Automated Flow Rate Control

[prodigal]

For those that have automated their home brewery setups, how did you automate the control of the flow rate? For example during sparging, you probably don't want to have the ball valve at the output of the MLT open full throttle as it may compact the grain bed.

From what I've read, motorized ball valves are either completely open or complete closed so that won't do it. Some have suggested using a PWM to control the input voltage to a 12V DC pump to control GPM but others have advised against it. Btw, I'm not set on using a DC motor; just using that as an example.

So what's the scoop? Anyone able to successfully automate flow control?

Thanks for the help!

 

[ssgross]

I am working on this as well. My current, non-automated process is motivation for how I plan to tackle it. My rig is all inline. I pump sparge water to the mash tun, and wort is pumped to the kettle with chugger pumps.

What I do now is set the ball valve at the pot, and watch the sight glass. I adjust the ball valve until I get about 1 qt in a minute, and confirm that I got a half gallon in about 2 minutes. I then toy with the output pump to keep the level of the mash tun the same. Nothing surprising here. The point is I fix the flow into the kettle, and never change it. The valve is marked so I hit 1 qt/min for a sparge time of 45-60 minutes.

My plan is to add a float sensor to the mash tun. The sensor will turn the sparge pump on and off. I will set the ball valve to be slightly more open than the kettle valve. This will act as a safety feature to keep the level from getting too low, and limit overcompacting the grain bed. The only problem with this idea is that, with the pump off and the valve fixed at partly open, the level of the hot liquor tank and mash tun will try to equalize via gravity. All I need is a check valve on the hot liqour tank to solve this.

At one time I was thinking of going with automatic valves everywhere. I decided this was just one more place for things to go wrong, and I already swapped out 3 of 10 ssr's that were firing when wanted. What if an ssr became unreliable during brew day? and the valve failed open?

 

[prodigal]

So once you adjust your flow rate via the ball valve, you pretty much never have to adjust it? My concern is that I may have to monkey with the ball valve from one brew to the next to achieve a desired flow rate. Even within the same brew, I'm anticipating having to adjust the flow rate to account for decreasing wort viscosity as the sparge progresses. Maybe this level of control is overkill but if I'm going to automate this, then I'm going to go whole hog!

That being the case, my current goal is to use automatic valves everywhere. I'll use motorized ball valves wherever I don't need to control the flow rate but I'll need to find another solution for the situations where I might want that level of control. Of course, I might change my tune once I total up the cost of all this automation.

I did some further reading and some of you have mentioned using actuated ball valves or stepper motors to regulate flow. I'll have to look into this. In the mean time if anyone has used actuated valves in their automated setup, then please let me know about your experience with it.

 

[ssgross]

my point was that you can control the flow rate by setting a manual ball valve at 1/4 open, and then cycling the pump on and off. setting the ball valve simply puts an upper bound on the flow when the pump is on, so that you don't have too much suction and a stuck mash. A float sensor is a lot cheaper than automated valves. you can tweak the flow rate by adjusting the ball valve on the kettle. As long as the kettle valve is open less than the sparge valve, you won't have to monitor the level in the mash.

I am sure there are other ways to do it. My goal is to only have to monitor flow at 1 place in my automated setup. right now I set it and forget it, and concentrate on the flow into the mash tun. My plan with the sensor means that I can set everything and forget it, or adjust things by only monitoring the kettle valve.

 

[ZVNJ]

Could use a Peristaltic pump


Sent from my iPhone using Home Brew

 

[jpalarchio]

I saw a setup once that had a 3-way valve at the top of the mash tun where the input was coming from the HLT, one output went to the mash tun and one output went back into the HLT. The valve was then tied to a float sensor in the mash tun.

When the liquid level was low on top of the mash tun, the valve was opened to the mash tun and sparge water was added. When the float sensor was tripped, the valve just redirected the sparge water back into the HLT.

I assume this allowed the pump to just run full speed the whole time and the manual output valve on the mash tun controlled the speed of the wort exiting the mash tun.

 

[Facinerous]

I like the idea from jpalarchio. That sounds super simple and logical.

I am currently developing my own system as well. It involves two flow meters. One from the mash to the boil, and one from the HLT to the mash. Then a solinoid valve in line from the HLT to the mash. I plan on using a arduino style micro controller to take input from the two flow meters and regulate the solinoid valve accordingly.

This setup is not inexpensive, thinking will cost me around 300 or so dollars, but will be truly automated. The output from the mash tank, whether pump or gravity fed will always regulate the flow from the HLT.

Its a long term plan and jpalarchio's idea would be much cheaper. Its just something for me to geek out on.

