A digitally-controlled-power power supply

If it plugs in, post it here.

Moderator: Site Moderator

Maritimer
Rumrunner
Posts: 716
Joined: Sun Jul 01, 2012 6:09 am
Location: Nova Scotia, Canada

A digitally-controlled-power power supply

Post by Maritimer »

In my thread Instantaneous RR Measurement Experiment 1, http://homedistiller.org/forum/viewtopi ... 17&t=32502, I found that my algorithm for controlling the power using a random function unfortunately did not work very well.

So, here is a new algorithm:

> There are 1024 levels of power desired.
> There are four 1000 watt elements
> There are four zero-crossing solid state relays to control power to the elements. Such a relay passes one or more half-cycles of voltage to an element.
> 1024/4=256. So assign 256 possible ON half-cycles to each element. If 1000 watts were desired, for example, each element would be on for 256/4=64 half-cycles.
> 256 half-cycles takes 256/120 = 2.1333 seconds
> Space the four elements apart in time by 64 half cycles = 0.5333 seconds

This is all controlled by a Zilog eZ8 Encore! development board (cost about $43)

In the main() function, we read the pot from the analogue-to-digital converter, getting a value between 0 and 1023:

int power_setting;
unsigned char elements_123, element_4;

power_setting= (store_adc_values[0]); //read pot from ADC
elements_123 = power_setting/4;
element_4 = power_setting - elements_123*3;


That last element_4 just makes up for power_settings that are not multiples of 4.

Now in the file interrupt.c, do an interrupt every half-cycle:

extern unsigned char elements_123, element_4; //these were defined in file main.c

#pragma interrupt
void isr_PA7(void) //the half-cycle is pulsed on port A pin 7
{
//This interrupt will turn on and off the four elements
//The first 3 elements all stay on for the same elements_123 1/2 cycles
//The fourth element stays on for the time of the first three plus 0 to 3 extra 1/2 cycles

static unsigned char counter1 = 0,counter2 = 64,counter3 = 128,counter4 = 192 ; //stagger the elements by 64 pulses


if(++counter1 < elements_123)
SETBITPE(0x10); //turn on element 1
else
RESETBITPE(0x10); // turn off element 1

if(++counter2 < elements_123)
SETBITPE(0x20); //turn on element 2
else
RESETBITPE(0x20); // turn off element 2

if(++counter3 < elements_123)
SETBITPE(0x40);
else
RESETBITPE(0x40);

if(++counter4 < element_4)
SETBITPE(0x80);
else
RESETBITPE(0x80);
}


By measuring the temperture difference across the Liebig on a pot still, the power is read directly by the temperture difference.

Here is a single 1000 W element. The trace is the temperture difference of coolant on the Liebig:
1000 Watts Single Element.jpg
And here is 1000 Watts by pulsing the elements sequencially:
1000 watts algorithm.jpg
And here is 2000 Watts by the algorithm:
2000 Watts on pot still condenser.jpg
Note that the digital command does not set voltage, but actually sets the power. So we have a digitally-controlled-power power supply!
Soggy Bottom Boy
Rumrunner
Posts: 746
Joined: Sun Sep 09, 2012 4:43 pm
Location: Somewhere to the north of the.. ..Peoples Яepublik of Kalifoяnia

Re: A digitally-controlled-power power supply

Post by Soggy Bottom Boy »

Four elements in a keg?! :crazy:
"Well, ......I don't want Fop, goddamn it! I'm a Dapper Dan man!" ...Ulysses Everett McGill

"Good thing you found HD. It's like the mythbusters of distilling." ...Prairiepiss
Maritimer
Rumrunner
Posts: 716
Joined: Sun Jul 01, 2012 6:09 am
Location: Nova Scotia, Canada

Re: A digitally-controlled-power power supply

Post by Maritimer »

No keg, a 32 litre stock pot.
inside of boiler during element installation.jpg
I'm a devout disbeliever in the principle of KISS!

