My attempt at controlling the Dephlem

[ShiFu]

I've been piddling with Law-of-Ohms on an idea to enhance my build.  So boxing will probably be delayed.

A couple of years ago I commissioned Olddog to build a controller for my 5" still. It worked well but the electric water valves for regulating the temperatures of the dephlegmator and product condenser drove me nuts. The constant slamming on and off of the electric valves (and the problems of overshooting the presets) were a constant irritation. But there was simply no affordable solution to be had. We're trying to fix that.   

So we're considering the marriage of a needle valve with a stepper motor, along with some Arduino PID code, to smooth things out. I envision being able to set a target ABV on the touch screen and the still will do its best to produce exactly that ABV by adjusting the water flow to the dephlegmator. It should not be a great leap as the Arduino already knows the temperature/ABV of the product.   

Details are still a bit sketchy but I've ordered lots of new parts to play with. Including a couple of sizes of stepper motors and drivers to see if the cheaper one will work or if the more expensive and powerful motor is needed. Also a GRBL Arduino shield to see if it can replace the more expensive stepper drivers. 

We've already determined that this drawing will not work as shown below but it is the groundwork that we are working on. A few modifications and some testing to see which stepper/valve/driver combination is the one that delivers the results at the lowest cost and then we can take some measurements to get a few of these brackets fabricated. 
I'm pretty sure these will be attached close to the controller instead of bolted to the still.   

My attempt at controlling the Dephlem
bracket render.JPG (654.7 kB. 350x196 - viewed 2760 times.)

[ketel3]

This is a other way as Istill is using they have a controled needle valve for controlling the take of speed.
I never came on the idea to control the reflux cooling this way.
For the LM it is not going to work I think but for vm cm or plated colums its a great way to control.

What a terrific ideas I have seen overhere its amazing.
cheers K3

[weddy]

I really like this topic
The stepper motor you use on the picture is the same one as i did for my 3d printer.
There are free programs for the arduino and the stepper motor so maybe you can steal some code of it.
I'm a complete noob whit coding so that's what i should do hahahaha
the stepper motor is pretty strong but most of the time they use a transmission to make it more accurate.
So the motor will run 10 turns for 1 on the valve .



Edwin

[ShiFu]

This is a other way as Istill is using they have a controled needle valve for controlling the take of speed....

How does Istill control the needle valve?
I can only assume the valve needs:
1) a motor
2) PID (software) to control the motor and make it do what we want it to do
3) a micro-controller (like Arduino, but there are other platforms) to execute the code
4) a driver, to link the motor to the micro-controller and supply the electric current that would otherwise destroy a micro-controller.

If there is something out there already that can be used for our purposes please let me know. 

I don't know much about an Istill. Is any of it open source that we can use? 
I'm not trying to make copies of anyone's products to sell. I'm trying to make the best possible still for myself without giving someone thousands of dollars. And around here, we share everything that we learn along the way.
All the code, drawings and ideas are free, we don't sell anything on this forum. Not yet anyway***. 
We try to get the most that we can get with the least amount of money. Nothing is free, even free code and free drawings require some work for the user.

*** Disclaimer...  I do vend distilling hardware to a few folks that usually re-sell them to others for profit. But that is another website. 

[ShiFu]

Well, I never said this would be a smooth road and mistakes are pretty much guaranteed. 

The first of the parts arrived and a quick fit-up showed I ordered the wrong part.
The coupler needs to be longer. Or cut the end of the valve stem off. 

At this stage its probably better to order the correct parts instead of needing to modify anything. 


My attempt at controlling the Dephlem
coupler one.png (246.24 kB. 350x262 - viewed 2228 times.)

[ketel3]

How does Istill control the needle valve?
I can only assume the valve needs:
1) a motor
2) PID (software) to control the motor and make it do what we want it to do
3) a micro-controller (like Arduino, but there are other platforms) to execute the code
4) a driver, to link the motor to the micro-controller and supply the electric current that would otherwise destroy a micro-controller.

If there is something out there already that can be used for our purposes please let me know. 

I don't know much about an Istill. Is any of it open source that we can use? 
I'm not trying to make copies of anyone's products to sell. I'm trying to make the best possible still for myself without giving someone thousands of dollars. And around here, we share everything that we learn along the way.
All the code, drawings and ideas are free, we don't sell anything on this forum. Not yet anyway***. 
We try to get the most that we can get with the least amount of money. Nothing is free, even free code and free drawings require some work for the user.

