Looking for ideas, voltage controlled air pressure reg

While there are many pre-built solutions out there if I had the money to throw at it I thought I'd pick a few brains here to see if anyone might know of or could help trigger my mind into figuring out a way to do this inexpensively.

What I need would commonly be known as a voltage controlled air pressure regulator but could simply be a tiny orifice the size of which is varied by a contol voltage.

I've got a 25 psi compressed air supply that I'd like to be able to adjust from between 0 to at least 20 psi on the output by applying a 0 - 10V or so control voltage. The cfm/volume of the output would be extremely low, almost a static value, so a divider setup composed of fixed and variable orifices would work out just fine so long as it doesn't consume (waste) a lot of air in doing so.

The closest thing I've come up with involves a tiny orifice feeding a small storage tank and the smallest automotive fuel inector I can find controlled by PWM as a bleedoff device to keep the pressure in the tank at the desired value.

Although it would work I can see several potential problems with this idea. First of all it might tend to be too noisy for the environment I want to use it in and the fact that most fuel injectors are going to flow a lot of air even at a very minimal duty cycle. I suppose that could be handled by installing another orifice in series with the injector but there's bound to be a better idea. If I knew of a tiny air comressor that would vary its output by the voltage you drove it with that might also work, but at the moment I'm just wanting to see if anyone else may have some thoughts on this.

Googling "voltage controlled pressure regulator" brings up a link that explains some technology, so that could be a good place to start for some ideas.

There are a bunch of 12 volt air compressors that are rated to pump up to 250-300PSI. Very low volume output though. PWM driving the motor, as you said, could work.

I'm aware of those air compressors but I'm dealing with very low pressure that must be controlled to within 10ths of a psi with some accuracy.

Not knowing much about pneumatics this suggestion could be way off the mark, but here goes...

Connect a (stepper?) motor, or maybe a servo, to the lever of an on/off valve, hook that up to a microcontroller along with a pressure sensor . Have the uC vary how much the valve is opened based on how far the current pressure is from the target pressure.

If my reasoning is right that should allow you to control the rate at which a tank is filled with or drained of air. You'll need 2 valves, one to regulate incoming flow and one to bleed off excess air when the target pressure is lowered. You could get away with one bleeding the tank constantly but that would be kinda wasteful.

Essentially you're making a regulator based on a uC controlled feedback loop instead of a mechanical one.

Hey marshallf3,

Got a busted laser printer? or a broken adjustable car seat (ahh the junk yards... can anyone say FREEE?!)

If so you got the right mix on creating a servo that you could fit in on the air pressure regulator that ether came with your compressor or can be picked up for few bucks elsewhere (you know eBay has it all, right?).

Oh and don't worry, as patentable as this is, I won't need any royalties from you brotherman !

DigitalReaper said:

Not knowing much about pneumatics this suggestion could be way off the mark, but here goes...

Connect a (stepper?) motor, or maybe a servo, to the lever of an on/off valve, hook that up to a microcontroller along with a pressure sensor . Have the uC vary how much the valve is opened based on how far the current pressure is from the target pressure.

If my reasoning is right that should allow you to control the rate at which a tank is filled with or drained of air. You'll need 2 valves, one to regulate incoming flow and one to bleed off excess air when the target pressure is lowered. You could get away with one bleeding the tank constantly but that would be kinda wasteful.

Essentially you're making a regulator based on a uC controlled feedback loop instead of a mechanical one.

Click to expand...

Ahhh! He beat me to it! But I got the first dibs on the junkyard electric seat adjustable servos

Oh and you can strip off that large plastic knob off the regulator, file it, and put a cheap metal gear on it (I think most nylon ones would split under a powerful servo that would move your ass back and forth).

marshallf3 said:

I'm aware of those air compressors but I'm dealing with very low pressure that must be controlled to within 10ths of a psi with some accuracy.

Click to expand...