M
Soggy Bottom Boy
Rumrunner
Posts: 746
Joined: Sun Sep 09, 2012 4:43 pm
Location: Somewhere to the north of the.. ..Peoples Яepublik of Kalifoяnia

Re: A digitally-controlled-power power supply

Post by Soggy Bottom Boy »

Maritimer wrote:........I'm a devout disbeliever in the principle of KISS!

M
keeps things interesting that way! =O)

Got a spare oscilloscope that I can have? lol
"Well, ......I don't want Fop, goddamn it! I'm a Dapper Dan man!" ...Ulysses Everett McGill

"Good thing you found HD. It's like the mythbusters of distilling." ...Prairiepiss
Maritimer
Rumrunner
Posts: 716
Joined: Sun Jul 01, 2012 6:09 am
Location: Nova Scotia, Canada

Re: A digitally-controlled-power power supply

Post by Maritimer »

My scope is a 1985 Tektronix 2430. It cost $15,000 in 1985 and you can get one used for $750 these days. These Tek scopes are immortal. Back in 1993 it got inundated by a burst ceiling water pipe over the Christmas holiday. There was six inches of water on the floor when it was discovered, and the scope was right in the spray of water. I took the case off and left it for a few days to dry out. Didn't make a bit of difference to its performance.

Sorry, it's the only scope I have. :esad:

M
maheel
Trainee
Posts: 936
Joined: Mon Jan 26, 2009 11:25 pm
Location: OZtralia in the lower half

Re: A digitally-controlled-power power supply

Post by maheel »

keep an eye one those rivets holding on the handles...
they might leak in a very short time
well one of my pots did....

it also looks like your fitting those elements wrong ?
are they ex water tea kettle elements ?
they normally are fitted with a special o-ring and you drill on larger hole with a hole saw
fit the split oring thing to the rim of the hole and then fit the element through it
then a backing plate compresses the element

the way you doing it looks like water could easily get under the seal and leak onto the wiring points

stay safe
Prairiepiss
retired
Posts: 16571
Joined: Sat Dec 18, 2010 7:42 am
Location: Somewhere in the Ozarks

Re: A digitally-controlled-power power supply

Post by Prairiepiss »

Maritimer wrote: I'm a devout disbeliever in the principle of KISS!

M
Really? Would have never thought that. :wtf:
It'snotsocoldnow.

Advice For newbies by a newbie.
CM Still Mods
My Stuffs
Fu Man

Mr. Piss
That's Princess Piss to the haters.
Maritimer
Rumrunner
Posts: 716
Joined: Sun Jul 01, 2012 6:09 am
Location: Nova Scotia, Canada

Re: A digitally-controlled-power power supply

Post by Maritimer »

The elements are holding up OK after at least 15 hours of boiling and two weeks of holding low wines while undergoing bicarb treatment (which seems to be useless). There is silicone underneath them and through all the holes. I used the screw hold-downs to press them tightly to the bottom of the pot. I'm keeping an eye out for leaks. In fact, one of my plans is to have a leak detector in the pot base.

M
ipee7ABV
Rumrunner
Posts: 623
Joined: Fri Jan 06, 2012 9:01 pm
Location: chain o lakes

Re: A digitally-controlled-power power supply

Post by ipee7ABV »

Maritimer wrote:The elements are holding up OK after at least 15 hours of boiling and two weeks of holding low wines while undergoing bicarb treatment (which seems to be useless). There is silicone underneath them and through all the holes. I used the screw hold-downs to press them tightly to the bottom of the pot. I'm keeping an eye out for leaks. In fact, one of my plans is to have a leak detector in the pot base.

M
silicone?
maheel
Trainee
Posts: 936
Joined: Mon Jan 26, 2009 11:25 pm
Location: OZtralia in the lower half

Re: A digitally-controlled-power power supply

Post by maheel »

Maritimer wrote:The elements are holding up OK after at least 15 hours of boiling I'm keeping an eye out for leaks. In fact, one of my plans is to have a leak detector in the pot base.

