Originally By Tony Ackland
Please note that the information contained on this page is to be considered largely outdated. There are far better methods being used these days.
For better information, please consult the Homedistiller Forums where heating and electronics are thorougly researched and discussed.
Heating Current Control Systems for Distillation
Peter summarised the two alternative methods of control required..
When you have a reflux cooler on top of a packed column outside, you have to regulate the flowrate of the reflux-distillate back into the packed colunm, to get constantly 94 % alcohol during the whole proces. The flowrate is regulated by the column top temperature.
I think the best solution is that suggested by Robert ...
http://www1.jaycar.com.au/ Indoor / Outdoor Thermometer with Memory & Alarm Cat. No: QM7212 you could also try Dick Smith or Tandy if Jaycar is not in proximity to you or not to your liking.
Probably what is of benefit to the lay distiller however is the control over the gross heat input. Given the long time it takes to first get the wash up to distilling temperature, there is some benifit in having a variable power input - so you can give it heaps to get it up to temperature, but then back off the power once you're distilling. Two ways of doing this;
Mike offers a simple way of getting variable heat input from just two elements, and a couple of switches :
1. Wmin = W1.W2 / (W1 + W2)
2. Wa = W1
3. Wb = W2
4. Wmax = (W1 + W2)
Be careful to rate the switches correctly. I = TWICE Wmax / V (twice the actual maximum current is good engineering practice) Taking the maximum wattage (in this case 3200W) this means Imax = 13.4 amp at 240V or 26.8 amp at 120V so get switches that can handle 30 amp (240V) or 60 amp (120V)
Implicit in this diagram is that the main power switch at the bottom incorporates a suitable quick-blow FUSE and, preferably, a neon to show when it is closed. Also implicit is that the element casings and the boiler are connected to a good EARTH.
TRIPLE CHECK that all connections are correct before connecting to mains power! Test first with a dry-cell battery and small lamp bulbs in place of the elements.
DISCLAIMER!!!!! I accept no responsibility whatsoever for anyone screwing up and electrocuting themselves, their stills blowing up and burning their house down, floods, pestilence, war, lightning strikes, or any other act that the capricious gods of distillation may visit upon any hapless distiller using this circuit.
Mike explains the difference between triac and transformers/variacs, and thyristor/triac circuits
A fixed multi-voltage transformer has one primary coil and several secondary coils, all with the number of turns needed to produce the required secondary voltages. A variac has one primary coil and one secondary coil, and the number of 'effective' turns on the secondary is varied by means of a sliding contact that 'taps' the secondary coil along its length. By this means, you can control the output voltage in tiny steps, each equivalent to one turn.
The important thing to remember with transformers is that the output voltage waveform looks exactly like the input voltage waveform, varying only in amplitude. It is therefore a 'smooth' voltage.
If you want to be precise (and confused), the secondary waveform also lags behind the primary waveform depending on the loading ... the lag increasing as the current load increases, and how inductive the coil is. This is what they are talking about with 'power factors' of motors, as the motor's spinning coil not only acts like a secondary coil in terms of extracting mechanical energy from the imposed primary field of the static coil, but also acts as if it's the primary of a transformer with the static coil of the motor as the secondary. The main supply therefore has a generated voltage 'reflected' back into it which is out of phase with it, and this can adversely affect the power supply if not compensated for. Luckily for you, as you will be dealing with a stationary load (a heating element), you can forget all that garbage! :-))
A thyristor is essentially a diode that passes current in only one direction. However, if you apply a brief 'trigger' voltage to a terminal called the 'gate' during the time that the thyristor is blocking current flow, then the blocking effect will be turned off until the applied voltage falls to zero. So a thyristor will always pass full current in one direction (forward bias), but can be switched from zero to full current in the other direction (reverse bias) at any time you choose. You can apply a steady 'trigger' voltage and have the thyristor act like an ordinary piece of wire, passing current both ways all the time, or remove that 'trigger' voltage and have it act like an ordinary diode. Alternatively, you can get cunning and apply the 'trigger' at the same time during each reverse bias cycle and have the thyristor switch off its diode blocking action at the same time during each cycle. As you would then be switching the thyristor in time with the phase of the applied voltage, this is called 'phase control'.
