StillDragon® Community Forum

Welcome!

Be part of our community & join our international next generation forum now!

In this Discussion

RC Cooling Automation

Hi all,

There is a lot of discussion lately regarding reflux cooling automation using motorized valves to control cooling water flow rates. Now I freely admit that there are many here with far more expertise than me in this field so please put me right if I'm wrong in this.

It seems to me however that it would be simpler, elegant, cheaper and more accurate to control the flow by controlling a mag drive pump on the input and providing a cheap needle valve in the RC output to provide the pump something to push against.

Like I say maybe I'm missing something obvious and I'll be very happy to be told so.

«1

Comments

  • Hi Ethan,

    Can you further elaborate when you say "cheaper and more accurate to control the flow by controlling a mag drive pump"

    StillDragon North America - Your StillDragon® Distributor for North America

  • edited March 2015

    I am by self confession not an expert in this field, it was just an idea that I had and already having all the required hardware I thought I would run it by the members to see if it was a practical idea.

    I presume that current endeavours use microprocessor control to fire a solenoid valve on the output (sort of like a machine gun going off), I've not come across a motorised needle valve at anything like a sensible price.

    Now considering the amount of current discussion this sort of setup must obviously work but

    1. Here in the UK such valves (other than cheap rubbish) are expensive and most discussions that I see already drive their systems with a pump anyway.
    2. Controlling a pump on the input would be a simple micro controller task and would automatically negate any supply pressure variations.
    3. It would not require any plumbing alterations to an already functional manual operation system
    4. Listening to a solenoid firing at, I don't know maybe a few hertz, for several hours would drive me up the wall.

    Re-reading my initial post tends to imply that I don't think current solutions would be accurate. I did not intend to give this impression, I should have said "as accurate" or even "sufficiently accurate". I justify my assertion regarding cost on the grounds that an already functional system could be converted at only the cost of the electronics, no plumbing alterations would be required.

  • Depends. Wouldn't work for me as I'm chilling with tap water and my only option is to control the flow.

    Generally I don't disagree. However, how are you going to control the output of the pump?

  • edited March 2015

    @Unsensibel I'm not an expert or even that knowledgeable but I presume it wouldn't be that difficult with micro-processor control, such hardware would be required in any case whatever form of automation is employed.

    My mains supply is crap, if someone down the street flushes their toilet it would de-stabilise my still. I currently use a cheap mains mag drive pump to supply my RC and the standard SD valve in the output to provide something for the pump to work against. I control this now using a PWM motor speed controller and it works great, I'm really only discussing a means of saving me twiddling the knob and maybe being able to just dial in temps

  • edited March 2015

    Agree, it seems more logical to simply control the flow with the pump speed versus spinning the pump at full speed, then using an additional valve to divert or restrict it to get the flow rate you need. You are right on.

    Problem is, most cheap mag pumps use cheap motors, you can't accurately control the speed of a cheap motor. To get it to work right is going to take a serious investment (or some fancy tinkering).

    At a minimum, a motor needs to be 3 phase to be able to be speed controlled with a variable frequency drive (VFD). It's not uncommon to find VFD that accepts a 0-20ma or 0-10v process input for speed control. The problem is, your range of variability is low, it's not 0-100% like a valve, it's more like 25-110% (yes, you can make it spin faster). So you would need to accurately size your motor to your flow rate needs.

    What's even more complicated is that you need a special kind of 3 phase motor, called an inverter grade or inverter duty, which is built to handle the additional stress associated with speed control. That lets you spin it slower and faster than a standard duty 3 phase. But even that doesn't give you the wide variability a valve would. Run a cheap 60hz 3 phase motor at 5hz under torque and watch it incinerate itself in a few minutes.

    There are other options, like servo motors and DC motors intended for PWM speed control, but these tend to sized differently than standard motors you find in pump applications, so to use one of these, you'd need to adapt the pump head to motor drive. That'll require some custom fabrication, which probably won't be cheap. This would be a much better approach than trying to use a 3 phase and VFD. However, now you need a serious dc power supply and a motor controller.

    There isn't a cheap pump you'd find that could do this, and an off-the-shelf 3 phase motor with a decent VFD (Hitachi, Omron, Fuji, etc) is going to get pretty spendy, and won't give you anywhere near the performance a proper control valve would.

    Decent control valves (Johnson, Flowserve, Belimo, etc) are made for exactly this kind of use in cooling systems.

  • edited March 2015

    The other issue is the pump. You might be in a low coolant flow rate situation. Most pumps aren't really good for this. however there are options. Perhaps a peristaltic pump could be used for low coolant flowrates, inparallel with a bigger pump? Or even justa big and small pump in parallel. Small for low flow rates, big for highflow rates.

  • You never, ever, ever want to restrict on the inlet of a centrifugal pump.... and the type of motors on 99% of mag drive pumps do not like being speed controlled... you may 'get away' with it, but you are hurting the motor just don't do it without knowing you may trash your motor...