 

[prodigal]

my point was that you can control the flow rate by setting a manual ball valve at 1/4 open, and then cycling the pump on and off. setting the ball valve simply puts an upper bound on the flow when the pump is on, so that you don't have too much suction and a stuck mash. A float sensor is a lot cheaper than automated valves. you can tweak the flow rate by adjusting the ball valve on the kettle. As long as the kettle valve is open less than the sparge valve, you won't have to monitor the level in the mash.

Ah, I got ya. That's not a bad idea. I'll have to keep that in mind if I can't find a fully automated solution within reasonable cost.

Could use a Peristaltic pump


Sent from my iPhone using Home Brew

I'll have to look into that. So far, I've only thought about a march, chugger, or 12V us solar pump.

I saw a setup once that had a 3-way valve at the top of the mash tun where the input was coming from the HLT, one output went to the mash tun and one output went back into the HLT. The valve was then tied to a float sensor in the mash tun.

I actually wasn't planning on having an HLT. I live in a condo and brew on my balcony so real estate is limited. I'm trying to minimize the footprint of this setup so I figured replacing the HLT with a hot-water-on -demand source (RIMS tube possibly) would be an easy way to do that.

Thanks for the input so far. It's always nice to hear how others have tackled these problems and try to borrow from their ideas.

 

[ZVNJ]

I would like to gravity feed the hlt through a float valve on the mash, with Peristaltic pump feeding the kettle set at 5 gph. Should work well, and without any controls

 

[ssgross]

From what I've read, motorized ball valves are either completely open or complete closed

Still working on this. In another thread I asked about the time it takes the 1/2" motorized ball valves to open or close. Looks like 10-20 seconds depending on the model used. I don't see this as practical for the purpose of automatic sparging. Is anyone using them for this?

 

[atoughram]

You could use an Arduino driving a radio control servo connected to a ball valve. For feedback, a hall effect flow sensor inline would work.

 

[ssgross]

what is the input voltage on the RC motor?
I have never set up an arduino, however this has been on my bucket list for a while now.
I have a BCS driven system. I think I will put a float sensor to control the HLT pump/level in the mash, with something to prevent backflow, and bypass the need for a motorized control valve. Right now, switching to sparge invloves sliding a wood block in above the valve. The block is measured so that the most the valve can be opened yields 13gal in one hour.
Not quite high speed, but gets the job done. My main goal is to not have to play with the valves - set it, forget it, tune back in for the last 5 minutes to switch proccesses.

 

[atoughram]

what is the input voltage on the RC motor?
I have never set up an arduino, however this has been on my bucket list for a while now.
I have a BCS driven system. I think I will put a float sensor to control the HLT pump/level in the mash, with something to prevent backflow, and bypass the need for a motorized control valve. Right now, switching to sparge invloves sliding a wood block in above the valve. The block is measured so that the most the valve can be opened yields 13gal in one hour.
Not quite high speed, but gets the job done. My main goal is to not have to play with the valves - set it, forget it, tune back in for the last 5 minutes to switch proccesses.

I've driven them with 5V dc right off the arduino.

http://playground.arduino.cc/Learning/SingleServoExample#.Uz9TkRCXGYk

 

[jgalati]

A 12v DC pump with volume feedback in the hot liquor tank and boil kettle will give you the most control

 

[ssgross]

What do you mean by volume feedback? I Google and found lots of 12v dc diaphragm pumps that say "X liters per minute", but nothing with volume feedback. are you describing something with a variable flow rate that you can set?

 

[jgalati]

Volume feedback would be a way of knowing how many gallons of liquid you have in the hlt and bk. As you spurge, your change in volume for each vessel would be the feedback used to either speed up each pump or slow them down. You want your change in liquid volume to equal each other. Knowing mash volume is also unnecessary. Look for march 12v DC 809 pump on tescopumps.com

 

[jgalati]

The most accurate way I've researched to measure volume is with an air bubbler system that uses an aquarium pump. A Google search including "oscys controls" should point you in the right direction

 

[muhteeus]

The most accurate way I've researched to measure volume is with an air bubbler system that uses an aquarium pump. A Google search including "oscys controls" should point you in the right direction

Interesting concept.

edit: LINK FOR THE LAZY. This guy gets a VERY ROBUST sensor with +-0.01gallon accuracy!

 

[ssgross]

For BCS users, there is a thread on the ecc forums about "tricking" a temp input to read volume using the sensor referenced above.

http://forum.embeddedcc.com/viewtopic.php?f=4&t=1102

This was back in 2012. Any other BCS'ers out there played with this?

I don't know how successful I would be reviving the old ecc thread. I will try.

EDIT: After a bit more snooping,
http://forum.embeddedcc.com/viewtopic.php?f=17&t=2193&p=6941&hilit=volume+sensor#p6941

here, the arduino sends out a signal when set point is reached. One step closer, but it would be nice if the readout could be incorporated into the bcs.

