to make the low gear a bit faster, but blah… I understand the current draw issues that may arise as soon as some good tread gets put on the wheels.
I agree that most drivers think that 12 fps is a top speed, but in my opinion that top speed in the gearbox is almost irrelevant. As long as you have enough torque to move your robot at that speed, then you can always drive slower. Just because most drivers push the sticks all the way forward when they drive doesn’t mean it is required. It is also possible to reduce the speed in software if you really wanted to.
As for outrunning most other teams…if most teams shoot for 12 fps you won’t 
The general rule of drive trains is you can always go slower, and you can always have use less torque; it is when you try to do things the other way that you run into problems.
Yeah, but he could get more torque AND the same speed if he’s got it designed to max out at that desired speed instead of reducing the power to get there.
Most teams I’ve talked to don’t shoot for 12 fps, they shoot for around 10 fps, so you still should be able to outrun them.
What part of the transmission is most likely to break? I am mostly worried that the first time that we get into a pushing match, the transmission will fail.
I am having trouble putting this next sentence in plain english, so bare with me.
When the dog’s teeth just leave the face of one of the two gears with holes cut in them to accept the dog’s teeth, how much clearance should I have between the dog’s other set of teeth and the other gear?
Thanks for all of the feedback so far.
<These are my thoughts, some suggestions from my experience, if some of my hypotheses are no good, please correct me>
If you want to shoot for 15 fps and your team has a good driver, I’d say go for it. Last year, we put our bot up on blocks to find its freewheeling speed, and high gear was around 14 fps. Once we got the bot on the ground, we never actually clocked it very accurately, but our rough figures said that it was running about 12 fps. You may not lose as much speed as we did, depending on the efficiency of your drivetrain and the type of wheels you use. (We ran a six wheel #35 chain drive with 4 go-cart tires and two pneumatic Skyway look-alikes. If you use smaller wheels I wouldn’t think your bot would slow down as much as ours did) Ours seemed blistering fast, it was fine for '06 (the field was open and our driver was great), but I wouldn’t recommend it for a playing field that has a lot of obstacles like it was in ‘05.
Also- two things we learned the hard way-
-make certain that the key for your output sproket/pulley is TIGHT. We had problems with shearing, yes shearing, keys off on our output because they were too lose and were getting slammed back and forth a lot due to our defensive strategy (ram, back off, repeat, etc.)
-try not to shift when your motors are under power, it tends to be hard on the dog. when we took our gearboxes apart this summer, the dogs’ corners were worn worse than they should’ve been. Just try to shift only when the motors aren’t running at full power, and I think you’ll be alright.
As for the clearance, I know exactly what you are talking about, but I can’t remember what ours was. I’ll check the bot tomorrow and get back to you. I would reckon that 1/8 would be plenty, but I honestly can’t remember.
Good luck!
While optimization is important in design, when it comes to your high speed of a FIRST drive train, especially with one that shifts on the fly you have no need for higher torque at your max speed. At that speed any robot running the kit bot drive and upgraded wheels will be able to push you anyway. Since you would never try and push someone at high speed and it is easy enough to change gears if it ever came up. The only purpose of high gear is to move your bot as fast as possible from one place to another, as long as you can move yourself and the game element if relevant, maximum torque at that point isn’t required.
True, it’s not required, but it sure is nice to have. Our '06 bot had the high- gear torque to move many a shooter bot, and it came in handy. It took about 1 1/2 seconds for our bot to lock into low gear form high and get moving again (shifting on the fly= not good), and that was enough time for a good shooter to put in two or three balls, enough to seal a loss in some cases. So, no, it’s not really a requirement, but I would highly recommend utilizing all the torque available.
The constraining point of this would be the length of your pneumatic cylinder stroke. Typically dog shifters use a .75" bore 1" stroke cylinder. You should design it so the dog is fully engaged when the cylinder is fully extended and fully retracted. As for the clearances, space of the neutral gear, you want to minimize it as much as possible. Assuming that your dog is at least engaged half the gear thickness on both sides I would figure out the minimum spacing by calculating the throw of the cylinder and the thickness of the width of the dog. I would have no more then .125 clearance, shooting for about .0625, remembering that the larger the clearance the less engagement with the gears you get on the dog.