    Restrict the outlet for flow control, even fully closed is better than cavitation and slipping the mag drive... ask any, and I mean ANY pump expert...

    There are many, many solutions, but the best one for you involves taking you absolute hard requirements (water temp, water pressure, and availability/cost being the main ones) and working from there.

    I have made systems work with city water at 70-80F, pool water and pump at 70-80F, and IBC totes full of water that heat from 70 to 150 during use...

    you want elegant... look at what is on the PC of a $1M carl/kothe.... a couple hundred dollar danfoss mechanical thermostatic valve...

    you want a smaller sensor, or the ability to change the set temp electronically, then you go to a PID and proportional motorized valve (unless you are BIG, then you go to a air diaphragm valve)

    having a non-automated valve on the RC is something you will never, ever, ever want to do again once you use automatic a couple times...

  • I get the concern on the bigger / more expensive pumps & larger set-ups. For a hobby level though, not that much water flow is needed & I'm wondering if a Harbor-Freight-Experiment would be worth a shot:

    $10 submersible fountain pump + $20 router speed controller

    Plug everything up & see how it runs. Worst case, you're frying $10 and have a router controller for the woodshed ;)

  • ummm.. if you are using that for the RC and have a sufficient source of PC cooling water for 0 RC water flow, I can see that... I know the chances are slim, and only really a possibility if you walk away and something happens, but worst case losing a cooling pump can mean loss of life or house...

  • I use a $30 HF submersible pump that pumps 640 gph. 5 months later and still working.

  • great to hear... how many watts is it knocking down, maybe any other info you think people would want to know?

  • Here's the one I use. Correction on price; $39 and its 620 gph. I'm sure it isn't the most rugged in the world but it has been working fine for me.

    620 GPH Submersible Fountain Pump @ Harbor Freight Tools

  • What do you use for a controller on it?

    StillDragon Australia & New Zealand - Your StillDragon® Distributor for Australia & New Zealand

  • My controller might be a little high tech for you guys. After watching these threads, I realize that you guys are way behind me in both ideas and ability to get it done.

    I use a vice grip on the return hose from the column. More cooling water? Turn knob on bottom of handle counter clockwise to open vice. Less water, turn other way.

    :))

  • Doesn't make much sense in the context of the thread is all where the guys are talking about the longevity of cheap pumps when they have a controller attached to the motor if you haven't got a controller attached to the motor of your cheap pump.

    StillDragon Australia & New Zealand - Your StillDragon® Distributor for Australia & New Zealand

  • I guess I should tell you also that I have a splitter in the line coming from the pump; the one to the PC is wide open as the condenser is way undersized and almost useless. The other one goes to the cooling tubes that run through the column. The line coming from the column is the one with the pinch.

  • Wow...first few all I hate my keyboard on my tablet...regardless of whether you are controlling water flow for a still, beer mash, cooling a room, cooling a house, cooling a process it's all the same in the long run. Too little cooling and you get hot. Too much cooling ( yes you can have too much ) and you get huge temperature swings. There are huge factors here that could take weeks to discuss, but you must (reset must ) have a method to modulate water flow to maintain precise and accurate temperatures.
    How do you maintain precise water flows? The simplest method is to sit there and look at the temperature of the water leaving the condenser and physical open or close the valve. Works 100% of the time.
    I am not an expert, but I have a lot of experience. You can vary the opening and closing of a valve by several methods. The opening and closing of a valve through a pid routine is the least precise method. Any form of modulating the valve is 1000% more accurate.
    Alsa

  • You can modulate a valve by several different methods. Most methods end up with extremely similar results For our intents. You can send signal that varies from 2 volts to 10 volts DC, you Can add a resistor to this signal and make this a 4 to 20 ma signal. You can incrementally open or close the valve by pulsing voltage to an end contact or a close contact ( called tri state ). Most of these methods will work very similarly, so you just need to choose one and saddle up for that ride. Any "modulating" control will incrementally open and close the valve allowing more or less water flow to the condenser based on the need from the controller. It doesn't matter how You are controlling ( 2 to 10 vdc, 4-20vdc, tri state ), but the valve will be opened to a set percent based on the controllers output. A pid controller will pulse an open /closed valve on and off to try and maintain a water flow which is not very effective. Yes it is a pain, but either find a controller ( 2 to 10 vdc, 4 to 20 ma, tri state ) and then look for valve to match that signal or pick the valve and then find a controller to match. All of the threads and info on PID controllers are all based on a communist theory that requires us to drink a lot to me it work. Pid algorithms are designed for extremely tight temperature or pressure or flow controls. I am talking 1/100 of degree situations. If you can deal with +/-1 for (and who cant) then any of the control scenarios I have listed above will work fine.

