Pump control

Few questions.

1. If your mash recirculation is automated to maintain temp do you use AC or DC puimp?

2. How do you control the speed if you use AC?

3. Has anybody developed anything in labview to control their pump and how do they interface the control hardware to LV?

gerrywalsh said:

Few questions.

1. If your mash recirculation is automated to maintain temp do you use AC or DC puimp?

2. How do you control the speed if you use AC?

3. Has anybody developed anything in labview to control their pump and how do they interface the control hardware to LV?

Click to expand...

I use an AC pump.

I don't control the speed of the pump. I control the output by closing or opening a ball valve. It won't hurt a magnetically coupled pump like the typical March pump used in this forum.

I don't use labview, so maybe you would need a stepper motor for a pump?

TheFlyingBeer uses LabView and has made a peristatlic pump prototype using a stepper motor for this. You could also just automate a ball valve with a standard mag drive pump like most of us use.

wyzazz said:

TheFlyingBeer uses LabView and has made a peristatlic pump prototype using a stepper motor for this. You could also just automate a ball valve with a standard mag drive pump like most of us use.

Click to expand...

I quickly gave up on labview to control the stepper motor for that pump, there is too much SW and OS overhead to drive a parallel port at high frequencies reliably. I am working on adding code to my LPC1768 to handle the stepper. Labview will send the MCU speed commands and the MCU will handle the lower level details.

stlbeer said:

I don't control the speed of the pump. I control the output by closing or opening a ball valve. It won't hurt a magnetically coupled pump like the typical March pump used in this forum.

Click to expand...

Does this setup maintain mash temps accurately. From the way i read it your valve is fully open or fully closed, no in between. when the valve closes does this knock off your pump at the same time or is there not any requirement to turn it off? I know you kind of answered this already but i want to be certain it's ok to drive these against a closed valve!

wyzazz said:

TheFlyingBeer uses LabView and has made a peristatlic pump prototype using a stepper motor for this. You could also just automate a ball valve with a standard mag drive pump like most of us use.

Click to expand...

What kind of vavles are you using? how are they controlled? Can you throttle the flow rate?

gerrywalsh said:

What kind of vavles are you using? how are they controlled? Can you throttle the flow rate?

Click to expand...

I don't use automated valves, but there are plenty that use asco valves and the like to automate their systems. I have seen them open/close partially based on which wires on the valve current is applied to. You could control this using a PLC, BCS-460, Brewtroller, manual buttons on a Control Panel, or perhaps even LabView I'd imagine.

gerrywalsh said:

Does this setup maintain mash temps accurately. From the way i read it your valve is fully open or fully closed, no in between. when the valve closes does this knock off your pump at the same time or is there not any requirement to turn it off? I know you kind of answered this already but i want to be certain it's ok to drive these against a closed valve!

Click to expand...

Since my system is a RIMS I am monitoring the outflow from it and the PID adds heat or turns off the heater as needed. It is a continuous flow system. I reduce the flow rate from the pump by partially closing the ball valve to < 50% open. It is a manual control for the flow. Fully restricting the pump is no problem.

If you want to automate the pump flow but only have digital outputs, 2 outputs will work when used with a 90 second open/closed valve from Belimo, or other manufacturers like Honeywell and Siemens. The term used for this method is floating control because valve stops when drive power is stopped, and it is pulsed to make changes in either direction and has been in use for over 60 years. Some form of flow feed back is needed to determine which direction to drive the valve when using floating control, easiest/cheapest would be a pulse output flow sensor and a formula to convert pulses to GPM. Some exploration into suitable low cost units has found GE-303 series flow sensors made in China in the $10 range, additional fittings are needed to convert metric threads to NPT.
Analog control is possible with 0-4.9 VDC analog outputs with an Op amp boosting control voltage to 0-10VDC, same valve manufacturers as above have 0 - 10 VDC analog control versions.

its far easier to use PWM on the heater to control temperature than it is to vary the flow rate of the pump (and gives more repeatable/accurate results), if that is what you are going after.

since the valve isnt the only variable restriction in the system (the grain bed restriction will vary considerably), having the ball valve open however much % will not give you the same flow rate from batch to batch (it will even vary over time during the same batch). you would need the feedback of a flow meter to accurately control the flow rate in a dynamic system like this.

what are you trying to do exactly?

stlbeer said:

Since my system is a RIMS I am monitoring the outflow from it and the PID adds heat or turns off the heater as needed. It is a continuous flow system. I reduce the flow rate from the pump by partially closing the ball valve to < 50% open. It is a manual control for the flow. Fully restricting the pump is no problem.