 

[helibrewer]

So once you adjust your flow rate via the ball valve, you pretty much never have to adjust it? My concern is that I may have to monkey with the ball valve from one brew to the next to achieve a desired flow rate. Even within the same brew, I'm anticipating having to adjust the flow rate to account for decreasing wort viscosity as the sparge progresses. Maybe this level of control is overkill but if I'm going to automate this, then I'm going to go whole hog!

That being the case, my current goal is to use automatic valves everywhere. I'll use motorized ball valves wherever I don't need to control the flow rate but I'll need to find another solution for the situations where I might want that level of control. Of course, I might change my tune once I total up the cost of all this automation.

I did some further reading and some of you have mentioned using actuated ball valves or stepper motors to regulate flow. I'll have to look into this. In the mean time if anyone has used actuated valves in their automated setup, then please let me know about your experience with it.

With an arduino and two pairs of wires it is simple to turn a pump on and off to control a level, no float switch required. The distance between the wire pairs (sensors) is the water level you want to maintain. When both pairs are uncovered the pump runs, when both pairs become covered the pump stops. I used this setup to control the level in my grant.

 

[neumeon]

To automate flow, and do it right is expensive. There are a lot of creative schemes out there but IMO they just aren't robust enough to be repeatable and reliable every time.

In the industrial world I use proportional or modulating valves to control flow but they are pricey. Using a flow meter and a PID feedback loop, flow can be controlled to a set point just like temperature can. Also, ball valves are really bad at controlling flow - they are extremely non-linear and all the flow control action happens in the last 1/4 to 1/8 turn. I've used 30 & 60 deg ball valves in flow control loops and they work much better but also expensive.

My question is what are you willing to spend for the task you want to accomplish? I have a highly automated brew system with full PLC and HMI control - and for this application of RIMS recirc flow control I use a simple MANUAL ball valve. IMO the costs simply don't justify automation here. My 2 cents...

 

[kladue]

I did not spend that much for flow control with flow meters and proportional ball valves, but this would not be difficult to implement with even an Arduino platform. It would take a pulse output flow sensor ~ $10, and a floating control valve from belimo~$140 new/ ~$30-$50 ebay. This would not work with the 12V valves from ECC as they move too fast to work as a floating control device, you need a 90 second stop to stop actuator to make this approach work.
Level measurement with dry sensors and bubblers works, but some math and temperature measurement is needed to correct for density change as temperature changes.

 

[neumeon]

... It would take a pulse output flow sensor ~ $10,

The inexpensive Chinese paddle wheel sensors I've used (from eBay, etc.) work great as long as there aren't any particulates in the stream. Otherwise they jam up with even very small particles - they don't publish much in the way of specs for these sensors but my guess is particulates have to be less than 50um to work repeatably. They will also bind up with dissolved and sticky residues - so using them on recirculating wort is probably not the best idea. My 2 cents.

I use one of these $10 sensors to measure volume into my HLT & MT but I placed it post filter and I've had no issues with over a year in service.

One more thing. The $10 sensors are notoriously non-linear across flow range. They measure on the low side at low flow and on the high side at max flow - up to 15% off at each end. I had to calibrate several times and came up with an average for volume/pulse figure. I also calibrated several different but identical sensors and the calibration was unique for each sensor.

 

[kladue]

I use Malema M10000 flow transmitters in boiler and wort circuits and have not had problems in the last 5 years of operation, but use SS screen over false bottom. I will take your advice on the cheap flow sensors, not good for most false bottom openings which can pass larger particles.
I prefer level sensing over flow for batch control for small volume brewing applications, needs more code and temperature feedback, but accuracy and repeatability is better.

 

[wbarber69]

Volume feedback would be a way of knowing how many gallons of liquid you have in the hlt and bk. As you spurge, your change in volume for each vessel would be the feedback used to either speed up each pump or slow them down. You want your change in liquid volume to equal each other. Knowing mash volume is also unnecessary. Look for march 12v DC 809 pump on tescopumps.com

Much cheaper 12v high flow pump.
http://www.ussolarpumps.com/index.php?main_page=product_info&cPath=71_116_146&products_id=241

 

[jgalati]

^I have one of those as well and can attest that they are more than enough for everything but whirlpooling

 

[wbarber69]

^I have one of those as well and can attest that they are more than enough for everything but whirlpooling

Follow the link us solar pumps has a new high flow pump. I doubt that we are talking about the same thing.

 

[LordUlrich]

While I admit I am not an automation expert by any stretch, I will add my 2 cents.

In brewing we seldom actually care about the flow, but are more concerned with the volume in a vessel. Flow can also be calculated by measuring volume (and time)

The motorized ball valves (as I understand without actually haven gotten one yet) do need power to close. You could use 2 lines to control the valve. By turning off the power the valve will stay where it is, the other line is used to control if you want it to open or close when power is applied.

