Epicyclic retro-direct 2 speed transmission

[12p3phPMDC]

Sounds like a custom planetary....nice!!

So, you're able to source the ring, sun and planets.

I need to look up the modulus stuff you've been talking about.

Are you going to use needle bearing on the planet axles?

So, 6:68:1 total reduction for reduction for 1st?
Still working on second eh? ***oops, forgot its direct*** 4:1 for second?

The CSK sprags are very low in cost relative to the others I've looked at.

 

[Miles]

So, you're able to source the ring, sun and planets.
Are you going to use needle bearing on the planet axles?

They are all standard parts. The only thing I'll have to fabricate (for the epicyclic bit) is the plate which goes between the ring gear and the final clutch. Yes, needle bearings on the 4 planets.

The two gear box ratios will be 4:1 (bypassing the epicyclic) and 6.67:1 (including it).

 

[12p3phPMDC]

Cool....

Are there formulas or algorithms that help you decide what gear combinations will work
due to the constraints of an epicyclic? i.e. packaging everything in the ring gear.

Or is it just a matter of lots of iterations...i.e. pick a ring gear, then pick a sun gear,
then hunt for a planet that will fit and give the right "mesh". If no go, then
change the sun by 1 or 2 teeth and try again...etc..etc?

 

[Miles]

Are there formulas or algorithms that help you decide what gear combinations will work
due to the constraints of an epicyclic?

- The number of teeth on the Sun plus twice the number of teeth on the Planets equals the number of teeth on the Annulus.

- The number of teeth on the Annulus and Sun gears together, divided by two, must be an integer.

- To have equally spaced Planets, the number of teeth on the Annulus plus the Sun must be divisible by the number of Planets used.

 

[12p3phPMDC]

Thanks Miles, I mean professor.

Are there formulas or algorithms that help you decide what gear combinations will work
due to the constraints of an epicyclic?

- The number of teeth on the Sun plus twice the number of teeth on the Planets equals the number of teeth on the Annulus.

- The number of teeth on the Annulus and Sun gears together, divided by two, must be an integer.

- To have equally spaced Planets, the number of teeth on the Annulus plus the Sun must be divisible by the number of Planets used.

Rule 1: Ns + 2Np = Na

Since the number of teeth is proportional to diameter, this makes total sense when thinking in terms of circles.

Rule 2: (Na + Ns)/2 => must be integer.

Ok, the integer part I get, because gear teeth are whole numbers....i.e. can't have half a tooth...unless it's a soft cyclone gear :lol:

But, I guess I'm wondering why the planet teeth are irrelevent for both the sizing rules and when it comes to the reduction calculation. I understand the planet is an idler gear...but uh....still confused. :| . Hmm..

Rule 3:
(Ns + Na)/4 (4 for your example) = a whole number with no remainder.

I can understand the even divisions of the annulus or the sun independently....So, just adding them together meets the constraints...?

Well, even without understanding everything, the rules are simple and elegant.
8)

 

[gogo]

But, I guess I'm wondering why the planet teeth are irrelevent for both the sizing rules and when it comes to the reduction calculation.

I think of the planetary gears to be serving the same function as a chain in between two sprockets.

 

[Miles]

Hi 12p,

Imagine turning the Ring gear inside out and running it on its own shaft, engaged with one of the Planets. The total ratio will stay the same, the only thing that will change is the output rotation direction. If you lock the Planet carrier, things are very simple.

Does this help?

 

[12p3phPMDC]

Ok, so with the planet carrier locked and the planet gears as intermediate idlers,
the ring driving a planet gear by itself would give an increase or overdrive...(big gear drives small gear).
But then when the planet gear drives the sun gear, it provides an reduction or underdrive.
And when you put the whole thing together the planet diameter/teeth cancel out.

Na/Np * Np/Ns = Na/Ns

Ok, so I get the idea with the planet locked, but am not fully understanding why ( Na + Ns )/2 must result in an integer.

I can understand the integer requirement... but I guess I don't understand why just adding the Na and Ns then dividing by two solves it. How is that derived....?

 

[HAL9000v2.0]

Miles, are you going to wheel on "other" side or to cranks?

 

[Miles]

12p,

( Da - Ds ) / 2 = Dp

If the difference between Da and Ds isn't exactly divisible by 2, then Dp won't be a whole number

Another way:

The number of teeth in the Annulus and in the Sun must both be odd numbers, or both be even numbers - otherwise, the difference won't be exactly divisible by two to give a planet gears with a whole number of teeth.

 

[Miles]

Miles, are you going to wheel on "other" side or to cranks?

Hi HAL,

I don't have a particular build in mind, for this, but I kind of envisaged it for driving the left side of the wheel.

 

[12p3phPMDC]

Ahh....Thanks Miles...

 

[Miles]

Revised spec. :

1st stage 18t to 72t 5mm pitch GT3 synchro belt.

Epicyclic: Sun 42t, Annulus 84t, (-2:1) Planets 21t (3 off), Modulus 1 pitch.

Sprag clutches: CSK 17 (2 off) CSK 20 (1 off ).

1st gear 8:1

2nd gear 4:1

I'm not sure that I should really call this an epicyclic transmission. With a permanently fixed planet carrier, there is no epicycle.... You are using the planetary gears for the reversing and the revertive functions, only.

 

[EVnewbie]

An 8:1 and a 4:1 sounds great
Looking forward to this tranny coming out, it will solve many problems I am having.

 

[Miles]

Here's a teaser

 

[12p3phPMDC]

At first, I thought you already had it built, then I realized it's model.

Nice model! Sounds like you're getting close.

 

[HAL9000v2.0]

nice 1, miles

Alibre?

 

[Miles]

Alibre?

Modelled in Alibre, rendered with HyperShot (first time I've used HyperShot.)

 

[HAL9000v2.0]

first time I've used HyperShot.

jea, right. 8)

 

[EVnewbie]

Miles,
Is it in your plans to make one that has a larger mounting surface to be used with different and larger motors? I was thinking the Currie/Cyclone style of motors would work well with your transmission.

 

[Miles]

EVnewbie,

I don't have any plans, that far ahead, yet...... Let's see how the prototype works first

 

[12p3phPMDC]

Miles,

just for reference, do you mind sharing the supplier of the gears you are planning on using?
Are they hobbed, sintered, hardened etc...
Can Joe-schmoe buy small quantities?

 

[Miles]

12p,

SDP/SI have a good selection in both Diametral and Module pitches: http://www.sdp-si.com/eStore/CoverPg/Gears.htm

Here's another source: http://www.hpcgears.com/

I was going to get everything case-hardened after re-machining.

 

[12p3phPMDC]

http://www.sdp-si.com/D190/HTML/D190T109.htm

Nice explanation of Diametral (American standard) and module pitches (Metric).

It looks like hpc will case harden stuff. Cool!

Thanks!

 

[comradegerry]

Hi Miles

Came across this interesting piece of hardware, HammerSchmidt made by SRAM. The gearing is too low for what I had in mind but if the peddles /arms were discarded and it was modded to be just part of a transmision with the thumb control between low and high gear it might have some use on an ebike? The best of luck on your engineering project, it sounds great, look forward to seeing the prototype in action.

http://magicmechanics.com/home.php?lang=en#

Best regards

Gerard