When you trigger the thyristor in the middle of a reverse bias cycle, then you get an almost instantaneous jump in the output voltage from zero to whatever the applied voltage is at the time. This very sharp 'ramp' generates high frequency 'harmonics' that can be so high as to reach radio frequencies. When this happens, your neighbours start swearing as their radio and TV start to be swamped with your 'jamming station'. This can be controlled with appropriate circuits that absorb and damp high frequency currents.
A triac is essentially two thyristors wired in series, but pointing in opposite directions. When the applied voltage is in one direction, one thyristor can pass current, but the other can't, and vice versa, so no current flows. They have a common 'gate', so you can control both at the same time. Apply a steady 'trigger' voltage to this 'gate', and current will flow freely back and forth all the time. Apply the trigger regularly at the same time each forward cycle, and you have got a device that acts like a single thyristor. Apply that 'trigger' voltage at the same time each forward AND reverse cycle, and you have control over the full cycle of the imposed supply voltage, both forwards and backwards, and you have in effect doubled your efficiency. It is also possible to design circuits that trigger at one point in the forward cycle and another in the reverse cycle, but that is getting into the realms of sophisticated applications such as radar pulse control.
So there you have the essential difference between transformers/variacs, which produce smoothly varying sinusoidal voltage outputs (and so need no EM suppression), and thyristor/triac circuits which produce 'sinusoidally challenged' outputs that have bits chopped off the sides of their waveform, so they end up looking like the teeth of a saw (and can radiate like mad is not firmly squelched).
Pilch writes ..
PID controllersJohn wrote ...
Now, when you go to the web site, you will see many different options as to what you can purchase. This is where I recommend you to go to the originating web site of the article:http://hbd.org/kroyster/
Once you get there, click on the link on the left hand side of the page that says "Design Details." This will take you to a page that explains his "Recirculating Infusion Mash System." If you scroll down toward the bottom of the "Design Details" page, you will find the author's details about how he used his unit. NOTE: the link that the author has on his web site to the manufacturer of his PID is not correct anymore. They, Omega, has re-arranged their web site a bit. The link that I have posted at the top is the corrected page.
I use a controller from a company called "CAL Controls". I purchased one new several years ago for about $120. Two months ago I purchased the same model from a surplus place for $50. They're out there if you keep looking. I imagine that any temperature-type PID controller would do the job.
..I use mine in a RIMS (that's another TLA... a three-letter-acronym ;-) ) That means "Recirculating Infusion Mash System". When I do a mash for beer brewing, the PID controller monitors the mash temperature. The wort (same as the wash in distilling) is recirculated from the bottom of the mash tun, through a pump, past an in-line heating element, past the thermocouple, and back into the top of the mash tun. The controller has an "autotune" or "learn" mode. When this mode is activated, the controller turns on the heater full-blast for a short time. It then monitors the rate of temperature rise and the resulting temperature achieved in a specific amount of time. It then "knows" how your system behaves. That way it can turn on the heat to get a maximum temperature in the shortest amount of time, and then start turning the heater off and on (very quickly) to control the approach to the setpoint. This results in reaching the setpoint as quickly as possible, but not going over the setpoint. Then, when the temperature begins to drop (normal cooling of the system), the controller gives the heater just enough power to keep the temp where you want it. You should expect to be +/- 1 degree C from the setpoint. In a still, this means that when the alcohol is about gone and the temp starts to rise, your heater will turn off and the distillate will stop. It will, however keep the remaining wash at the setpoint, so some water vapor will still be condensing.
Check these out:
Also - there is a "Manual Heat %" mode on these controllers that will allow you to set the heater power as a percentage anywhere from 1%-100%. This mode uses no thermocouple. So, if you have a 2KW heater, set it for 100% for heat-up, then crank it back to 50% (for 1KW) for the run. This would make far more sense than trying to control the vapor temp, 'cause it's going to boil at whatever temp the ethanol/water mix allows...