 

[audger]

a dual-channel peristaltic pump with a shared stepper motor to drive it would ensure that for every ounce of wort coming out the bottom, an ounce of water is replaced at the top of the grain bed... controlling the speed of the motor controls the speed of the lauter directly.

easy and simple; no messing with flow sensors, valves, control circuitry, etc....

 

[wbarber69]

a dual-channel peristaltic pump with a shared stepper motor to drive it would ensure that for every ounce of wort coming out the bottom, an ounce of water is replaced at the top of the grain bed... controlling the speed of the motor controls the speed of the lauter directly.

easy and simple; no messing with flow sensors, valves, control circuitry, etc....

Can I get one from you. I got 100 bucks.

 

[hightechlofi]

I am currently in the process of building my system out. I am using 12v pumps coupled with Hall effect style flowmeters. As stated above, they have to be calibrated as they are non-linear, but my plan is to use a pid algorithm (I ported the arduino pid library to javascript as I am controlling with a Beaglebone black using node.js) to allow me to set a desired flowrate (the setpoint) and letting the pid adjust the pump output via pwm to maintain the desired flowrate.

I have al of the equipment, but am still in the programming phase, so I cannot attest that this will wot out in practice, but it is what I am gunning for.


Sent from my iPad using Home Brew

 

[wbarber69]

I am currently in the process of building my system out. I am using 12v pumps coupled with Hall effect style flowmeters. As stated above, they have to be calibrated as they are non-linear, but my plan is to use a pid algorithm (I ported the arduino pid library to javascript as I am controlling with a Beaglebone black using node.js) to allow me to set a desired flowrate (the setpoint) and letting the pid adjust the pump output via pwm to maintain the desired flowrate.

I have al of the equipment, but am still in the programming phase, so I cannot attest that this will wot out in practice, but it is what I am gunning for.


Sent from my iPad using Home Brew

What kind of 12v pumps are u using. I've tried to use pwm on the topsflo pumps from ussolarpumos to no avail..since the pumps use a built in controller to get maximum flow out of whatever voltage is provided.

 

[hightechlofi]

http://m.ebay.com/itm/251457815300?nav=SEARCH

Thee are cheap, foodsafe, and seem to work. They ship from China though so it takes a while for them to arrive.


Sent from my iPhone using Home Brew

 

[wbarber69]

Hmm 2.6 GPM that's not bad for the price. It would be great to control a mash tun.

 

[hightechlofi]

Yep, I have them on my hlt, MLT, and my rims tube. I still use a march 809 on my bk for whirl pooling and chill, but these guys are great for sparging, recirculating, and transfers.


Sent from my iPhone using Home Brew

 

[wbarber69]

Yep, I have them on my hlt, MLT, and my rims tube. I still use a march 809 on my bk for whirl pooling and chill, but these guys are great for sparging, recirculating, and transfers.


Sent from my iPhone using Home Brew

Check out ussolarpumps new TD-5 6gpm pump if u want to go all dc. I use one for my whirlpool and it sick.

 

[mattd2]

I am currently in the process of building my system out. I am using 12v pumps coupled with Hall effect style flowmeters. As stated above, they have to be calibrated as they are non-linear, but my plan is to use a pid algorithm (I ported the arduino pid library to javascript as I am controlling with a Beaglebone black using node.js) to allow me to set a desired flowrate (the setpoint) and letting the pid adjust the pump output via pwm to maintain the desired flowrate.

I have al of the equipment, but am still in the programming phase, so I cannot attest that this will wot out in practice, but it is what I am gunning for.


Sent from my iPad using Home Brew

Hightech how are you going with this setup? Have the pumps arrived yet? I am looking to implement something similar with an AC pump (Iwaki/Resun - similar to a March/Chugger) using a non zero-crossing ssr. I have tested it with a scr and can turn down the flow rate in my pump from 12 lpm down to 4 lpm. Still more than I would like for a 20 / 40 litre system but gives hope that with a smaller March 815 type pump it would give good control. Just need to find a few spare dollars for the elec parts needed.

 

[ssgross]

I am building a bubbler type volume sensor with an arduino handling the computing. I am programming it so that I can either output volume, or the rate of change of the volume.

 

[fuxorfly]

You could use an Arduino driving a radio control servo connected to a ball valve. For feedback, a hall effect flow sensor inline would work.

Is this a picture of your project? Could you link me (or give me more details) as to your build?

 

[lael]

With an arduino and two pairs of wires it is simple to turn a pump on and off to control a level, no float switch required. The distance between the wire pairs (sensors) is the water level you want to maintain. When both pairs are uncovered the pump runs, when both pairs become covered the pump stops. I used this setup to control the level in my grant.

That is cool. How do I do this? Would it work as a sensor in wort? (Auto sparge using a pump)