*** Disclaimer...  I do vend distilling hardware to a few folks that usually re-sell them to others for profit. But that is another website.

Hi Shifu,
No Istill is not open source.
Iám not shore how they do it,I think the opening of the valve is adjusted by temperature control compensated with a pressure sensor , they speak about opening 18   opening 19  and so on..
Further you can choose the distill methode you want to carry out and the software makes the right choise according to the valve opening for that special program,so you can do potstill or lm  all depending and controled by the micro + valve,.

  Iám shore Odin does not want to sheare his knowledge on that part ,about distilling he is very open and helpfull

cheers have work now,

[law-of-ohms]

The router gets fired up tonight for prototype version 0.0000000000000000000000001a

[ketel3]

The router gets fired up tonight for prototype version 0.0000000000000000000000001a
[/quoteHi LOO can't wait to see it.

Now I have to work

[ShiFu]

Hi Shifu,
No Istill is not open source.
Iám not shore how they do it,I think the opening of the valve is adjusted by temperature control compensated with a pressure sensor , they speak about opening 18   opening 19  and so on..
Further you can choose the distill methode you want to carry out and the software makes the right choise according to the valve opening for that special program,so you can do potstill or lm  all depending and controled by the micro + valve,.

  Iám shore Odin does not want to sheare his knowledge on that part ,about distilling he is very open and helpfull

cheers have work now,

So they are not open source and are useless to us, very typical.

",I think the opening of the valve is adjusted by temperature control compensated with a pressure sensor , they speak about opening 18   opening 19  and so on.."   

That somehow says everything and nothing at the same time. Again, so typical of greedy closed source. 


If that is all that Istill offers for the big money they want then they can keep it locked up.

We will work our way through this so everyone can create their own at very low cost.
I'm not discounting their work that they want to make big money on, I'm saying I don't want to pay it. 
And when I get mine going I share all the details for free, as always. 

I buy equipment for a customer. He pays 1/2 before I order it and the other half before I ship it. For that I charge 30% markup for my time.

I buy equipment for a customer that refuses to pay even the deposit until the goods are delivered, then our markup is AT LEAST 100%. 

The customers that pay the 50% deposit have a perfect track record of paying me. The customers that do not pay the deposit have a terrible (stealdragon) record of not paying so we can no longer ship goods that way. 

I do not care much for any company that is closed source and try to avoid dealing with them. I often buy parts from open source companies because they supply the code and drawings freely while providing a good deal on their hardware. But when everything is closed source you are the slave of their (inept) customer service. 
Screw that!

There seems to be a relationship between customers that pay (or not) and companies that are open source (or not).
And the good guys suffer because a few bad guys screw up everything for others.

I say, forget them! 
Lets make the best still that can be made and if they steal it (yes, they absolutely will) then so be it. 

[law-of-ohms]

Here's a DIY trick to making brackets without a break press....

Add routed slots along the fold lines and then bash them over in a vice.

I love it when shit just works first time.

But, as shown in the last two pics, every man and his dog already has a push connect needle valve.

I think we should develop the bracket kit for a standard push connect, what y'all think?

My attempt at controlling the Dephlem
2016-09-14 20.28.29.jpg (1100.64 kB. 350x196 - viewed 2601 times.)
My attempt at controlling the Dephlem
2016-09-14 20.28.35.jpg (960.96 kB. 350x196 - viewed 2234 times.)
My attempt at controlling the Dephlem
2016-09-14 20.30.17.jpg (1194.66 kB. 350x196 - viewed 2278 times.)
My attempt at controlling the Dephlem
2016-09-14 20.30.22.jpg (1031.55 kB. 350x196 - viewed 1710 times.)

[ketel3]

   +1 Shifu,


But Istill is not  for hobby use in the the first place,and I only mentioned the differance between what they use and your idea how to control the proces  by controlling the ammount of  reflux,and your idea was new to me.

I did not know about your bad experiances as I came to this forum ,and meanwhile I did some reading so I can understand your point of vieuw:)
There are bad people around and very bad people,and sometimes they come into your life.

But back to the point your idea will work on Vm and plated colums  , but my question wil it work on LM I think not as the proces is controled in another way,but lets see maybe your system is able to control the output needle valve of my VM also.

I had another idea of controlling the vm output by a small dosing pump , arduino shows on youtube some project with it and it is able to control drop by drop till full open,unfortunately my coding is bad,and the hose used in it has to be silione.