M
cool just something i noticed in your pics and thought i would mention

i have used those elements my self in my beer brewing adventures :)

keep us posted as its and interesting "different" design
User avatar
sambedded
Trainee
Posts: 758
Joined: Tue Jul 10, 2012 11:25 pm
Location: Canada

Re: A digitally-controlled-power power supply

Post by sambedded »

Maritimer wrote:I found that my algorithm for controlling the power using a random function unfortunately did not work very well.
I'd recommend you to use Bresinham algorithm. It will evenly spread your "on" semi-periods across control interval.
> There are 1024 levels of power desired.
I believe such precision is not necessary 100 levels more than enough.

Good luck with your controller!
Maritimer
Rumrunner
Posts: 716
Joined: Sun Jul 01, 2012 6:09 am
Location: Nova Scotia, Canada

Answer to maheel

Post by Maritimer »

maheel,

Thanks for the tip about properly mounting the elements. I checked the resistance to the stock pot and found that only one was fully insulated. The other three read 60 Megohms, 48 Megohms, and 52 Megohms.

Last night I went to bed and started wondering about your instructions. Although there isn't any leakage coming through, there is obviously liquid getting to the elements' wiring. What would happen, I wondered. Electrolysis--generation of O2 and H2. Where would the gasses go? Out the leak that let the water in. Making it bigger, letting in more water, generating more gases, ugh.

I kept the split o-rings, but I wasn't able to use them because the stock pot has an aluminum heat spreader that is about 1/4 to maybe 3/8 inch thick. The or-ring is made to fit 1/8 inch material.

The proper way to connect the power would seem to be to have the element connections floating in the large hole, so that even if there were some leakage, the water would not contact the wiring. And if it did, the gases would be outside the stock pot.

What to do?

M
mash rookie
Angel's Share
Angel's Share
Posts: 2228
Joined: Tue Dec 07, 2010 8:20 am
Location: Pacific Northwest

Re: A digitally-controlled-power power supply

Post by mash rookie »

Maritimer I have to say you do jump in with both feet. I am not understanding how this might work to control a pot still or what advantage that fine of control would be.

Knowing exactly how much power you are applying can offer good information.
With boil temperature being determined by the wash ratio of alcohol to water, power applied determines vapor speed to your column or head.

Is your effort to preset coolant speed then control reflux ratio by power control? Something we did not get to with your first thread is that reflux condenser surface area and coolant flow rate make it all but impossible to make a calculation. Add in changing temperature during a run and wow!

Measuring temperature at your Liebig product condenser is of no value that I can see other than conservation of coolant water.

I monitor boiler temperature to understand my wash abv and to know when nearing tails. For my best repeatable high ABV, I monitor amperage, reflux coolant flow and exit temperature than product ABV.

Sounds like your having fun!
Maritimer
Rumrunner
Posts: 716
Joined: Sun Jul 01, 2012 6:09 am
Location: Nova Scotia, Canada

reply to sambedded

Post by Maritimer »

I checked out the Bresenham algorithm. It would do the job nicely.

I was also thinking of going back to the rand() function, using the same thresholding that I used originally, but keeping track of how many on half-cycles there are and comparing that to the desired number during the 256 half-cycle period. If the desired number has been reached, do not allow any more on half-cycles. If the number of remaining half-cycles is equal to the desired number minus the number passed, leave the remaining half-cycles on.

Yes, 1024 levels is crazy. But it's easy to fit into 8-bit chars.

M
Maritimer
Rumrunner
Posts: 716
Joined: Sun Jul 01, 2012 6:09 am
Location: Nova Scotia, Canada

Reply to mash rookie

Post by Maritimer »

Hi MR,

Yes, this is all about having fun!

My goal at this point is to make a variable power supply that produces a constant power, unlike my previous attempts which did not. The ridiculous precision is only a product of the computer byte size and the resolution of the ADC on the microcontroller.

I'm treating the still as a mystical creature that I'm trying to understand. I have to feed her and listen to what she is telling me.

The reason for measuring the temperature difference across a pot still Liebig is to monitor how steady the power supply is. The wild variations of my random algorithm power supply ruined my reflux ratio measurement experiment. Now that I have a steady power supply, I can try that experiment again.

