How do you calculate Gear Ratio?

Hi everyone,
I was wondering, how other teams go about calculating the perfect gear ratio for your robot mechanisms.
What factors does your team consider when determining the ideal gear ratio?
Have you encountered any unexpected challenges related to gear ratios, and how did you overcome them?
Are you using any kind of online calculates (like JVN) or do you calculate it manually?
Personally, my team usually uses online calculators but if your team usually calculates Gear Ratio manually, I would love to know about the process and what formulas are involved!
And if your team also uses online calculators do you have any favorite tools or software that assist in gear ratio calculations?

  • Thanks, Guy from team D-Bug 3316.
1 Like

I start with JVN as a ā€˜quickā€™ guess. Recalc and other models are more accurate, and I double check with those.

We frankly always start very conservative. We generally start at a higher ratio that moves more slowly than is ideal, so that we know it will move consistently and ā€œjust workā€. Later, we go back and iterate to a faster ratio. Usually after the first competition. We do this because we usually play week 1 if possible, and winning in week one usually comes down to having a machine that plays effectively. Not necessarily quickly.

Our second comp is usually week 4 or 5. That one, youā€™ve got to be hustling to win.

Itā€™s funny, but we tailor some of our design based on when we are playing. Early on you can play it safe. Later on you better be pushing your motors to the limit.

5 Likes

Typically I prefer using dimensional analysis (conversion between different units), coupled with some basic algebra, and occasionally some basic rotational physics. Otherwise, try using design calculators such as JVN, AMB, or ReCalc. Design Calculators are often good tools for finding a range of ratios that should work, but Iā€™d suggest double checking by hand after the fact. This is mostly because JVN (in my experience) only accounts for motor specs at more-or-less full speed. To better explain what why this is important, check out the documentation pages on motor curves (WCP, REV, VEX, etc.). By finding target amounts of current draw and speed, you can work the math based on values not always considered in the motor data tables utilized by design calculators. If you choose to iterate based on an over-geared mechanism that you believe could be improved in terms of speed, I would definitely consider doing that math manually. Hope this helps!

(NOTE: Just because something ā€œworksā€ in a calculator does NOT mean it will work in real life)

1 Like

When we designed and built our own WCD gearboxes (an era I hope is over), we would start with one of the calculators giving us our ā€œoptimumā€ ratio for our expected cycle sprint distance. By designing our gearboxes with sufficient space between gears and standoffs, we knew that if it didnā€™t actually turn out to be the ideal ratio in the real world, that we could always easily swap gears in the second stage. Say our second stage started out 30:30, then 32:28, 34:26, 28:32, 26:34 etc would all also fit.

There is seldom an ā€œidealā€ ratio. It always a trade of different needs. Say for an elevator. The ratio is going to need to travel its working range in the speed required. It will have to be low enough to have sufficient torque. If you use the stall of the motor to hold the elevator position, The required stall current will have to be low enough not to burn out the motor. Many elevators will require 2 motors to meet all of the requirements.

If you are using planetary gears it is generally easy to swap gear set to change the ratio. To a lesser degree with chains and belts.

It really depends what the ratio is for. Weā€™ve used the old JVN spreadsheet, and its variants, as well as recalc to get good estimates for what we want. Iā€™ve used the spreadsheet calculators more for predictions on total current draw than for speeds.

For a ball shooter, you can use your wheel/roller radius and desired exit velocity to calculate an RPM. Then from that, choose a gear ratio that will allow your motor(s) to be faster than the desired RPM, and control it with encoders and code.

For a climber in 2017 and 2022 we did the physics calculations first and then put the gear ratios into the spreadsheet to see if we would be tripping breakers.

The AMB mechanism calculator has documentation that explains how all of the gear calculations work. If you want to do the calculations by hand, you should be able to using those equations. There are also a number of examples on the page for how to decide what gear ratio is right for different applications. If you have any questions about how the equations work, feel free to send me a DM here or you can get my WhatsApp from your mentors.

3 Likes

This is a really excellent point. Iā€™d have to go back and check the math on teeth and pitch diameters, but is it really as simply as drop 2 teeth from one and add 2 teeth to another?

I bet a lot of people donā€™t realize that.

1 Like

Yes, the center-to-center distance for a gear pair is the sum of the teeth between the two gears divided by twice the diametrical pitch (dp). So you can change the gear ratio freely without changing the center distance so long as you shrink one gear by the same number of teeth as you enlarge the other. My reduction swap calculator works essentially on this principle, finding possible gear/belt/chain reductions with a different ratio that fit in the same c-c distance.

2 Likes

So in other words the tooth count needs to stay the same for the same CC distance.

This works except in the case the some one it messing with the DP of one of the gears. Which is how you get small ratio changes by changing one gear and not the CC distance.

ā€¦ And this is a never ending source of confusion on this forum come build season as teams use whatever-ideal-ctc-table from the Internet vs looking at the product page.

I like this video from Citrus Circuits 2016 Fall Workshops - Motor Sizing, https://youtu.be/U4pgviiEwLg?si=cjfH7sYEEtvKCcC4

1 Like

This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.