Ok lunch is almost over,see you around

[ShiFu]

Sorry for the rant. It seems to pop up uncontrollably from time to time... it can happen when people (stealdragon) steals from you 

I am personally not very good at LM (liquid management). I found it frustratingly too slow vs EVERYTHING else. 
I prefer VM (vapor management) with a dab of PM (power management) which is the essence of a bubble plate still. 

If running a scrubby packed column, I prefer VM for take off speed. LM is just, to me, SO PAINFULLY SLOW.

Like you, I doubt that having the dephlegmator in precision mode via a PID controller will benefit a LM column.
But thinking my way through, I think maybe LM could actually benefit form the tighter control. More real-life experiments will be needed to prove this as right or wrong.     

[ShiFu]

My attempt at controlling the Dephlem
holycrapThatIsIt.png (875.97 kB. 350x163 - viewed 1528 times.)
Click on the box above to view.   

And then, out of the blue, Law of Ohms comes up with this gem!!!   

The Holy Grail of controlling condensers.  VERY WELL DONE! 

(You think the motor is overkill, Dan?) 

This solves a few problems and makes everything much cheaper to implement. Kudos, Sir LawMan, for thinking up such an ideal solution.  The push needle valve torque is WAY less than the 3/8" stainless needle valve so the stepper and its driver can also be much less powerful and expensive.
The push valve is "friendly" for us to easily supply the condenser(s) with plastic tubing vs adding push fittings to the stainless valve. This save a few fittings and that expense. 
I'm Really Liking This. 

[Myles]

ShiFu, I know there are many ways to run a plated column and often they are used in combination, but can I just clarify why you wish to control the dephlegmator temperature. Just wish to get it clear in my head what you are trying to do. For me the plated column is mostly a CM device.

Are you trying to control both boiler power and dephleg temp, to get your target ABV?

What I think of as a generic mode of operation, is to calculate the max power to the boiler that your dephleg can handle, and then to set the boiler at a %age of that. Anywhere from 80% to 100%, and control the run from that point onwards with mostly the dephleg temperature. (Possibly with different power set points for the different phases of the run)

The problem is that different recipes will have different values for max power, and it will also be influenced by ambient conditions. You can offset that by averaging it out and always running at less than max power, but that might not suit some folks.

Setting the dephleg to a specific temp (lets say for argument 80°C, sort of like you might do on a packed column) and then controlling with boiler power is not something I do, but it can be done.
You can also control dephleg temp to get a constant reflux flow rate into the column, but I don't really bother with that either.

The product temp is a different issue and I would try to control that independently from the dephleg, if at all.
To be honest I used to adjust it to match the calibration point of the glass hydrometer, but if I was using an e-parrot I might just max out the PC coolant flow and hope that the product temp would be fairly stable anyway. As long as the e-parrot knows what it is, the actual product temp may not be that important.

What is your end game with regards to dephleg temperature?

[ShiFu]

End game...  control the still (especially the dephlegmator) to produce the desired %ABV that I enter into the touch screen. 

Side benefit of doing so will also allow me to control the coolant to the product condenser so I can maintain a "normal" product temperature without wasting too much unnecessary cooling water (again, PID).   

Not saying I can do this now. This is my goal.   

[Edwin Croissant]

The iStill as I have seen it at Odin’s place some years ago was a re-branded Polish still from destylaty.pl. This website now redirects to g-still.com. It looks like the business relation between Destylaty and iStill ended as Odin sold his stock of imported stills and is now developing his own stills based on rectangular boilers and automation with PLC’s.

The needle valve controller in the Genio still, as they call it now, uses a stepper motor and a valve position feed back (I have seen it but can’t remember if this was a multiturn potentiometer or a rotary decoder)

To get some idea’s do an image search on Google on the words: diy servo water valve for some examples for pinch and quarter turn valves with RC servo’s.

Hope this is useful 

[ketel3]

The iStill as I have seen it at Odin’s place some years ago was a re-branded Polish still from destylaty.pl. This website now redirects to g-still.com. It looks like the business relation between Destylaty and iStill ended as Odin sold his stock of imported stills and is now developing his own stills based on rectangular boilers and automation with PLC’s.

The needle valve controller in the Genio still, as they call it now, uses a stepper motor and a valve position feed back (I have seen it but can’t remember if this was a multiturn potentiometer or a rotary decoder)

To get some idea’s do an image search on Google on the words: diy servo water valve for some examples for pinch and quarter turn valves with RC servo’s.

