What is Regulator gain?

[Shatter]

Sorry if I am asking a silly question, but what exactly is regulator gain? It is a term at work I keep hearing but I cant place exactly what its supposed to mean. Our product is fuel modules in cars that contain a fuel pressure regulator which opens and dumps excess fuel back into the reservoir, so what would the gain constitute? Thx for any help

 

[BigInch]

If your device functions in the usual "control sense", the Gain is similar to "leverage". The controller for your regulator device takes the difference between the controller's input signal (say that would be the actual fuel pressure discharged from the regulator) and the controller's set point (which would be the desired regulator discharge pressure). The controller determines relatively how far away the discharge pressure is from the set pressure and calculates a correcting signal (probably) based on the following formula,

Output signal = Gain * ( E + Td + 1/Ti [∫] E(t)dt ) + Vs

E = Error = Input Signal - Setpoint
Ti = Reset Time
Td = derivative Time
Vs = Bias

The controller then sends the new output signal to the regulator and the regulator moves to a corresponding position.

The output signal strength is effectively the error signal multiplied by the gain setting, so in combination with the derivative and reset times, gain determines both how quickly and how "hard" the regulator responds to any differences between input signal and set point.

BigInch-born in the trenches.

http://virtualpipeline.spaces.msn.com

 

[Shatter]

Thank you for the reply, thats about what I figured it was but I still am gray in 2 things. First off these regulators do not use a controller, like in the electrcial sense, they are mechanical. The regulator operates via pressure in the system, thats what makes it open. When the 400kPa system pressure gets too high the regulators are set to open at 400kPa. The excess is then dumped back into a reservoir so it can be re-used. Based on your description, I still dont think im sure I understand what "Gain" is. You said its like leverage, can you elaborate? Thanks again

 

[JLSeagull]

Pneumatic controllers included proportional, integral and derivative actions before WWII. The proportional or gain function is a simple straight lined slope relationship between the input action and the output action.

 

[Shatter]

Ok that makes sense, any idea why would gain be important if all the regulator is doing is dumping back fluid to a reservoir?

 

[BigInch]

Gain controlls the swing of the controlling action. If gain is too low, the swing will be too slow and it will take too long for the regulator to reach the set pressure. If gain is set too high, the regulator may overshoot too far and take a long time to stabilize, or may go unstable and over/under shoot for a long time.

BigInch-born in the trenches.

http://virtualpipeline.spaces.msn.com

 

[dcasto]

The gain in a mechanical spring regulator would be set by the orifice size that balances the forces of the spring?

 

[iainuts]

I've not heard of the term "gain" before when used in conjunction with a mechanical pressure regulator.

Mechanical pressure regulators have some amount of 'unbalance' built into them however, which is a function of the poppet seal area, diaphram or sensing area and spring rate. As inlet pressure drops, the outlet 'set pressure' (or the pressure the regulator will hold under constant flow conditions) will increase. This is due to unbalanced forces on the poppet which are opposed by the spring and sensing element.

Do you think this is what is being refered to?

 

[JLSeagull]

The term droop is often applied to the inlet to outlet characteristics for self-contained pressure regulators.

 

[Shatter]

In this sytem in particular, the regulator is simply dumping fuel back into a reservoir, the "flow" coming from the output of the regulator is meaningless really. In a fuel system on a vehicle the regulator monitors mechanically the pressure in the system, once the pressure exceeds 400kPa the regulator opens and dumps the excess fuel. Would the gain be somehow tied to the pump in the system?

 

[BigInch]

Just substitute the pressure signal for the flow signal I gave in my example. The analogy is valid for most control systems based on flow, pressure, temperature, light intensity, whatever. A signal is just a signal from one kind of any of a number of different types of transducers.

BigInch-born in the trenches.

http://virtualpipeline.spaces.msn.com

 

[Compositepro]

For a self contained spring loaded regulator, which is what Shatter appears to have, the term gain is rarely used. The term droop is commonly used. Gain is usually used by control engineers and droop by most others. But, control engineers aren't often need where self contained regulators are applied.
In any case the terms are simply the inverse of each other. Gain is the amount of flow through the regulator devided by the error (set pressure-actual pressure). Droop is the error devided by the flow. A larger valve orifice and longer spring (lower spring constant) give you more gain.

 

[iainuts]

Just to clarify what droop is. From the Fisher Regulator web site:

14. Droop is the reduction of outlet pressure experienced by pressure-reducing regulators as the flow rate increases. It is stated as a percent, in inches of water column (mbar) or in pounds per square inch (bar) and indicates the difference between the outlet pressure setting made at low flow rates and the actual outlet pressure at the published maximum flow rate. Droop is also called offset or proportional band.

Ref:

http://www.fisherregulators.com/technical/tips/

Consider the forces pushing 'up' on a regulator's various internal parts.
1. Upstream pressure times poppet seal area is pushing "up". [1]
2. Downstream pressure times diaphragm/sensing element area also pushes up.
3. Reference force from spring or gas in the case of a dome loaded reg is pushing "down". Note that in the case of a spring, the force down is a function of poppet movement and spring rate. F = k x

Let's disregard changes in upstream pressure for a moment. Downstream pressure pushing up on the diaphram then balances the spring load pushing down. If the poppet is open some amount (dx) and downstream pressure drops, the force up on the diaphram drops and the spring extends which pushes open the poppet to increase flow. But the down force due to the spring is a function of spring rate, so the new opening on the reg is larger but the pressure is necessarily lower by the amount dP = k dx / A. This phenomenon of pressure droping due to the regulator needing to increase flow (increase poppet lift) is known as droop.

Further down on the same page, I see they mention "gain".

26. Cutting the supply pressure to a pilot-operated regulator reduces the regulator gain or sensitivity and, thus, may improve regulator stability. (This can only be used with two path control.)

So it seems "gain" in this case is being used to indicate the restriction of gas flow to the dome of a pilot operated regulator. I've not heard of this being called "gain" before, but I can understand why after reading that.

I used a spring in the above example regarding droop, but a dome loaded regulator (being called a "pilot operated regulator" by Fisher) is very similar. In the case of a dome loaded reg, the change in dome pressure is also a linear function of sensing element movement. If dome volume is large, relative spring rate is low and droop is minimized. The use of a dome loaded regulator is commonly used where very accurate pressure control is needed over a wide range of flow due to the droop issue. The problem is often that the reg becomes unstable with a very small spring rate.

The way I've seen that typically controlled is isolate the sensing element from downstream pressure and putting an orifice in line. This reduces the volume and increases spring rate, but also allows sensing element pressure and downstream pressure to slowly equilize as gas flows through the orifice which minimizes droop.

Another option is to have a small dome such that the relative spring rate is high. But this means droop is high. Not a very good option.

Another option is to have a small volume on top of the sensing element which has some restriction to flow similar to orificing the sensing pressure. Apparantly, Fischer is calling this "gain". Not sure from reading their definition whether or not increased gain is increased restriction to flow or not though.

~

[1] Note that a balanced poppet minimizes this area and helps reduce set pressure change as upsream pressure changes, but let's disregard this for now.

 

[BigInch]

Quite interesting, but I still see the underlying fact that gain is gain, whether it is accomplished via electronic amplification or by a lever, no matter what namne is preferred amongst the various system types.

BigInch-born in the trenches.

http://virtualpipeline.spaces.msn.com