Is this an improvement?

There are four lantern pinions in my skeleton clock. The pins were a press fit of steel into brass, and unlikely to work loose.

Although it’s highly unlikely I’ll make any for reasons I’ve mentioned on previous occasions, I’ve pondered about the benefits of trunnion-style pins as shown in this illustration.

The pins could roll over wheel teeth instead of sliding with the obvious benefits and disadvantages …e.g. assembly of the pinion, lubricant spreading to the wheels, etc.

Any thoughts?

Sam

Looks impressive but difficult to make, particularly in small sizes. I also wonder if the rollers would actually turn easily on the pins ?

IIRC I think Harrison used rotating trunnions so you'd be in good company.

That’s typical of those old blokes, John.

Leaving us nothing to get our teeth into.

Ifoggy - Judging from your excellent work, I'd reckon making one would be a cinch for you.

Sam

I am just wondering if the surface contact area between a wheel tooth and a roller would be sufficient to overcome the friction between the roller and its pin, so that it will revolve if you see what I mean.

It would be an interesting experiment to prove it either way though.

Phil

I agree Phil.

Some sort of torque/friction relationship?

I’m beginning to think in addition, that even the tiniest amount of lubricant needed for each end of the pins (even minimal), might eventually migrate onto the wheel teeth.

Are there dry (or non-migratory) lubricants suitable for clocks?

Sam

Sam

As you suspect with clocks being low powered devices you need to worry about drive energy rather than simply thinking about friction.

If you have a plain bored roller on a plain pin the bearing surface area will be large correspondingly multiplying the actual friction. Probably need to reduce the plain pin to a set of ridges, perhaps three, to cut down the bearing area and drag.

The roller won't be turning with constant velocity either. It will speed up and slow down as the gear tooth moves past it. Acceleration and deceleration takes energy. The larger the roller the greater the demand.

Maybe plain pins in jewelled pivots at each end would run easier. Good luck with making that!

Although lantern pinions are known to be somewhat inefficient compared to close tolerance involute gear pairs the actual effect is strongly influenced by size. Small ones wit appropriately shaped gears can be pretty good. As everything is rotating the contact between gear and pinion is not simple sliding.

Clive

.

Lignum vitae [self lubricating] on brass pins ...

See detail at about 2m:20s here: **LINK**

https://www.youtube.com/watch?v=sBKMTKl0wkY

MichaelG.

.

P.S. ... Igus Drylin might be a reasonable modern substitute.

Edited By Michael Gilligan on 19/02/2020 01:06:34

No need for a substitute it can still be bought new (FSC cert) or cut form old bowling balls.

.

Yes, I realise that, Jason

But I thought ‘engineers’ might want to consider an alternative.

MichaelG.

There's something I don't get about lantern pinions. AFAIK they mesh with standard "cycloidal" wheels. Now, "normal" cycloidal pinions actually have the only true cycloid in the gear, the straight radial flank of the dedendum. The addendum of the wheel, which is a circular arc, engages with the straight pinion flank when on the line of centres and stays engaged until the next wheel tooth engages. The pinion addendum is also a circle, but shouldn't do anything and not touch the wheel - it's just there to keep the pinion tooth out of the way.

So what is the logic of replacing the straight cycloidal dedendum with a circular pin, and why use the same wheel tooth form which is normally expected to "roll" on the straight part of the pinion tooth? I see that for example Thorntons, doyens of clock tooth cutters, don't have a version for lantern pinions.

To the best of my [limited] knowledge and understanding, John : The mesh of a lantern pinion with the standard cycloidal wheel is geometrically correct, so there is no need for a different wheel cutter profile.

MichaelG.

.

P.S. 

I have [literally, in the last few seconds] just found this:

https://www.tec-science.com/mechanical-power-transmission/cycloidal-gear/lantern-pinion/

Edited By Michael Gilligan on 19/02/2020 09:18:50

The reason for that is that lantern trunnions are just made from the appropriate guage of pivot steel, not cut like regular pinions.

Russell

Yes. As drawn it can't be made. One flange would have to be removable. Lubricating when sevicing is needed would be a problem as one flange will be adjacent to the wheel.

Russell

Thanks for the link Michael, but it rather proves the point I think. The pinion expects the addendum curve of the wheel to be a cycloid with a generating circle specific to the lantern pinion. But clock gear cutters such as Thornton's have circular addenda with fixed radius for a given number of teeth and expect these to roll on straight radial pinion flanks. Also with non-rolling trundles there seems to be no point in using a lantern pinion at all (other than ease of manufacture) because you still have single point sliding rather than rolling.

I meant a wheel with a tooth profile to mesh with a lantern pinion.

Not being a clocksmith myself, and therefore possibly ignorant, but what's the difference between this and the relationship between a roller chain and sprocket?

Before moving on to the now, quite clearly, ancient trunnion concept, it seems appropriate to reiterate a (perhaps) minor issue. This cropped up towards the middle or final exercise of meshing (read messing) with cycloidal clock teeth in a CAD simulation. With an appropriate warning, the extent of my verbosity can be unearthed here …

Would you mesh with this? Some thoughts on cycloidal clock teeth

**LINK**

Meshing with Lantern pinions - Part A Another slice of CAD, and how cycloidal clock teeth might mesh.

**LINK**

Meshing with lantern pinions - Part B Due to my zealous output, I have been 'advised' to split the original thread.

**LINK**

continued ...

Back shortly,

Sam

... continued

My back-step is in regards to velocity ratio. I had become rather tired of attending to all the manual (CAD) iterations and abandoned the exercise when I got around to messing with lantern pinions.

In that case, I retained the number of teeth, the lantern pinion having 8 (eight) pins (leaves) meshing with the John Wilding’s 290 tooth large wheel. The teeth of the latter being of conventionally accepted cycloidal design.

Consequently, I can only show what was happening where Mr Wilding’s pinion was cycloidal not a round pin.

The graph shows that the velocity ratio (theoretically 290/8 = 36.25:1), gradually increases to almost 40:1 as a single (large wheel) tooth makes its transition. Then, as the ratio gradually decreases, surprise, surprise, the ration drops (from 38.77:1 to almost 36.25:1) as the next pair of teeth/pins meet.

I'm obliged to split this thread a second time ...

Continued ...

Tell yah wot ... split three times ... it's taken me twenty-five minutes to get this b******y graph into the thread.

Cobbled in Excel, it went into MS Word where I prefer to edit, and then into Photoshop, before becoming a .jpg file and palatable for this forum.

Now, if you haven't lost interest, you'll have to go back to the text on the previous page. EDIT Now on same page,( Jason)

Sam

Edited By Sam Stones on 20/02/2020 01:23:49

Edited By JasonB on 20/02/2020 08:01:49