Hope this is useful 

Hoi Edwin,
Yep your right it was a Polisch brand in the first place,you can see movie's of the Polisch version on youtube,the same brand as Istill.
How they do the control in my opinion for differant kind of destilations just by the control of the needle valve , only pottstill needle valve full open. VM  with controled opening , but that is not all , the build in program can control fores heads tales and stop distilling pure on the programmed temperatures,pressure compensated and compairing the temp on the colum bottom to the colum top.
And even change the colum opening automaticly during the run to keep the abv high,and if I have read well between the lines of al postings extra stabilisation during the run automaticly  if needed.
So it is a quit good design,I think.

But all the projects here are showing great steps forward,for do it your self still control,the idea of e parrot is now a step further with controlling the proces.
The resent step of LOO and Shifu with the stepper gives hope for LM control in future too.
Let see how thing develope   

[Edwin Croissant]

I have been thinking about controlling the dephlegmator. Now I have no experience with this kind of reflux control so the following may not make any sense whatsoever

When you want to control something with a lot of inertia you need to cancel out the inertia factor. For example controlling the lift of a helicopter with a fixed blade pitch rotor is very difficult, as changing the lift is only possible by changing the rotation speed of the rotor, and due to the inertia of the rotor that takes time. Controlling the lift by changing the pitch of the blades is instantaneous as the rotor can keep a constant speed.

I expect that controlling the reflux rate with the coolant flow is a slow process as there is a lot of thermal inertia in the water content of the dephlegmator. So the obvious way to control the cooling capacity of the dephlegmator is by controlling the water level inside the dephlegmator. I think this is the way the old goose condensers were controlled, by varying the water level in the cooling tank.

A brief search on Google revealed the myles  came up with this idea about 4 years ago.

So what I propose is the following system:

Water supply is connected to a thermal control valve with a remote sensor (Danfoss produce these things) to keep the cooling water output at a fixed temperature. This will minimize the amount of water required.

Output of this valve is connected to the product condenser business end. The other end of the product condenser is connected to the top of the dephlegmator. The bottom of the dephlegmator is connected to a low pressure check valve to create some back pressure in the system. This is also the point the remote sensor ot the thermo valve is connected. Other end of the check valve is connected to the drain.

Dephlegmator control is accomplished by manipulating the air pocket in the top of the  dephlegmator with a low pressure air pump (aquarium air pump?) and a small solenoid air valve. Put a level gauge in parallel of the dephlegmator to observe the water level.

As the flow through the condensors is constant the reflux ratio can be calculated with the temperature drops over the condensors. So the inner control loop keeps the refux ratio constant by letting air in or out of the dephlegmator and the out control loop keeps the vapor temperature constant by manipulation the reflux ratio.

I hope this brain fart make sense

[Myles]

You can try this but you may have problems.
You might also try changing the temperature of the water going into the dephlegmator, and maintaining a constant flow rate. Warmer going in means lower deltaT and less energy removed and control over internal temp.

Having air in the system complicates things but there are ways to have variable depth AND have the condenser open to atmosphere. Usualy involving an external reservoir on the dephlegmator.

In general a full condenser and either variable flow or variable temperature input are more controllable options.

[ShiFu]

Thank you all for the great replies and ideas.
The condenser is actually quite responsive to any changes and the lag time between "water off" and "water on" can be seen with temperatures changing in only a few seconds and stabilizing within a minute or less. 

Danfoss temperature control valves are very expensive here, the best price I've seen is over $300 USD, ouch.

I'd mentioned that Olddog built a great controller for me but it had solenoid valves (fully ON or fully OFF) and operating with a PID controller it actually did work pretty well. Not perfect, because of the inrush of cooling water when the valve was activated, but the controller would absolutely cause the still to stop producing when the heads were depleted. Then the PID was set to a higher temperature and the hearts were taken until they were depleted. Finding those temperature settings was a slow process but it was nailed down during the very first run.

The biggest problem was the fully ON or fully OFF nature of the solenoid valves. The constant clapping sound of them cycling on and off does get irritating but even worse was the temperature swings within the condenser.
The product came out in drips, then a nice stream, and back to drips as the PID cycled the water On and Off.
So I CAN state that PID control of solenoid valves DOES work to control what the still produces. I can also state that there is great room for improvement.

Far better would be to have PID control of a valve that could be set to any flow rate between the On and Off extremes. And that's where the stepper motor comes in. Making very tiny adjustments to the coolant flow instead of fully On or Off would be a huge improvement.