The Thrifty Bot - Product Releases & Updates

Also I think a bushing that has an ID to fit thunderhex/similar product would be great. There are a lot of applications in FRC that really don’t need a ball bearing, a bushing would be fine. We bought 0.500" ID 0.625" OD flanged MDS-filled nylon bushings and used a 0.543" reamer to open the ID up. We found that thunderhex spun really well in this and we used this on our 2018 intake shafts without issue. A COTS part that does this (with maybe 0.750" OD) would be fantastic.


I prefer both set screws on the same side for maintenance purposes (you can remove the set screws without rotating the shaft, which isn’t always trivial for arms or other large manipulators). You can balance any flywheel shaft using an even number of shaft collars by pointing the screws in opposite directions.


Flanged bearings with 1.125" OD and 13.75mm round ID would be amazing…


i want to know the weather on the day they were cast… and how the casting recovered to cast again.


Were any children forced to pack your products?


My 4 year old is really good at opening the individual bags that the bearings arrive in :slight_smile:

The real MVP here is my wife - a lot of the bearings and collars were grouped together in small bags by her.


I got a quote and it doesn’t add too much to the cost to do this. If there’s enough interest (AKA if more people say they’d buy some in this thread) then I’ll definitely pursue this. Sounds like it may make sense to stock both if there’s demand. Thanks!

1 Like

I’ve had this thought stuck in my head for no real reason for the past day. But you can actually balance a hex shaft with any quantity of shaft collars other than 1.
Evens are obvious (point screws 180deg apart).
Sets of three are also rather easy (point screws 120deg apart).
Any number that is a factor of 2 or 3 can be done with those.
But any other number can be broken down into sets of twos and threes* (2+2+2+2+3=11, for instance). Not that you should ever need that many shaft collars on a shaft, but it’s possible.

Now is this a better practice than using a balanced shaft collar on a high rotational inertia shaft or a reason not to also provide the option to purchase a balanced collar? No, it’s not. But it was just a fun thought experiment I had lurking in my brain.

*I’m not going to a mathematical proof for this, so if someone wants to set aside the time to prove me wrong here, I’m all for it. In fact, I encourage it.

1 Like

What you’ve described is “balanced” only in static conditions, but it’s still dynamically unbalanced.

To make the system dynamically balanced you have to make the center of mass colocated with the center of rotation at every infinitesimal slice of the system along the axis of rotation.

EDIT— actually no that’s a bit too strict of a definition. This shaft is dynamically balanced:

The statement from the article is better:
" For a system to be in complete balance both force and couple polygons should be closed."


We’re getting a limited number of these in the next week or so to make sure the quality is solid on them. Once I validate that I’ll order more.

Taking a stab at this too, quotes are in and may have some up for sale in the next few weeks if the manufacturing process goes well.

Give the people what they want right? What else do people want? :thinking:

  • 16t #25 1/2 hex bore sprockets. We used over 40 of these last year between 2 robots.
  • a tube clamping bearing block made of aluminum. We used 16 of these between 2 robots last year, and another 16 we manufactured ourselves with the bearing standing away from the tube.
  • Metal 1/2 hex hubs—used 24 between 2 robots last year
  • dozens of squishy wheels used each year

Would buy all of these things in bulk if we could


Here’s a proof for those who are interested.

RTP: All integers greater than or equal to 2 can bet expressed as the sum of multiples of 2 and 3. i.e. For all z >= 2, z = 2a + 3b, where a & b >= 0 and are integers.

Consider a parity argument.

Case 1: z is even.
Thus z can be written as

z = 2k, \, {k \geq 1}

Thus it matches the proposed form with a = k, b = 0.

Case 2: z is odd
Thus z can be written as:

\begin{align*} z = &\, 2k + 1, \, {k \geq 1} \\ z = & \, 2k + (-2+2) + 1 \\ z = & \,(2k-2) + (2+1) \\ z = & \,2(k-1) + 3(1) \end{align*}

Thus it matches the proposed form with a = k-1, b =1



Or we can use the Chicken McNugget Theorem :wink:

Essentially for two relatively prime positive integers a and b, the largest number that cannot be expresseed as a linear combination with positive coefficients is a * b - a - b. In this case, a = 2, b = 3, and a * b - a - b = 1. Therefore, every integer greater than one is expressable.



Big shout out to Ryan and The Thriftybot team! We placed an order last week to check out the quality, and we’ll definitely be making more orders throughout the season. The on-line order process was super simple and the items arrived two days later in excellent condition. We really like the split hex collars! They seem rugged, good fit, and solid fastener engagement. I’d definitely recommend these to any teams that use a lot of Hex shaft elements.


Thanks Dave! Glad to hear the products meet your expectations :slight_smile:

Orders shipping after today will have a fun sticker or two in them, for those who enjoy that sort of thing.


Any chance you have a ballpark price estimate, provided they hold up to your standards, and you sell them

1 Like

Looking to price them at $35 each right now assuming everything stays the same.


I received a message asking if Thrifty Bot bearings were similar to Boca hex bearings. I had never actually used or held a Boca hex bearing before, so I bought a few to compare. The Boca hex bearings seem to be very similar to the ‘gen 1’ thinner inner race style hex bearings. You can see our bearing on the left and the Boca bearing on the right below in a side by side comparison -

Also a few updates -

We should be receiving the 4" OD x 3.5" ID swerve bearings on Wednesday. We’ll have a limited quantity to start on this initial run but plan to order more. Lead time is only about a week on them, so don’t worry if you don’t pick any up on this first run. We’ll be getting more assuming quality is good.

The @Nick_Coussens nylon bushing is happening. Once we have them in hand (probably next week sometime) we’ll verify them. If quality is good expect those to go live next week or the following week.


The 4" OD x 3.5" ID x .25" swerve bearings are now in stock. Let me know if you have any questions!


Did you make a typo on the price? This is substantially cheaper than I’ve seen elsewhere.