Notice in the oscilloscope traces that 2kW is double 1kW. Measuring the temperture across the Liebig measures the relative power that the Liebig is condensing (at a constant flow rate).

In the Experiment 1 post, I wondered about using RR measurement to control the coolant flow rate of a CM still. That's the first practical use for measuring RR so far. I guess it's a matter of a solution looking for a problem. But first to verify that the solution is real!

M
mash rookie
Angel's Share
Angel's Share
Posts: 2228
Joined: Tue Dec 07, 2010 8:20 am
Location: Pacific Northwest

Re: A digitally-controlled-power power supply

Post by mash rookie »

In your first test you were correct on what practical application would be of managing RR. I run CM stills. Conceivably a program could be created to manage a run of a particular ferment batch. We know the boiling points of different fractions. You would have to have the ability to manage power input and reflux condenser coolant flow. The reflux ratio could be manipulated to create desired end product.
The ability to ring out more hearts on a neutral run requires different settings during heads compression, hearts collection and tail avoidance. Whiskey uses a different protocol of less RR after heads compression.
Just to give you something to think about. :)
Maritimer
Rumrunner
Posts: 716
Joined: Sun Jul 01, 2012 6:09 am
Location: Nova Scotia, Canada

Re: A digitally-controlled-power power supply

Post by Maritimer »

MR, you know how to pull my strings! Every morning I walk three miles along a path through the woods. It is a marvellous environment for thinking. This morning I was thinking about your post.

This precisely controlled power supply might be another one of those solutions looking for a problem. The mass of vapour produced is directly proportional to the power applied (assuming perfect insulation). The condenser temperature difference is proportional to the mass of vapour condensed and inversely proportional to the coolant rate. This is the condenser power.

I wouldn't imagine that the product condenser is in series with the reflux condense as in a VM still--that would require a perfect balance of vapour production and vapour condensation. Nevertheless, we can still measure the mass of the product from the power it takes to condense it. It's all numbers.

Controlling the reflux coolant flow rate might not be necessary, as long as it is constant. The same effect could be obtained by controlling the power. (Is this PM?) Measuring the reflux temperture difference at a (constant flow rate) measures the condensation power.

RR is defined in The Compleat Distiller as the ratio of the returned vapour to the total amount of vapour produced, so the RR can be calculated and set by the power applied to the still.

RR = (power of reflux condensation)/(power applied to still) [Ideally, but inefficiency factors can be accomodated.]

M
mash rookie
Angel's Share
Angel's Share
Posts: 2228
Joined: Tue Dec 07, 2010 8:20 am
Location: Pacific Northwest

Re: A digitally-controlled-power power supply

Post by mash rookie »

Maritimer wrote:MR, you know how to pull my strings! Every morning I walk three miles along a path through the woods. It is a marvellous environment for thinking. This morning I was thinking about your post.

This precisely controlled power supply might be another one of those solutions looking for a problem. The mass of vapour produced is directly proportional to the power applied (assuming perfect insulation). The condenser temperature difference is proportional to the mass of vapour condensed and inversely proportional to the coolant rate. This is the condenser power.

I wouldn't imagine that the product condenser is in series with the reflux condense as in a VM still--that would require a perfect balance of vapour production and vapour condensation. Nevertheless, we can still measure the mass of the product from the power it takes to condense it. It's all numbers.

Controlling the reflux coolant flow rate might not be necessary, as long as it is constant. Yes. KISS but any master distiller would argue that you need to be able to control both to meet certain goals.The same effect could be obtained by controlling the power. (Is this PM?) Measuring the reflux temperture difference at a (constant flow rate) measures the condensation power.
I have to get you away from thinking that vapour temperature will tell you anything. It only measures the make up of the vapour.RR is defined in The Compleat Distiller as the ratio of the returned vapour to the total amount of vapour produced, so the RR can be calculated and set by the power applied to the still.
Yes. Watts converted to vapour and measured against take off speed of condensed vapour. If you do those calculations and compare to collected data of coolant temp you can start to understand the capabilities of your still and particular packing or plate system. Different packing will behave differently. Some will produce well at a fairly low RR and allow for a fast take off. Some will require a high RR at lower vapour speeds to function well. But, this is just the beginning to understand fractional distilling. You have to consider your goals and work backwards some or the information is useless. If we first consider fractioning of partial pressure compounds, their BP’s and concentration you can consider what your goals are for your still.RR = (power of reflux condensation)/(power applied to still) [Ideally, but inefficiency factors can be accomodated.]

