Ball Valve Controller

Projects General Guidance
1

Any advice for explaining a problem/general tips are most welcome.

Hi,

I am trying to build an automatic controller for my distillation unit (whiskey still). It is a 200L vessel with a reflux cooler at the top, then a condenser to the outlet. When we run a brew through it probably takes us 6-7 hours and we are constantly (checking every 5 mins once its stable) adjusting the reflux water flow (with a manual ball valve), while monitoring the
temperature at the reflux outlet.

I have purchased a mechanical ball valve which has a geared DC motor to adjust the water flow rate, it has limit switches at the fully closed and fully open ends of movement. I intend to use a PID controller with input from a thermocouple (MAX6675) as feedback, but need to be able to apply power to the valve to get it to the correct position/flow rate.

I'm trying to see if my position control idea is too flawed and i should just add a potentiometer or if it really may be possible.

My idea was to jog the valve with a fixed duration pulse so it just slightly opens or closes a set amount each time. In reality because of increasing friction due to seal surface contact as the valve closes, more pulses are required to close the valve than open it. Also the gearbox has some slack when changing direction, a pulse is needed to take up that slack so the next movement is to the same set amount.
I'm not confident at this stage that getting accurate control is possible.

I'm just learning about PID control and have never implemented it, but i understand that it should adapt to external changes on the system, maybe even control instrument drift (thermocouple seems to be stable at least).

Sorry the code has lots of irrelevant stuff for the problem at hand. Most of the code i'm trying to fix is just above the commented out PID control in the loop.
from:

"if (mode ==0) {"

Hope this makes sense. Any advice for how to explain my problems/general tips are most welcome.

In short is a position control possible on a ball valve (see link below) without a potentiometer?

Casey

Mechanical Ball Valve link:

brew_w_lcd2.ino.ino (17.7 KB)

2

Attached photo with labels

3

#1) the mechanical device has to be able to be moved so that movement will effect change in the process.
a good rule of thumb is that your device should be 4 times more sensitive then the desired thing.

if you want to control within 1 gallon per minute, then a change should be able to deliver a step change of 1/4 gallon per minute or better.

#2) flow through an orifice is dependent on the pressure drop across the orifice.
if, in reducing the orifice, the pressure increases, then the flow increases so flow rate will increase with the hopes of the volume decreasing.

I would offer that you can send a 1ms pulse, once every 10 seconds. then see if that is enough to create movement.
increase to some point you do get movement. there will be a point where the pulse does not offer enough power to create movement, and some point of longer duration where it does.
what you would want to do is to try that out.

I would offer that is you made a step change, the temperature will change over time. so, you would want too know that time.

consider another alternative, get the smallest valve you can, connect it on the discharge of the pump and let it drain into the reservoir, essentially giving the pumped water a place to go, instead of your process. this is called a run-around loop. it can reduce the pressure on your main valve. see #2 above

4

Problem with trying to control flowrate with a ball valve is near closed position, the flow increases a large amount with a small rotation and near full open, very little change in flowrate with a large rotation, it's non linear.