So our team has been thinking about designing custom gearboxes to achieve a fast high gear and strong low gear. How do teams create such fast/strong gearboxes? I tried playing around with the JVN calculator to see what I could get but couldn’t get what we wanted.
I’m not really sure what you mean by “we couldn’t get what we wanted” out of the JVN calculator. Were you using the single speed or two speed tabs? Were you using all the tooth counts of gears that are available from FIRST vendors?
My typical process for drive train gearboxes is to determine what my design parameters are for the gearbox (shifting or no? max speed/min speed? certain time to distance requirements?), then playing around with JVN until I get a set of gear ratios that work.
I guess the main thing about the JVN calulator is that, like all calculators, it’s only as good as what you put into it. It’s not magic-- one of the often ignored parts of design is determining what exactly you want out of what you’re designing. In my experience there hasn’t been a single gearbox I’ve designed that the design wasn’t at least in part simplified by the JVN calculator.
When I said “we couldn’t get what we wanted”, I meant that we couldn’t get a fast speed for high gear and a strong low gear. I was using the two speed tab, and yes, I was using the gears that Vex offered. We were hoping to get at least 13f/s for high gear, and a reduction of at least 10:1 gear ratio.
What’s driving these numbers? How do you know you want a 10:1 gear ratio? Wouldn’t that depend on the wheel size a little bit? And what you’re doing?
Well we just want a general gearbox that would have a fast high gear and a strong low gear, and we use 4 inch wheels.
Actually, I’ve alwasy been confused about something in the JVN calc: it lists the draw at maximum output load, but is that the normal drive load or pushing match load?
So, why build your own? That sounds like a pretty cut and dry use of COTS components? For example http://www.wcproducts.net/wcp-shifter-3cim/ (scroll down to the charts for the speeds) has several options that meet your needs.
I’m not saying don’t go custom, just saying that you might want to look at options and see what concepts you can steal.
WCP’s drivetrain calculator gives a more in depth analysis than the JVN calculator. http://www.wcproducts.net/how-to-drivetrain/
I think it would be nice to learn making custom gearboxes. If it doesn’t work, we could always buy them. And I never knew that site that you gave me existed, so I didn’t know if there was something like that.
This calculator only works with Vexpro and WCP gearboxes, unless you want to use the same gears and ratios for a custom box.
With 4 inch traction wheels 1678 Citrus Circuits ran a high gear of 22 fps(theoretical free speed) and a low gear of 8 fps on our 6 CIM drivetrain using VEXPRO ball shifters.
As they say, steal from the best, invent the rest. I’d agree that looking at COTS options and borrowing their ratios is probably the best option for you. You might also want to be a bit more flexible with your design-- if you’re designing for a certain speed and ratio, you might need to be flexible with your wheel size, for example.
Did you guys have any breaker issues?
Custom gearboxes are nice, but WCP 3-cim gearboxes are hard to beat.
I would go custom only if I am confident that I can get an appreciable reduction in weight (at most a 2.25lb gearbox) or size (because WCP gearboxes are pretty large). Or if you have a gearbox design you want to try in the offseason, such as a bevel box or something similar then it would be good to try custom.
Neither, it’s an overly simplified model of total current draw. It doesn’t take into account stuff like battery voltage drop.
JVN’s spreadsheet is a very good first pass, but you really should look into it deeper. Check out Ether’s drivetrain simulation program that takes that stuff into account.
This year CV Robotics did our first custom gearboxes since 2008, because we wanted a more durable and flexible to our design option. When I calculate the speed for gearboxes I design, I usually just do a rough calculation (eg. stoichiometry to get from CIM RPM [5000ish] to feet per second). Using this I can compare it to previously used gearboxes to see about how fast we would be going. This year’s gearbox was calculated at around 18ft/s:
You can see our robot on the far side in red here:
The robot may not be going 18ft/s, but next year if we want a faster robot, we can change the gearing so that our calculated value is a little larger, maybe 21ft/s.
Hope this helps,
Michael
We did custom gearboxes on our 2013 west coast drive and were very happy with the results of it. We went about finding our gear ratios by looking at what other teams had done for speeds in the past as well as analyze how our own drive trains had preformed in the past. It looks like a “fast speed” is around 18 feet per second theoretical (free speed * gear ratio * wheel dia in inches * π / ( 12 * 60 ) and low gears usually run ~ 2 - 2.5 times lower that a gearbox’s high gear. The highest high gear that I have ever seen was 24 fps theoretical on 3061’s machine this year and in talking to them they said that they felt that their high gear was too high because they couldn’t accelerate to top speed on the field and they found it difficult to control.
In finals 2 at SVR they blew the main breaker: http://www.youtube.com/watch?v=fi68AQWPtCw
I believe they said this was pushing in high gear by accident or something along those lines. They can give better clarification than I, though. 
Edit: Found it! Talks about blowing the breaker: http://www.chiefdelphi.com/forums/showthread.php?t=128530
Truck Town Thunder has built their own transmissions for** 16 **years and 2014 was the first year we bought them. I guess you can say we were experts at designing and building the strongest transmissions out there but it turns out WCP does exactly the same thing. Buying was difficult for us to do because we had never liked what reductions and speed they gave us, but those WCP 2 speed trannys are pretty sweet. We had a low speed of 5 (strictly for pushing) and a high of 15 (which I think could have been higher but we wanted a lower low speed). Buying them saves a ton of time (like 3 weeks for us) and a lot of resources. If you are set in your ways, then good luck to you, but I strongly suggest a store-bought transmission.
Yeah, it was far too fast. Our driver couldn’t react in time and drove like he was drunk when in high gear.
That said, it was fun to be able to get across the field in just over two seconds when there was nobody in the way.
I wouldn’t recommend going higher than 18 theoretical.