M
Just giving you a little more to chew on.
MR
King Of Hearts
Distiller
Posts: 1503
Joined: Wed Jul 21, 2010 2:09 pm

Re: A digitally-controlled-power power supply

Post by King Of Hearts »

I just got a splitting headache. :?
Maritimer
Rumrunner
Posts: 716
Joined: Sun Jul 01, 2012 6:09 am
Location: Nova Scotia, Canada

Re: A digitally-controlled-power power supply

Post by Maritimer »

Hi MR,

"Controlling the reflux coolant flow rate might not be necessary, as long as it is constant. Yes. KISS but any master distiller would argue that you need to be able to control both to meet certain goals.The same effect could be obtained by controlling the power. (Is this PM?) Measuring the reflux temperture difference at a (constant flow rate) measures the condensation power.
I have to get you away from thinking that vapour temperature will tell you anything. It only measures the make up of the vapour"

How do you get that quote thing with the boxes to work?

I wasn't talking about vapour temperature, but the temperture difference across the reflux condenser. See my latest post for Experiment 2 http://homedistiller.org/forum/viewtopi ... 17&t=32860 . The temperature difference technique works just as proposed. RR is now measureable. A different technique for CM will be necessary. This is just one of the parameters, which, once measured, can be controlled.

"...but any master distiller would argue that you need to be able to control both to meet certain goals."

The goals need to be stated and defined. Then we can go about controlling them.

This is getting funner and funner!

M
mash rookie
Angel's Share
Angel's Share
Posts: 2228
Joined: Tue Dec 07, 2010 8:20 am
Location: Pacific Northwest

Re: A digitally-controlled-power power supply

Post by mash rookie »

I missed that post. Lets see if I understand what you are saying. You are measuring reflux coolant differential correct? Are you calculating reflux ratio by comparing temperature at full reflux / head temp with no flow? Then expressing this as as a percentage?

High temp at 0%. lower temp at full reflux as 100% ?
Maritimer
Rumrunner
Posts: 716
Joined: Sun Jul 01, 2012 6:09 am
Location: Nova Scotia, Canada

Re: A digitally-controlled-power power supply

Post by Maritimer »

There are some edits, shown in blue.

Hi MR,

Watching the RR as the run progressed in Experiment 2 was really cool. I could change the power or coolant flow rate and watch the RR go into a slow transient and return to the orginal value. I ran the boiler at 1734 W and the RR was pegged at 0.83 until the tails started to appear, where it shot up into the 0.9s. I didn't wait for 1.0, though, because the product was stinky. There were three liters of azeoptrope hearts.

Another morning walk and another idea.

The power put into the boiler is transferred to the condensers, but some also goes into losses as radiated heat.

Instead of watts, use cal/sec to make things easier. 1W= 0.2388 cal/sec.

Some terms:
Pb boiler power in cal/sec
Tc temperature drop across the reflux condenser in degrees C
Fc flow rate of coolant in grams/sec
Pc power of condenser in cal/sec
Pl power losses in cal/sec
Ta ambient temperature degrees C

At 100% condensation,
Pc = Fc x Tc x ( specific heat of water=1cal/gram/*C) [g/s x *C x cal/g/*C= cal/sec for water]

Pb = Pc + Pl [the boiler power becomes the sum of the condenser power plus the losses]

Pb = Fc x Tc + Pl [substitute Fc x Tc for Pc]

Now, run the still at two different power levels with full condensation. There will be some cooling of the condensed vapour, but we can ignore it. We will have two simultaneous equations with unknowns Fc and Pl:

Pb1 = Fc x Tc1 + Pl
Pb2 = Fc x Tc2 + Pl

So, we don't even have to know the value of the flow rate, but we can solve for it if we want to.

