This is a personally financed effort, so I don't have an aerospace sized budget, but my budget is non zero.

I'm looking for advice on valves, my Servo driven ball valves have too much slop/hysteresis.

I'm looking for a valve:
Liquid oxygen compatible.
CV varies linearly from .13 to 3.0
It needs to shut off bubble tight , but can be very non-linear from off to a CV of .13.
Max pressure will be 300PSI
Safety factor of 2
Our test facility has a nice bunker ,the valve will not be pressurized with humans in the blast radius.
As light as possible.

I don't care how long it takes to get from off to .13, but from .13 to 3 I need it to make +/-10% steps in 50msec with very little hysteresis.

I currently have a solution using a 3/8" servo driven ball valve followed by a faster 3/8 butterfly with the seals removed. It weighs about 1Kg.
The current issues is that the 3 peice ball valves seem to leak after 30 or so temp cycles, and combination of multiple
valves is larger more fragile and less reliable that I'd like.

I'm going to test a 3/8" swageloc plug valve, and might possibly machine a new plug for that with triangular orifices for better throttling.  I've also been advised that the ideal solution would be a pressure balanced poppet valve driven by a voice coil arrangement, but that seems like a lot of development.

Any suggestions of valve experts?

It sounds as if the seals are not the ones intended for Cryo, or if they are automating a valve that was originally for manual service.  

Worcester cryo seats are configured somewhat differently than seats for standard service, to allow for the not-insignificant dimensional change. There are other mods to the seat as well.   Also Worcester valves for modulating have a high-cycle packing stack with extra seal rings and 4 belleville washers (instead of just 2) to compensate for wear and thermal changes.  

I was involved with some test-stand globe valves once where NASA stipulated that they wanted 20 hz response(!). The testing procedure assumed that the 20 Hz was continuous.  It was possible to do this with an electrohydraulic actuator, but the valve packing melted. With LO2 in the valve, it's a good thing the packing did not ignite!.  

I know at least one Rocket group using Worchester Ball valves
is looking to switch. They have a really fast custom actuator and are wearing the seals out with the rapid movement.

For straight on-off I'm sure these are fine, but the xprize competitors need to rapidly throttle the motors for hovering and that seems to wear out the seals.

MORE NEWS