For a really good tutorial on thermocouples, go see: http://www.tinaja.com/glib/muse144.pdf (This web site is amazing! - go read everything there - you'll learn a lot!)
The controller has only one thermocouple input. There are 2 outputs from the controller. One is relay contacts (slow operating), and one is the "contacts" of a solid-state-relay (SSR).The SSR is used to control a mash or still heater.
.. I did some checking and found a web site where a guy built and programmed his own RIMS controller. However, it's not PID, but it probably would work if the code were changed some. He provides a schematic and the code to program a BASIC STAMP. Check it out here: http://chattanooga.net/~cdp/rims/rims_inf.htm
Also do a search for "RIMS Controller" on a search engine. eBay has quite a few temperature controllers for sale.....
The idea came to use a stove element. By measuring the resistance in the stove element and adding it to the resistance of my 2400W heating element in the formula P=(I^2)*R and V=IR. I found I could cut the power in half, or in thirds, depending on what size stove element I used. (Stove elements are almost free in 2nd hand stores). So My problem was solved very quickly and VERY cheaply.
a)This system makes it very easy to electrocute oneself, so electrical knowledge is ESSENTIAL. (It's only a simple circuit but it can kill very easily if stuffed up).
b)The stoveplate gets very hot and needs to be in a spot where it can't get trodden on or have stuff dropped on it.
c)The element needs to be in a ventilated area, any open space will do
Apart from that it's all hunky dory, it's not pretty, or clean, and has a terrible power factor, but it works extremely well and can be made in 5 minutes.
I hope you find that an original idea to a common problem. I have built a triac controller now, but in the beginning this was my original method.
Andrews MicroprocessorAndrew writes ...
DavidDavid Reid writes ...
If you dont want to do that because you cant afford it or are too tight to spend the money at least go and rip a decent simmerstat out of a stove from an inorganic rubbish collection round and at least use that.
You definitely need some type of control over the heating system. The chances of it being exactly spot on are almost negligible; it will more likely than not be too hot at times which means it needs turning down a bit. (Too cold and you could have done with a bigger element for a startoff but 750w should be adequate apart from bringing the mixture up to temp. Too hot and you are boiling the guts out of the mixture stuffing up the separations).
Remember you get what you pay for.
I use a solenoid valve to turn the water off and on and a manually adjusted needle valve to control the flow. Please note temperature at takeoff point is different to temperature in boiler, and can vary depending on still. This is a point where there is a bit of disagreement over location of sensors. Some people argue that because you want to control the temperature in the boiler which is what controls most of the other temperatures in the column etc that this is where you need the sensor and work from that point. To a degree they are right. I and most people who know what they are doing adopt a slightly different attitude knowing that the most critical temperature point in any still is at the takeoff or condensation point. I therefore mount the critical sensor there working backwards using insulation and other aspects of the design process to minimise the differences between temperature at this point and the boiler. The ideal situation is where you have 2 or 3 sensors each monitoring different points but then the controls and the monitoring situation become a lot more involved.
The temp sensor must not touch the cooling coils. You need to measure the vapour temp at this point. Please note that water vaporised expands to something like 1500 times the volume. Location and depth of sensor is critical. So many of the people designing stills dont have the first clue and there is so much false information out there. Before going further and designing a controller that might work go out and research the subject properly and read a few decent books. Most of the people designing stills are too lazy or too arrogant to do even that. If you can see how someone else achieved a certain goal you can admire it or fault their thinking. Quite often you can even improve upon ther original concept. Firstly you need to separate the chaff from the wheat. That way you may design a controller that works. PIC controls certainly have their place and proper intergration of these in the stills of the future will be commonplace and is one of the better solutions available.
Jan-Willem's TriacJan writes ...
I don't know if you can build it yourself, if you do on my site is a very simple triac controller, and that one can do 2.5 kW. Its a very basic straightforward controller. I experimented with a proportional controller with temperature feedback but that was a bit overkill in my opinion. But than again i don't run a MEGASTILL with continuous output etc etc Its just hobby and when its running I am VERY nearby...