For example, suppose Pb1 is 200 cal/sec [838 watts] and the flow rate is 40 grams/sec. With no losses, the temperature drop across the condenser would be:

Tc = Pb/Fc = 200/40 = 5*C.

But suppose in real life we measured 4*C.

That would mean we have 200 = 40 x 4 + Pl = 160 + Pl

So Pl = 40 cal/sec. This will be the same when the boiler is run at higher power because the temperature difference to ambient doesn't change. However, as the charge in the boiler loses its alcohol, the boiling temperature will increase. The next post will deal with this. Now we run the still at 300 cal/ sec and find that the temperture drop is 6.5*C.

So the simultaneous equations are:

200 = Fc x 4 + Pl
300 = Fc x 6.5 + Pl

Solve simulataneously to get Fc = 40 grams/sec and Pl = 40 cal/sec [unfortunate choice that the numerical values are the same]

OK, now to apply this to the RR equation:

RR = Pc/(Pb - Pl)

Suppose we want the RR to be 0.85:

0.85 = (40 X Tc)/(Pb - 40)

Ramp up the power from 300 cal/sec (full reflux). Measure Tc and use the known Pb in the equation until the RR goes to 0.85.

If this is all controlled by a microcontroller, the program can set the different power levels, measure the temperatures, solve the equations, and then use a feedback loop to set the power to the RR that you want.

M
Maritimer
Rumrunner
Posts: 716
Joined: Sun Jul 01, 2012 6:09 am
Location: Nova Scotia, Canada

Re: A digitally-controlled-power power supply

Post by Maritimer »

What happens when the condenser is not condensing all of the vapour around it? The vapour is going past at a certain velocity, so it either gets condensed or it goes right through the condenser. If you have a lot of vapour going past, does more condense than when you have a smaller flow?

There is a constant flow of coolant in the condenser. The vapour is at a fixed temperature, determined by its composition.

If vapour is getting past, then there is not 100% condensation, so the condenser power must be less than the boiler power.

Some terms:
Pb boiler power in cal/sec
Tc temperature drop across the reflux condenser in degrees C
Fc flow rate of coolant in grams/sec
Pc power of condenser in cal/sec
Pl power losses in cal/sec
Ta ambient temperature degrees C

The condenser power is:
Pc = Fc x Tc x ( specific heat of water=1cal/gram/*C) [g/s x *C x cal/g/*C= cal/sec for water] or
Pc = Fc x Tc for water.

Let's consider extremes.
> If there is no vapour, Tc = 0. The coolant temperature doesn't change at all.
> If there is 100% condensation, the condenser power will equal the boiler power, so
Pc = Pb = Fc x Tc
And
Tc = Pb/Fc
> So, in-between, Tc changes from 0 to Pb/Fc.
> But when the boiler puts out so much vapour that some is not condensed, does Tc continue to rise, or does it stay pegged at the 100% condensation temperature?

Does anyone know from experience or reasoning? My feeling is that since the vapour temperature is constant, the temperature of the coolant will stay at the 100% temperature.

M
mash rookie
Angel's Share
Angel's Share
Posts: 2228
Joined: Tue Dec 07, 2010 8:20 am
Location: Pacific Northwest

Re: A digitally-controlled-power power supply

Post by mash rookie »

http://homedistiller.org/forum/viewtopi ... n#p6995382

Check out FD's thread. He is a brilliant organic chemist. I learned a lot from him. When I get a little time I will engage you on how we could use your control theories for product manipulation.
Maritimer
Rumrunner
Posts: 716
Joined: Sun Jul 01, 2012 6:09 am
Location: Nova Scotia, Canada

Re: A digitally-controlled-power power supply

Post by Maritimer »

Thank you, MR. I'm re-studying chemistry (my past 35 years experience has been totally mechanical, electronic, and control). But from as physics viewpoint, how does a condenser react to various power regimes:

0-full condensation
full condensation and beyond

?

