Although Vex has its own differential, the bevel gears always wear out when I use it. So I made a straight-tooth differential myself.
Since many of us (including me) responded to your poll with: āI donāt know how straight teeth diff work but I know what is a diff,ā it would be helpful if you could explain how the mechanism works (especially if the image you provided is from a 3-D model you made/ have access to).
I know we could all individually scour the internet for an explanation, but it would be even better for as detailed of an explanation as possible to be here for future reference, especially for users who might find this topic through the search feature later on.
The easiest way to understand how a differential works is to hold the carrier (the piece that holds all the gears and spins around) still and look only at the two output shafts.
A āstandardā bevel-gear differential relies on the fact that two bevel gears on opposite sides of a single bevel gear can only spin in opposite directions.
A straight-gear differential relies on a rather clever trick: each of the ācenterā straight gears mesh with the other ācenterā gear and ONLY ONE output shaft.
If you hold the carrier still and turn the ātopā shaft clockwise (looking down from the top), the following will happen:
If there was a single gear connecting the output shafts, they rotate in the same direction if the carrier is held still (exactly wrong for a differential).
The fact that there are 2 meshing gears, each connected to only one of the output shafts forces the shafts to rotate only in the opposite direction if the carrier is held still, serving as a differential.
Very clever use of standard and high-strength gears, @Lao6.
When you mean only one, do you mean only one of the two output shafts on both sides of the strraight teeth differential?
On another note, how did they get a patent for a design that has existed since 1911? It doesnāt seem that different from the patent fundamentally.
After reading your post and doing a bit of digging online, I realized that I was for some reason operating under the misconception that there wouldnāt be planetary motion of the whole mechanism in this design. This video helped me realize this visually (see the end - where itās spun while in a carrier). I now understand how the system works (great work @Lao6), so if anyone is still confused, I recommend giving the aforementioned video a watch-through, especially if you need to see it in motion to understand it.
It doesnāt different too much, but it is different.
Can you say this again and use different words, Iām confused.
I freehanded these a few years ago, so thereās no CAD:
IIRC I was intending my next step to be to add strength in torsion between the framing gears to stop the axles gripping in the bearings. It was sticky despite all the teflon washers.
It didnāt see action on a robot - I was just playing with ideas to work into mentoring. Youāve flagged some of the basics to some extent, and thereās more or different design work to do before my version is any practical use, I think.
What in the worldā¦ I meant āit isnāt too different, but it is different enoughā.
English is hardā¦
'Tis hard, using the words, and finding the phrase that makes the words work together is hard, words on their own can mean one thing, words working together can mean something else"
Words are always about communication.
Thanks ā¦
how straight teeth diff work (perhaps)
Yep, thatās one approach.
That is seriously impressive! I would assume that white lithium grease will most likely be a requirement for that system to operate well.
Iām glad you like it 
More generally to the crowd:
Read @teresajude 's explanation.
The wikipedia entry for differential gears is pretty good Differential (mechanical device) - Wikipedia
Watch the .mov @Lao6 found above.
Hereās a more exposed view of the spur gear motions in my CAD. Iāve included a clue pointing to a potential alternative scheme for mounting the pinions.
My inputs arenāt answers to the problem - theyāre just illustration of stuff that can happen, and you need to find what works for you.
A lot of folk seem to ask for the ābestā way to do something, whereas - especially in competition - you need to find something good for you by trying alternatives and variants. The importance of fumbling around in the dark a bit canāt be overstated. That you donāt have the answers yet means that you canāt see them yet, and when you do see them, your first impressions will be incomplete. Whereās Wally? How many wallies are there? You donāt know yet.
There isnāt even an undisputed, universally best way to move sideways, for example.
And donāt forget there are actual rotational bearings now. Theyāre quite chunky, but a diff can be as big as it needs to be.
Graphite (pencil lead) works wonders and is FAR easier/cleaner to deal with. Or use powdered graphite if you must.
Be aware that pencil lead contains abrasive clay that will help rotating elements grind through their mountings. As for graphite, consider that conductive particulate matter is bad news for connectors and electronics.
A tiny dab of lithium grease that escapes will just join the general sea of finger grease already covering your robot and its wheels. Clean your robot, and especially wheels/rollers and surfaces they have to climb on the field.
Differentials have things spinning within a spinning thing and so can distribute loose material far and wide.
Build to run as freely as possible without grease. Grease in the amounts allowed by the rules doesnāt create freedom of motion, it takes the edge off friction.
Think about how fast or slow various parts of the differential could rotate in relation to the mechanisms theyāre attached to - you get to choose/design those speeds.
Couple of comments, to dive deeper into this.
First, I had forgotten that pencil lead does contain a clay binder.
Second, Iām not an abrasive expert, but from using pencil lead in a bot that sustained a lot of use and heavy loadsā¦ there was no noticeable wear/polishing on the aluminum (was used on a sliding mechanism), so by gut said that itās not a problemā¦ at least under limited use that a robot will see. I probably wouldnāt use it in industry for a long time.
Asking chat gptā¦ and FWIW, hereās what it said.
Powdered graphite is presumably āpureā but a heck of lot harder/messier to work with.
As far as conductivity goesā¦ this would actually be a good thing. Let me dive deeper. Iāve seen some bots have problems with motors on arms/lifts frying. I bought a static electricity meter and learned that bots can generate a LOT of static (I posted on this here).
However, if a team is using plastic/nylon bearings, parts of the bot can be electrically isolate, build up a differential charge, and eventually get enough static charge to fry components. If graphite (or a grounding connection) bridged this gapā¦ it would let the charge bleed off and avoid killing parts. We performed tests on this.
That saidā¦ I wouldnāt want graphite in my brainās connectors.
I misremembered the nature of the clay. Thank you for that correction.
