KEY ASPECTS

Being very simple from the conceptual point of view, the stick is often a homemade device; however it is easy to underestimate some key aspects that must be respected:

• there must not be any risk of introducing too high fractions of Oxygen into the compressor (some professional units have feedback controls that automatically stop or adjust the metering valves when the oxygen fraction exceeds a defined limit)

• the composition of the mixture stream must be stable during the whole process (in this issue also the compressor inlet performance is involved)

• the metering inlet valves must have a metering needle with very fine adjustment capabilities

• turbulence in the chamber must ensure a perfect mixing before the intake of the compressor and also before the analyzers

• real-time mixture analysis system must be reliable and accurate ! Even though this statement may seem obvious,  this issue opens up a big debating point. It is not just about the choice of the sensors but also their positioning, calibration, durability, redundancy and so on....  Moreover, when blending Trimix, it is not recommendable to rely on the oxygen sensors only and calculating Helium by difference; in this case a new uncertainty element has been introduced

CONTINUOUS FLOW PROS

The Continuous Flow method has important pros:

• It is not affected by the  Real Gas and temperature issues  (as long as  the cylinder is empty or that the residual mix has the same composition of the target mix)
• It allows you to use almost all of the storage gas
• The compressed mixture is immediately homogeneous

CONTINUOUS FLOW LIMITS

But even some defects, limits and complications ...

You are completely dependent on the reliability of the sensors:

• before (analysis of the residual mix),
• during (during the filling process) ...
• and after (check of the final mixture) !

It's true that the analyzers work in an environment that is not as extreme as in a  rebreather (they work at ambient pressure and low humidity) but they still remain a weak link.
To study the analyzers issue in depth, you can refer to the CCR forums  (Closed Circuit Rebreathers)

• in case of residual mix different from the target one, you still need to then calculate by partial pressures and Real Gas (keeping the temperature stable)  
• unless you have an oil-free compressor, the maximum fraction of oxygen is limited for safety reasons (more in the sections safety precautions and advanced safety precautions
• a stable inlet compressor flow is needed  in order to avoid fluctuations in the composition
• it is impractical to fill a sequence of cylinders with different target mixtures: for each different mix you have to adjust the valves flow until the stream is correct, you then have to wait until the whole system (compressor, piping, whips etc.) is flushed from the previous mix and that the new stream is stable