Acceleration is a factor, as well. I agree with Jeremiah that last year (2006) was a good year to have a “zippy” speed to fly across the field, but in 2005 (and, arguably 2004) a superfast robot could have proven to be a hinderance. In 2004/5, all the stop and go and fine adjustments left the fastest of robots not usually going above 9-10 fps, even though they were capable of more speed.
Joel raises an excellent point that I meant to mention earlier. In many games there often isn’t enough open space to reach full speed for any extended period of time (even in a full open field like 2006, teams wouldn’t be at full speed for more than a few seconds). Designing and anticipating such a fast speed may be wasted, as it may never be reached for a long enough to make it be worth it.
A majority of the “racing” in FIRST actually happens when a bot breaks free of a defense in one “axis” and then moves along the other to create space from the defender. For example, Team A is trying to score in goal A, which is down the length of the field. Team B is blocking them by moving back and forth along the width of the field, preventing Team A from getting by. Team A eventually slips by Team B, and moves down the length of the field towards the goals as fast as possible, as to score as many points as they can before Team B and orient themselves along the length of the field to resume defending against the Team A. In this situation, the maximum speed of the bot is only reached for a couple seconds, and the speed in which Team B can turn is usually far more important than either of the robot’s top speed. Additionally, in this situation, both teams pushing power and speed come into play when Team B is blocking Team A in the beginning, as the having both pushing matches, and the speed the correct their positioning is critical. The ability to shift-on-the-fly is typically fairly helpful for both teams in this situation as well.
While the maximum torque typically isn’t required (with the exception of a few rare case pushing matches), you do have to make sure that you can get enough torque while running your 15 fps gearbox, with inefficiencies, at 80% to move. If your just moving your 130ish (battery weight) pound robot, that should be no issue, but if there is a heavy game element (such as goal, or multiple goals) that you have to move as well, it may take near your maximum torque. Or even a worst case scenario, your robot, multiple goals (and other game objects), and other robots (see Beatty 2002, although they didn’t exactly move in a conventional manner or a blistering pace).
While all of the points you guys bring up are valid, and great things to keep in mind when choosing gear ratios during the season, I just chose those two speeds because those were speeds that I thought might be useful for a game like last year.
When the new season starts, I am going to re-evaluate the gear ratios, and redesign that part of the transmission. I see no point in speculating about the next season and choosing a specific gear ratio before the challenge is released. I just started with 2 gear ratios so that I would have a basic design to build off of and adjust to the challenge when the season starts. Most likely, I will keep the 5 ft/sec, and adjust the top speed to a speed that fits the challenge better.
Thanks for all of the great input on my design.
What dimensions do teams normally use for the size of their dog, and the depth that the teeth of the dog go into the gears? I am currently using 1/2 inch hex, and the dog’s teeth are 7/16 of an inch long.
I use a square dog with an outside shift plate. ‘Teeth’ are .160" long and engagement in the gear is .140". You can see this is the white papers in our 2005 gearbox.
This design has had no problems in two years of competition. The only thing about building your own is you have to build, and break in, new gearboxes every year.
I think last year I did something like .020 neutral space. I mean, if it clears, it clears; no need to clear by a big amount. Just make sure that under no circumstances are both gears angaged at once.
Our shifting dogs went something like .13" into the gears and were something like 1" OD.
Yea, we found out after getting the kit tranny’s that it was alot more efficient to use them. Instead of spending 2 - 3 weeks designing and machining tanny’s, we can use that extra time to design and perfect another aspect of the machine. Remember it’s not always about speed and strength(but even if it was, the kit tranny’s provide enough speed and torque), but how efficient you are at doing a certain task as well.
But kudos to all the teams who design bot kicking tranny’s! haha 
If anyone is interested, here are the Solidworks files for the transmission and drivetrain. If you end up using any part of them durring the season, I would love to know about it. Please let me know if there is a problem with the files, or something doesn’t make sense.