  • Can you modulate water flow with a pump? Yes. As mentioned above it requires an inverter duty 3 phase pump motor and a vfd (variable frequency drive ) controller. You still have to have one of the controllers I listed previously ( 0 to 10 vdc 2 to 10 vdc, 4 to 20 ma....) to control this vfd and you have to have a pump that will operate on 3 phase power and will be an inverter duty motor. So for most hobby or small operations this will not be an economically viable option and for this particular phase of operation have not advantage

  • Guys it's late and yes I have been drinking, hence some serious typos on the threads above, but after thinking about this I need to provide some explanation that will most likely be confusing.
    PID is an acronym for Period Integer and Derivative. I kind of know what this is, but will deny any knowledge if ever asked. Basically it is a program that looks at the difference between the actual temperature, the setpoint and how long the temperature has deviated from the setpoint ( to help anticipate external temperatures influences). In reality all 2 to 10 vdc, 4 to 20 vdc and tristate...signals use pid technologies.
    However in most of these forums the term PID is used to describe a controller that pulses a signal on and off ( versus a modulating signal ) which is often used with a SSR ( solid state relay ) to control heating elements. Trying to use this PID signal to control an On /Off solenoid valve will result in an unacceptable control. In reality most of the controls we use will be a pid. However we need controllers and valves that use a 2 to 10 vdc or a 4 to 20 ma signal. They need to match though.

  • I work with all kinds of fancy I&C shit all the time with my day job. The last thing I want to do is over-automate my stilling. I enjoy the heck out of reading the simple SD digital thermometer on the defleg outlet and adjusting the flow using my thumb and index finger on the needle valve stem.

    When you fly jets all the time, sometimes its just fun to hop in a Piper Cub and fly by the seat of your pants.

    I'm more like I am now than I was before.

  • Nice. I agree with Kapea, but if you wish to automate I won't tell you not to. Theres enough room here for everyone. :D

  • Through my 10,000 words above I haven't addressed how to automate control. Honestly I can't answer that. I have a dash 4 plate column. I typically run about 20 c on water temp leaving dephlag while I am compressing heads and ignore my vapor temperature. After an hour or so I shoot for a vapor temp above the dephlag of 67 to 68.6 and a liquid temperature leaving the dephlag of around 40. I try to hold the product condenser at around 20 the whole time. Typically I am running about 4 to 4.2 kW while collecting. If I were going to automate I think I would shoot for a 68.2 above the dephlag and a 25 leaving water temp on the dephlag while compression heads. I think I would control on water temp leaving the dephlag to a max of 40, water leaving the product condenser to 21 and a max temp above 5 hr dephlag 68.5?

  • edited March 2015

    I'm not sure what is over or under automation, or if we'd all draw the line in the same place. But I know one thing, maybe two.

    When I don't need to worry about the product condenser coolant flow, and I don't need to worry about the power input to the boiler changing unexpectedly, and I don't need to worry about fiddling with the dephlegmator temperature, I can focus a whole lot more time and energy on the important bits.

    By being able to see exactly the temperatures, flow rates, power inputs, etc, it provides me with the tools to produce a product more consistent than otherwise possible.

    What I would define over-automation is, is any attempt to try to automate cuts, which I would say is absolutely worthless and a waste of time.

    Batch consistency isn't overrated. We've all had that magic batch, where you did everything exactly the same as the time before, but the end result was so exceptional, you are left scratching your head. After running with controls, especially on the dephlegmator, I don't know how the hell I was doing anything before. Adjusting the dephlegmator completely blind? No way I'd go back.

  • PID stands for Proportional Integral Derivative. Here we go again :D Plenty of people here use PID control to control on/off solenoid valves using @olddog control systems and they work well. PID is a process that can be used to control an output that modulates, pulses or floats.

  • I could throw all of my automation away with the exception of my danfoss PC and my PID-proportional RC... those you will have to pry from my cold dead hands.. ;-)

  • edited March 2015

    I understand the desire to automate. I meant no disrepect to those who enjoy using instruments and controls on their stills. Refining all of that down to a fine point is very satifying I'm sure.

    I enjoy reaching the fine point empirically. There is satisfaction in repeatability from experience. But, this is a hobby for me. If my bottom line depended on wringing out the most volume possible without dragging up tails, I'm sure I would have all kinds of sensors and data loggers all over the place. I love finding minute clues in the details of strip charts, when I'm on the clock.

    I'm more like I am now than I was before.

  • +1 kapea.

    StillDragon North America - Your StillDragon® Distributor for North America

  • I have both options available to me when I using my machine. Generally I use the manual on the first go around and the auto on the next and subsequent.

  • I have been using my home made closed loop system for two years now and it works great, I use a 5 gallon bucket or 5 gal IGLOO cooler to contain the cooling water and ice. I have a small aquarium pump pumping cooling water to the condenser I use an inline needle valve to control flow to the dephleg and the condenser. The bill for 40# of ice for a 4 hour run is only $7 I cant beat it. As for automation, I have not needed it yet the needle valves do a great job set it and forget it. Check out the photos I have included. I am showing two rigs. A pot still with a packed column and a reflux column with the dephleg.

    image

    image

    image

    1.jpg
    600 x 800 - 87K
    2.jpg
    600 x 800 - 82K
    3.jpg
    600 x 800 - 78K
Sign In or Register to comment.