StillDragon® Community Forum

Welcome!

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

In this Discussion

My PID Experiments

13»

Comments

  • @Smaug said: Thinking out loud.....With a PC large enough to hold a measure of "reserve" water, you would be able to use preheated RC water supplied by the PC with out affecting the knock down capability of the PC.

    If you were doing one process, the same process, all the time, I would agree... but RC heats water for all but full-time mutli-shift distilleries...

    another note.. it is not the outlet temp or the internal temp of the RC that is what is directly related taking heat away and causing reflux, it is the change in temp... if you only control the outlet, but have a changing inlet temp, you will not have stability... maybe control the temp of the input with a mixing valve, and the output with a flow valve?

  • If you add a flow transmitter and transpose the equation I put up last page you end up with the power going through the RC.
    With a few other tricks you can get the total power.
    If you divide the RC power by the total power it gives you a real figure of the Reflux Ratio regardless of feed temps or flow rate or any other variables that affect the outlet/RC temp that everyone knows and loves.

    I think this what we should be controlling directly rather that playing with rubber temps.
    I've made a RR meter that is up in another thread. It works fine and it is a lot easier to go off a 0% 100% figure than trying to interpret some seemingly random temperature range with seasonal fluctuations.

    The two downsides are the need for flow measurement and the fact it needs some sort of mc to to do the calculating on the fly.

  • @smaug @CothermanDistilling that's exactly how I run in manual mode. And I usually dial in my take off rate after seeing my power and don't touch it anymore. My gut tells me that this should be replicable. On the other side, I'm unable to tune right now without oscillation, so my gut might hit be the right indicating...

    @jacksonbrown link please

  • To what? This?
    Trust me. It has more to say with some instruments plugged in :)

    image

    I was using it with some of my parrot testing but the intention was auto control.
    I also got my proportional solenoid valves working manually through the controller too.
    All the components were working well individually but I got side tracked with a few other import projects before I got a chance to combine it all into one.
    The idea is to just twist a knob to the desired reflux ratio and it'll sit there regardless of water temp fluctuation, water pressure fluctuations, element power adjustments. Whatever.
    It should be an extremely consistent and repeatable way to run.

  • edited January 2016

    You could use a small plate heat exchanger to temper the RC input water and use a single PID controlled solenoid on a bypass to control the RC input temp. This would give you more stable input temp and would decouple the flow rates.

  • edited February 2016

    Thinking about the next experiment to make the valve work:

    If flow is too high, cut down the pressure...
    Thinking about this valve as it's not too expensive. Thoughts?

  • Normally you correctly size the valve or orifice to give the flow you want with the pressure drop available.
    If you are trying to make do with an oversized valve acquired through other means, your other option is to tune the pressure to suit.

    The ultimate aim is to get the correct pressure drop across the valve. This is what governs the flow rate. Not just the feed pressure but the deltaP across the valve.

    Pressure reduction on the feed line is a good option.

    A constant back pressure valve on the return is probably a better one (if you can find a cheap one). That way there’s still plenty of pressure downstream to play with if you are not going straight to drain.

    Everyone’s playing with pretty serious processes now so correct sizing is getting more important.

  • @jacksonbrown thanks! I'm not really familiar with fluid dynamics so advice is always good

    I'm looking to contain the cost a little. The back pressure valves seem like an expensive option though

  • edited February 2016

    Can someone explain, in layman terms, what pressure drop really means? Not in terms of valves and plumbing systems, but in terms of dephlemators, product condensers, recirculating loops, and using pumps or straight mains water. How does it apply to us, really, not theoretical.

  • @Unsensibel said: I'm looking to contain the cost a little. The back pressure valves seem like an expensive option though

    Your right. Sometimes it pays to ‘Do it right, do it once’.
    There’s nothing worse that throwing bad money after bad money.

    @grim said: Can someone explain, in layman terms, what pressure drop really means?

