Er umm - well, one 'lifts' the valve off its seat where the spring has (better - has) held it.

Desmo is about No spring (er maybe weak one for taking up clearances) - no? It 'places' the valve up + down exactamente. And its essence would be / nay is: to permit even higher RPM without fear of float. So it is the very essence of a modern major ge - er Screamer, no?

Maybe my description didn't convey..

Imagine a swaged-on donut about midway along the valve stem. From the side, a (hardened) fork abuts this collar (the donut has a flat face on that side). There's a step in the valve stem dia. upon which the collar sits, in addition to being a tight fit on assembly.

Consequence too, less valve mass overall. Larger valve-guide dia. at bottom, where seat is - smaller at top = just to prevent side wobble. Two guides better than one -- when ya gots such room! 'Austenitic iron' valve seat inserts, to complete the aim for longevity. Many made it over 100K fast miles.

A bearing in middle of rocker of course; the other side of rocker has a spherical socket - to match the ball-end of a double-tapered pushrod. Light alloy rocker with hardened 'business' inserts. *Modern* design - as any Screamer, I assert: Every gram was contemplated, and while "no expense" is alwys an exaggeration - Irving fought against any serious compromises. (An early bronze timing gear was.. finally replaced with lightened, steel one with fancy hardening process).

It is a short pushrod because of the high placement of the cam, in crankcase ~ near the lower visible part of juncture of the cylinder fins. This-all, instead of pushing on the Top of the valve stem, as normally.

Saves height of cyl head too - rocker need not be Above valve. Saves valve guide wear; on Phil Irving's design: these two guides unique. Add-in superior finishing of such sliding surfaces as are unavoidable in any cam/valve assembly.

Or were you just being a smart-ass, odd as that thought is?

OTOH - IIRC the VOC (Owners..) kicked around the natural exploitation of the valve collar approach. One problem re returning valve to seat desmodromically: collar is er 'swaged' onto something which runs as hot as anything (inside) an engine.

You can only pay so much attention to How-tight that collar must remain at all temps, via friction VS simply cracking it from too Much interference fit. 'Close' motion wants to lift up that collar!

Pinning it.. is out for all obvious reasons.. Maybe special forged valve with lumpy stem + split-bearings for the lower guide? Never heard more. Why bother? Well folks in US and UK were pushing it - also in Oz. Alcohol, 11:1 pistons one-piece cranks, etc. to where more rpm could have been useful. Nothing exceeds like excess...


A.

PS there's a Shadow running ~ these parts I hear. So far no spot.