Simple automated spunding valve

[sounddoc]

I've got the BrewPi set up now (I'm using dotps1's Arduino uno / Raspberry pi build) and I'm now hooked on microcontrollers. I'm going to go down this route unless there's a better option in designing a spunding valve that can be tied into the version of brewpi that I'm using (former web developer here).

So far I have these two bits to start with:

Adafruit MPRLS Ported Pressure Sensor Breakout - 0 to 25 PSI

We stock a few barometric pressure sensors, great for altitude and weather measurements. This pressure sensor is special because it comes with a metal port! Unlike other pressure sensors, ...

www.adafruit.com

6V Air Valve with 2-pin JST PH Connector

Here's a cool new valve for makers interested in pneumatic systems, hydraulics, inflatables, projects with compressed air, and the like. These air valves, also known by other names like, ...

www.adafruit.com

Obviously a 2nd Ardiuno will have to join the family (can't have too many, I'm finding).

Opinions? Other pieces of hardware (other than plumbing) that I'd need? It would be my first 'from scratch' Arduino project.
Thanks!

 

[sounddoc]

I should also note if possible I don’t mind tying this directly to the pi too.

 

[Vale71]

Make sure the valve vents through as small as possible a pinhole otherwise with the combined hysteresis you might find your system constantly undershooting the set point.

 

[BrunDog]

Do you know @Die_Beerery? He has done this successfully. I don’t think he is active here these days but you might ping him at **********************.

 

[sounddoc]

Make sure the valve vents through as small as possible a pinhole otherwise with the combined hysteresis you might find your system constantly undershooting the set point.

Is there a source I could learn about with this? I'm not familiar, but I do appreciate the opportunity to learn!

 

[BrunDog]

Per @Vale71, you need a needle valve to slowly bleed off pressure so you can monitor the pressure and determine exactly when to close the valve. Otherwise you’ll dump all the pressure and undershoot as he said.

 

[sounddoc]

Ah I get it. Dumping the pressure when it hits a target means it’ll continue and the transducer won’t reliably read a value where it should close the valve back up. I wonder if a timed release would help, instead of dump until it hits the target, dump for a few seconds, close then check, rinse and repeat.

it’ll make more sense once I have a test circuit with the hardware set up. Thanks all!

 

[day_trippr]

It is an input change rate verses response time issue. Using a very tiny orifice on your solenoid valve would better allow the sensor - and full feed back loop - to hold a target pressure.

A similar paradigm is using a 1000 watt hair drier as the heating element in a small fermentation chamber. By the time the controller can react the overshoot is massive...

Cheers!

 

[crane]

From my experience the solenoid valve provides enough restriction to not cause any undershoot issues. No need to overthink this, just use simple hysteresis.

 

[crane]

Removed

 

[Bilsch]

And it's worth every penny.

 

[Paulaner]

Oh oh, is this thread going to get deleted now?

 

[moorejl57]

I had made a software based pressure regulator for a work project years ago to control a 2000 psi pressure chamber using a PLC. The designer had spec'd a $1500 mechanical regulator that performed very erratically and had a large hysteresis. I removed the regulator and made use of the pressure sensor already present on the system along with a fuzzy logic controller that kept the pressure within 20 psi of target. It did employ small orifices to both fill and bleed the pressure in the chamber.

 

[pocketmon]

I implemented pressure control in BrewPiLess. However, my hardware building skill sucks and I haven't run it in real life.
Some people do report working, though.