    I’m not sure if this answers your question in laymans terms but I like to use the electricity analogy.
    Think of pressure as voltage potential.
    The pump is like a battery.
    It raises the potential.
    Put it in a circuit and you will have flow (current)
    Any gear you have in that circuit will have resistance to flow, including the line itself.
    Just like measuring the voltage drop across a resister, you also have pressure drop across a condenser (or whatever else).
    The control valve is just a trim potentiometer.
    If it isn’t sized correctly for the voltage potential we have available we either get all the adjustment as soon as it is cracked or the other extreme is it’s never open enough for the flow we need.

    Pressure drop is like voltage drop.
    Without it, you get zero flow.

    The closer you get to zero (open, 0PSIG) the less you have to work with.
    That’s why I suggested leaving the supply pressure and adjusting the return.
    I was having similar issues with my system.
    With a shitty chugger pump you only have so much pressure to play with so you kinda need to get the most out of it.
    If you drop close to zero at the start just to get your valve working there's not much potential left to get through what everelse is down stream on the line. In my case a watercooler and a bit of elevation.
    This is also why I suggested earlier on that a better option is to install a flow transmitter and use that to control the valve via PID. Not the cheapest option but you have direct control of the property that actually matters in this instance.
    Only having control over the power to the solenoid leaves you very open to other factors influencing the flow rate.

    If you were designing an electrical circuit you wouldn’t just grab random size battery’s and resistors etc. When it didn’t work it wouldn’t be much of a surprise.
    With some of the complicated process circuits that everyone’s having a crack at it’s very similar.
    A bit more thought into design and sizing can save a lot of fuking around and money later.

    You could make similar analogy’s with heat transfer too. Zero deltaT, zero heat transfer.

    Is that what you were after grim?

    It's not a perfect analogy and there are some terms that are probably inverse which might confuse things a bit but it works for me.

  • All this talk of pressure drop confuses me. As far as i thought the pressure should be pretty much the same anywhere in a closed, level system and it's the flow that changes.

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

  • edited February 2016

    cV = flow rate at 1 psi and 60f

    As defined by

    Max flow = cV * SqRt(pressure / 1)

    So a valve with a cV of 1.6 and a supply pressure of 10psi will flow 5.06 gallons a minute. 5.06 = 1.6 * SqRt(10)

    At 60psi a 1.4cV flows 10.84gpm max open.

    A characterized valve with a linear response should see half the max flow at half open, etc.

  • edited February 2016

    DetlaP, not just the supply. That's the key.

    https://www.youtube.com/watch?v=wxGXHAleItM

    I've never heard of a PICCV, they look pretty trick.
    I found another video that contradicts what I posted but it was a hydraulics vid using a PD pump.

  • edited February 2016

    I just use a calculator (just don't get your units mixed up)

    Water Flow Rate through a Valve

    Burkert have a good one too.

    I just reread your post @punkin, I think I know what you meant.
    I might be complicating things but there are two components to pressure. Static and dynamic.
    Static is what I think your talking about. That's pressure head from elevation (like the mains).
    You also need to add dynamic head to get the total head. That's where all the back pressure and pressure drop crap comes in.
    If there is no flow then the is no resistance so zero dynamic head.

    Total dynamic head @ Wikipedia

  • @jacksonbrown fwiw - the voltage analogy is helping me...

  • The pressure drop across my dephlegmator is zero - gauge before and after - how do I interpret that?

  • It's proportional to flow rate.
    Fuck all flow equals fuck all pressure drop.
    Open the tap and see what happens.

  • :) I get what your saying. Let me think about it.

  • edited February 2016

    @punkin said: All this talk of pressure drop confuses me. As far as i thought the pressure should be pretty much the same anywhere in a closed, level system and it's the flow that changes.

    If you got your 50ft garden hose and put a differential pressure gauge accross it, high to the feed end and low to the discharge end, and put a pressure gauge accross the tap, high to the supply side and low to the to the feed side, turned the tap on slowly and watched both gauges, the pd accross the hose will go up as the flow increases and drop accross the tap at the same time. Pressure drop in any system is a combination of static resistance (physical size of the pipe) and dynamic resistance (the more flow the less area to cope with that flow. There wont be much of a drop accross the dephleg at the flows you are using, you would probably need a gauge that reads pascals to measure it. Condensers that have low flow are better suited to a full flow 3 way mixing valve setup. A 1/2" pipe wont present any resistance at 1 l/min. Thefore trying to make this work using a 2 way valve even with a really low CV is not really the way to go. Feel free to tell me I'm wrong.