You will need some type of precision regulator for this. a normal air regulator on a compressor is not accurate enough to control a few tenths of a PSI.

I think I'd start like this:



The main reservoir gets filled directly by the compressor via a line dryer (neither shown). The main reservoir can be held at an arbitrary pressure at least several pounds higher than required at the output.

The left side injector is used to fill the intermediate reservoir to a pressure just above what is required as the output pressure; perhaps 1psi or so.

The right injector is the "fine tuning" injector. It's only opened when the output pressure is below the required pressure.

If desired, another injector could be added to the output reservoir to release pressure.

You would need to monitor the pressure in both the intermediate and output reservoirs.

I would use PVC pipe for the reservoirs, perhaps 3" to 6" diameter. The larger your intermediate and output reservoirs are, the finer you can make the output pressure adjustment, but you may experience variations over temperature due to the temperature coefficient of expansion of the PVC material. However, PVC pipe is strong, very inexpensive, easy to work with, and available at any hardware or plumbing supply store.

Far too complicated for what I'm wanting to do.

I'm wanting to control a continuously variable pneumatically controlled valve (<=8 psi = full off, >=14 psi = full on) by the temperature of the chilled water I've got available in the system. I could use temp sensor with a pneumatic output but those are really expensive for one with the ratio of temp to psi change I'd need.

There's two air handlers feeding a certain area of the warehouse and while I can run one with full chilled water to it the other one I can only give a little to, however there are times when my overall chilled water temperature is low enough that I could give them a little more. At present I just adjust any extra I give them by manually adjusting the large butterfly valves to the unit however I get tired of climbing up to where those particular units are located.

What I was going to do, and probably will end up doing, is giving them some extra with the manual valves but have an electronic signal shut off the air to the pneumatic valve when the chilled water temperature exceeds 50*F. If I let the system get much above that temperture some of the other more critical areas throughout the building aren't able to cool as well as they should and, as you can imagine, the CEO's office takes priority over the warehouse.

I do have a friend at Johnson Control who may have a simple solution, sometimes he'll give me parts to play with as a lot of what I've done to this antique system has already fascinated him. He's a pnuematic control expert, I mix some electronics in with mine.

I'm thinking that due to the low pressure and virtually no volume situation I could weld a fuel injector on an old propane tank and get away with that. I'd feed it through a set size orifice and bleed it with the injector hooked in series with a slightly larger orifice. This way if the injector was fully open I'd drop and stabilize below 8 psi; if the injector were fully closed I'd stabilize at full supply pressure.

By low volume I'm talking orifice sizes of 0.005" and 0.008".

That is another idea to look into, thanks.

Sounds like you could just hook up a (probably fairly powerful) servo/motor to the large butterfly valves and control them directly...

1) Servo motors are far too expensive and complicated to drive
2) No easy way to mount anything like that on the valves (two, both freestanding on the pipes + I'd have to add a third)
3) They'd need to have at least 100 lb*ft of torque each
4) A third valve would have to be added as a bypass to keep the water in the loop moving f the main valves were closed

If would be far cheaper for me to relace the pneumatic diaphram type actuator with an electronic one on the three way valve.

There are several reasons and advantages as to why pneumatic controls are used in large HVAC systems

1) They're extremely simple
2) They're inexpensive
3) They're fully proportional
4) They're relaible
5) They don't require large wiring runs, just a simple 1/4" tube

We could change this system over to electronic digital proportional control but it would cost around $100,000

If I had that kind of money at my disposal I'd buy a larger chiller and not have the problem to begin with.

I've actually got two extra 400 ton chillers I don't/can't use because there's no cooling tower for them and they'd need to be overhauled since they haven't been in used for almost 10 years.

In other words I'm kind of stuck with what I've got and very little discretionary money I can spend on my own without going through the corporate funding maze so it's either going to become a combination of cheap technolofy on the air pressure control or it's going to be a simple go/no go situation.