Factors that allow regulator input pressure to affect output pressure

[cloverleaf]

We all know that in an ideal world a pressure regulator should output a consistent pressure; regardless of input pressure.

So in the example of an air rifle reg system with an output set at 100bar, the reg should consistently produce this pressure whether the gun's cylinder pressure is at 100bar or 250bar.

In reality as we know, regs like everything else are not perfect and invariably their output pressure varies to some extent throughout their input operating pressure range. A practical example of this is an HW100 I recently set up. Reg output pressure was set at about 93ish bar in the middle of the input / cylinder pressure range (around 145bar) however with the cylinder filled to the maximum of 200bar the reg output was more like 96bar, while at a cylinder pressure of 100bar the reg output had fallen to around 88bar - a fall of around 8 bar or 8% in output pressure for a decrease of 100bar or 50% in input pressure.

I won't go into why this is a bad thing (I'm sure we are all aware of how this behaviour could be detrimental to rifle performance); suffice to say our rifles will perform better the more stable the reg output pressure.

The point of this post is to ask those with more experience / aptitude than myself, what features of a regulator determine the extent to which input pressure affects output pressure?

I know that on moving piston regs (like the one illustrated below) the ratio between the areas of the pistons on input and output sides of the reg directly affect the influence of inlet pressure on output pressure. This is pretty simple to model mathematically (I have a spreadsheet ) and understand.




What confuses me is the cause of "inlet pressure effect" in regs of other designs - most specifically (but not limited to) that used in the Weihrauch HW100; a diagram of the older version of which can be seen below:



In this design the air pressure is only acting on the reg piston in one direction (as opposed to both sides as with the moving piston design); against the spring stack. From a static perspective, I can see no way how changing input pressure can alter the output pressure.

I'm thinking the inlet pressure effect on this design is probably due to a lag in the valve closing completely (due to "stiction" in the O-rings and the time taken to accelerate the piston the small distance it needs to travel); this lag allowing the admission of a little more air when the charge is at a higher pressure - being more dense and with more force behind it to to push it through tiny gaps in the not-quite-closed inlet valve.

To the wise, does the above sound plausible and am I missing any issues that can be modelled as a static problem; taking into account only forces, pressures and areas? Obviously these are far easier to understand and quantify than flow rates through tiny gaps or accelerations of masses against springs due to varying pressure loading..

Finally if the issues with this reg are due to seal friction, piston mass and increased flow rate through the inlet valve at higher pressures; what are the best ways to combat them? Low friction (PTFE) seals, lightweight reg components and a massive restriction at the entry of the reg to slow the filling process?

Thanks - I very much look forward to hearing what the good members of the board have got to say on this matter 

Cheers,

Mike

[rsterne]

The inlet pressure is bearing on the ball which rests against the seat.... As the pressure decreases, the force required to unseat the ball and open the inlet decreases.... However, we are not concerned with OPENING the valve, which happens instantaneously upon firing from the force of the spring stack, but how the pressure builds on closing as the outlet pressure nears and then reaches the setpoint.... Until that moment, there is little force on the ball, depending on the clearance in the hole it is in and how far off the seat, of course.... Therefore the predominant force is, as you say, the outlet pressure closing the valve vs. the spring pressure opening it.... Depending on the actual diameters of everything, while the ball is off the seat, in theory the valve should close at the same outlet pressure regardless of inlet pressure.... In practice, due to piston mass, the momentum of the air passing the ball and seat, the resistance to that movement, plus sticktion in the seals, literally any small variations may be possible from that ideal....

Fortunately, none of this matters if the gun is properly tuned.... If the gun is tuned to the "knee" of the velocity vs. hammer strike curve the gun ends up operating on the plateau of what would be the bell curve in an unregulated gun.... As an 8% change in pressure won't effect the velocity hardly at all, it won't matter if the reg. output pressure changes slightly....

My apologies for the number of times I rewrote this post, it was more complex than I originally thought....

Bob

[Bill G]

The second pic is of a externally adjustable reg I suppose?  That is interesting and would be handy.  It would seem that this manner of regulation wold be easier to manage, more straight forward so to speak. any opinions to this?

Bill

[Motorhead]

Pending SEAT design used and if or not the HP side bears against the spools HP stem once seats closed makes a difference !!
If HP side of spool sees ACTUAL HP storage pressure once seat is closed, the area of the HP side of spool bears against the spring stack changing its resistance value.
As HP pressure changes so does the force exerted against the spring stack.

In a Reg where the seat is such incoming HP pressure gets shut off by the seat closing, spools HP end is at the same regulated pressure as the LP side.  So no value of exerted pressure on spring stack happens and reg does not creep at HP storage pressure changes.

Regulators such as the NINJA take HP air in and have there seat on end of the spool.  As air is allowed to transfer from HP to LP side it passes over the end of spool entering a side hole that intersects a center hole down length of spool to the LP end.
* When seat closes HP air is shut off right on the end of HP spools end using a small diameter seat.  Upon seat closing ALL THE AIR surrounding spool is at the LP set point.

Regulators that BRING HP air into the space ALONG SIDE the HP end of spool ( Side intake types ) have an HP air path that always has WHATEVER THE HP PRESSURE ACTUALLY IS surrounding the HP end of spool because the seat which is also located on HP end of spool only controls opening and closing a straight passage LP to HP side down length of the spool.

** So the AREA of the HP end of spool will indeed exert differing amounts of push against the spring stack which governs seat open/close values ( Set Point ) 

Hope this is making sense

[rsterne]

I agree with what Scott says about the Ninja regulators that once the seat has closed the air in the HP chamber and the LP chamber is at the same pressure.... However, the Ninja style regulator is still subject to some creep (higher output pressure when the tank is full)....The seat on the end of the piston is subject to HP air even when closed, and although small in diameter, takes additional pressure on the output side to overcome it.... There is also likely a slight time lag between the piston starting to move to close the seat and it closing fully.... During that time, additional HP air flows to the output side.... While these effects are much smaller than a side port HP side design (as pictured in the first diagram in this post) they will, and do, cause a slightly higher output pressure when the tank is full than when it is empty.... It can be measured with a gauge, and it is enough to cause a slight velocity change in the shots between a full tank and one at the setpoint pressure, particularly if the hammer strike is either much too strong (velocity decreases as tank pressure drops) or much too light (velocity increases as tank pressure drops)....

As I said before, if the gun is tuned properly, to operate on the "knee" of the curve, any regulator creep won't matter because the gun is operating right in what would be the "sweet spot"....

Bob