  • A two way feeding into loop should be ok.
    Grim, continuing the voltage analogy if it were a multimeter you effectively measured the drop across a bit of wire so it should make sense that way. Move them the either side of the pump, get out your pump curve and have a play.

  • edited February 2016

    I’ve been lurking here for years but unable to post very much due to being in the middle east. I’m now back in civilization with my control gear in a shipping container on the way here.

    I purchased a Burkert 8605 for $30 on Ebay to drive a proportional solenoid valve.

    image

    My water pressure was 36psig, inlet water temp was about 18-28DegC, 4” Super Deflag.

    I picked up a used Burkert Valve 6022-A-5/64-FPM-BR also from Ebay. I believe the CV was 1.2 if I remember correctly.
    The controllability of it was amazing, never had any issues and could keep the cooling water outlet temp within 0.1DegC. These proportional valves work very well on the low end, you can control down to just a trickle.

    This is the only picture I have with me, the proportional valve is on the bottom left of the deflag. An emergency hand valve is on the bottom right (normally closed).

    image

    Notice at the top right of the Deflag there is a bronze T fitting, normally my RTD was plugged into that fitting, this way the RTD extended about 1/3 of the way into the Deflag. In the picture it is connected to the Vapor Temp location.

    The panel on the right is for temp monitoring and a P&ID for the cooling water.

    The panel on the left is the heat control panel.

    Cheers

    Setup.jpg
    800 x 519 - 60K
    8605.jpg
    202 x 196 - 7K
  • I've got six of them :D
    Best bit is you can set the limits in them and they have a built in PID on PWM output to maintain valve position. They're fantastic.

  • I would like to see a video of one of them working, they sound really cool.

  • Do they just convert proportional current or voltage to pwm or is there some kind of special sauce that does something additional? Linearity?

  • edited February 2016

    @CothermanDistilling said: I would like to see a video of one of them working, they sound really cool.

    https://www.youtube.com/watch?v=HQGEd4MRffo

    They take all the main input types then convert it into a PWM signal that powers the coil. A bit like a dimmer for an LED.

    I looked into it and from what I understand you can actually control the valves directly without the control modules but the frequency range for each valve type is very specific and they draw a few amp which is obviously a lot more than a standard analogue signal output on a microcontroller.
    I think they do a heap of other tricky stuff too that is specific to each valve type.

    My understanding is that they also have some kind of feedback involved with the valve stem position and also some sort of compensation for coil heat as well.

    Have a read.

    Type 8605 - Control electronics for solenoid control valves
    8605 Control Electronics for Proportional Valves (PDF)

    Group buy?

    I've got some spares ones of these.

    8mm is probably a bit big for small guys on mains pressure though.
    Good thing with Burkert is they go right down to 1 or 2 mm size range and also come in Explosion-protected coil option if that floats your boat.

    I think this one is 3mm which is probably a more sensible size.

  • edited February 2016

    I controlled the 8605 with a 4-20mA signal from my PID. The 8605 converts it to a fixed frequency DC signal that varies the duty cycle of the signal in proportion to the 4-20mA signal.

    Change the duty cycle = change of power = changes the amount it opens.

    Every type of solenoid valve needs a different freq and consumes a different current. The valves themselves are dead simple, there is a coil that the fixed frequency signal is applied to. If you take it apart there are less than six pieces to the valve body itself.

    As the coil gets hot the resistance changes, the 8605 looks at how much current it is applying and as it changes it adjusts to keep things more linear. I think this is why the controllers are so bloody expensive. Asco also makes a good controller but I could never find a used cheap one. They seem to have sold a shitload of Burkert valves based on how many used ones are on E-Bay.

    Linearity:

    image

    image

    1.jpg
    334 x 215 - 11K
    2.jpg
    389 x 298 - 21K
  • OK, I was going down the wrong idea path, when you said 'solenoid valve', I though this controller could hook up to an existing solenoid valve coil.. those regular solenoid valves seemed pretty 'light switchy' in being on or off..

    but these are still cool... and faster than my johnson... lol

Sign In or Register to comment.