You can find my homepage (if my provider is still running) at http://home.planet.nl/~jwdob/ then go to distilling
Ross's Time-clocksRoss uses time-clocks to help with the long distillation times involved with operating a high-puirity Nixon/Stone still ...
Now I have a solenoid water valve on a time clock which refills the 200lt drum every 12 hours. This is kinder on me and the pump. Running time clocks and solenoids on the both stills, I can start the stills at 1am and the SS unit has dropped about 6-7lts of 60-70% impure and [my still] has just equilibrated nicely by 7 am. The whole system is very user friendly.
It is a little early to get accurate times yet, but 5lts of water (to cover element) and 20lts of 60-70% impure takes about 48-50 hours before the temp rises above 80C and I turn off
Smithers TriacSmithers (of http://go.to/distil), who is a electronic technician by trade, suggests the following ..
DO NOT ATTEMPT TO BUILD THIS CIRCUIT UNLESS YOU ARE QUALIFIED OR EXPERIENCED WITH MAINS VOLTAGE. THIS CIRCUIT INVOLVES SWITCHING AN ELEMENT THAT CAN DRAW 10AMPS, IF YOU FUCK UP-THIS WILL KILL YOU. YOU CAN BUY CONTROLLERS LIKE THESE (COMMERCIAL) FROM ELECTRONIC SUPPLY SHOPS, BUT FUCK THEY'RE EXPENSIVE.
I ran with the I.C controlled (TDA1023) without using a temp probe. The specs on this (and application data) can be found at http://www-us2.semiconductors.philips.com/cgi-bin/pldb/pip/TDA1023#applications (or download it from me here (267 kB).
You will need to download the above PDF file (page 14,15) to understand what I am talking about below.
There were a few changes that I made for safety and to compensate for the 240V 10Amp supply that we have. They are as follows :
I also placed a neon indicator across the load to give me a visual on when power was applied to the element
The Triac requires a fair amount of cooling, I used a finned heatsink (100mm x 45mm x 45mm) this keeps it reletively cool, I also used plastic screws, heat compound and a mica washer to electricly insulate the triac from the heatsink.
The Resister RD dissipates approx 5W, so I advise that you heat sink this as well, this item does get fucking hot if you don't. Mount it all in a well ventilated preferrably plastic box
That is about all I have to say about that. It works really well and I have used it faultlessly for about 25 (6 hour) runs now. The neon indicator is a good idea and lets you know everything is working ok.
Andrews Temperature Control CircuitAndrew Graham supplies details (diagram, parts list, and explanation) for how to build a cheap, simple temperature controller at http://www.shortcircuit.com.au/EVCA/tcc.htm.
Reimas controllerReima writes
Here in Queensland you do not need any more if you do not have a lake of your own for cooling water.
I have also a teeny weeny 100W cartridge element in the barrel, this is hooked up to an ordinary dimmer switch, so this gives me 600 + (100W with stepless adjustment for fine tuning).
PS. Dimmer switches can take up to 300W for fine adjustment-heater and would be better wintertime, but you take what you have ;-)
See schematics below (click to enlarge).
Neils phase angle controllerNeil writes ..
Controlling Gas fired BoilersCraig offers :
The calculator is programmed with the following equation:
Ti = Condenser Inlet Temperature °C
To = Condenser Outlet Temperature °C
Mr = Condenser H2O Mass Flow Rate Kg/sec
Cp = 4200 (Watts · sec)/(Kg · °C) Specific Heat of H2O @ approximately 50 °C
The specific program for the HP 48SX calculator is:
Set up the variables "Watt", "Ti", and "Flow"
Key in < < Ti-Flow*4200*60 / > > and store it in "Watt"
Store the inlet temperature in "Ti"
Store the flow rate in "Flow"
Enter the outlet temperature into the display
Press the Watt's key to run the program
The watts should display
Note: The flow, even though measured as a volumetric rate, closely equates to a mass flow rate at the temperature ranges involved (1 Liter H2O ˜ 1 Kg H2O)