Click to expand...

Cheers, i intend going down the herms route. I dont think a PWM on the element in the herms tank would achieve much, perhaps i'm wrong tho?

kladue said:

If you want to automate the pump flow but only have digital outputs, 2 outputs will work when used with a 90 second open/closed valve from Belimo, or other manufacturers like Honeywell and Siemens. The term used for this method is floating control because valve stops when drive power is stopped, and it is pulsed to make changes in either direction and has been in use for over 60 years. Some form of flow feed back is needed to determine which direction to drive the valve when using floating control, easiest/cheapest would be a pulse output flow sensor and a formula to convert pulses to GPM. Some exploration into suitable low cost units has found GE-303 series flow sensors made in China in the $10 range, additional fittings are needed to convert metric threads to NPT.
Analog control is possible with 0-4.9 VDC analog outputs with an Op amp boosting control voltage to 0-10VDC, same valve manufacturers as above have 0 - 10 VDC analog control versions.

Click to expand...

When you say these valves are pulsed to make changes in direction, why would you do that? i thought recirculation was only done in one direction. Does direction refer to how open or closed this valve is?

If this is the case could a temp sensor in the grain bed be used in conjunction with digital outputs and some software to achieve the same as opposed to a flow sensor. I'm sorry if i've totally misunderstood

A 0-10V analog control valve would be a nice solution, do you know any makes models?

audger said:

its far easier to use PWM on the heater to control temperature than it is to vary the flow rate of the pump (and gives more repeatable/accurate results), if that is what you are going after.

since the valve isnt the only variable restriction in the system (the grain bed restriction will vary considerably), having the ball valve open however much % will not give you the same flow rate from batch to batch (it will even vary over time during the same batch). you would need the feedback of a flow meter to accurately control the flow rate in a dynamic system like this.

what are you trying to do exactly?

Click to expand...

I'm trying to use temperature readbacks from the grain bed, take them into labview and use either the speed of the pump or how open or closed a valve is as the process variable to achieve a setpoint.

Maybe i'm wrong but would PWM on the heating element in the herms tank be useless in terms of controlling the temperature in the grain bed?

I'm trying to use temperature readbacks from the grain bed, take them into labview and use either the speed of the pump or how open or closed a valve is as the process variable to achieve a setpoint.

Click to expand...

the problem with that is the large disconnect between the temperature control (heating element) and the temperature feedback (the sensor). the time it takes to register a temperature change in the mash tun itself is easily enough time to overshoot the RIMS heater temperature by a large amount. reading the temperature at the output of the heater, instead of the mash tun itself, is the way to go.

you then can adjust the integral band settings (the "I" in "PID") to only calculate error based on a smaller number of past readings, like 10-30 seconds, in order to have it react quickly when necessary. setting the "I" so low is acceptable in a system such as this, as the rate of change is very small, comparatively, and the thermal mass is very large. there wont be any large up/down/up temperature swings that you would normally see in a system with much less thermal mass (like an oven or refrigerator when you open the door).

Maybe i'm wrong but would PWM on the heating element in the herms tank be useless in terms of controlling the temperature in the grain bed?

Click to expand...

that is how almost all HERMS and RIMS systems work. what would be useless about it exactly? the time it takes for the RIMS output temp and the mash temp to converge would be slightly longer, but you gain overshoot protection. when you overshoot, you irreversibly denature mash protiens. if there is too much distance between the heater and the temperature probe, you can denature all of them before you (or your temp controller) even realize whats going on.