M
mash rookie
Angel's Share
Angel's Share
Posts: 2228
Joined: Tue Dec 07, 2010 8:20 am
Location: Pacific Northwest

Re: A digitally-controlled-power power supply

Post by mash rookie »

Maritimer, not being a physicist, that is either a particularly silly question or of over my head to answer. Power to condenser reactions vary depending on surface area and or conductivity of surface material, time, thickness, excreta. You conclusions are not as simple as you might think. There are many factors to consider. When we stop and consider our end goals, chemistry must be our first consideration then practical application applied.

nice pinch though. :thumbup:
Maritimer
Rumrunner
Posts: 716
Joined: Sun Jul 01, 2012 6:09 am
Location: Nova Scotia, Canada

Re: A digitally-controlled-power power supply

Post by Maritimer »

MR,

It might be a silly question. There is no such thing as 0-100% condensation. There is full condensation until the condenser can't handle any more. Why can't it handle any more, though? Because the coolant is at the same temperature as the vapour? Because so much liquid is collecting that it insulates the condenser from the vapour? Or the vapour is travelling so fast that the condenser hasn't got time to touch it?

What I'm wondering, because I don't have a CM still, is how do you go about setting things up so that you get the desired RR and boiler power? Now that I've got the VM RR measurement verified, I'd like to try out some theory and maybe even a collaboration with someone who'd be curious to have a Reflux Ratio Meter for a CM still. Seeing that number gives you a whole new view of your still!

I think this problem might be more physics than chemistry. The VM solution cancelled out all the chemistry--vapour composition, for example.

M
mash rookie
Angel's Share
Angel's Share
Posts: 2228
Joined: Tue Dec 07, 2010 8:20 am
Location: Pacific Northwest

Re: A digitally-controlled-power power supply

Post by mash rookie »

I think this problem might be more physics than chemistry. The VM solution cancelled out all the chemistry--vapour composition, for example.
I think it is equal parts maritimer. Your system worked with the simplicity of a VM. Just to create azeotrope.
I think it will could work with a CM once values are known for producing azeotrope at maximum speed and target fractions.
Many things affect that. Column packing, insulation and reflux condenser efficiency just to think of a few.

My still will make 96% or 75% whiskey with a few changes in managment.

My suggestion for your efforts would be to create a protocol for a CM, managing power and reflux coolant using physics to manage chemistry. Manage fractional distillation. How do you make the best tasting whiskey? Rum? Certain flavor congeners are in the higher BP fractions. What is the best protocol for heads compression? I cut this from the post I linked. That post also contained a flavor wheel that may or may not note BP's.

Acetone 56.5C
Methanol 64C
Ethyl acetate 77.1C
Ethanol 78C
2-Propanol 82C
1-Propanol 97C
Water 100C
Isobutanol 107.9 C
Butanol 116C
2-methyl-1-butanol 128.7C
Isopentanol 131.1 C
Amyl alcohol 137.8 C
Furfural 161C (322F
Maritimer
Rumrunner
Posts: 716
Joined: Sun Jul 01, 2012 6:09 am
Location: Nova Scotia, Canada

Re: A digitally-controlled-power power supply

Post by Maritimer »

Hi MR,

I remembered back to that post I found wherein someone was lamenting how boring it would be if everything worked first time. Woe is me! VM works so easily and perfectly that there isn't really any reason to think of automation.

The reason I've gotten into distillation is that, being a retiree, I just can't afford to buy alcohol any more. In Nova Scotia, it costs over $1 per fluid ounce. At that rate my wife and I are consuming $10 per day, or $3650 per year. Our preferences run to rum and gin--She loves gin and I love rum, and so we have a lot of fun!

But I love examining devices, and the still is my latest facination. My training is in physics and math, and I'm always examining things around me in terms of energy transfer, forces, power, etc. The still is completely new ground, but I want to take my experience in machines, and devices for experiments (which is what I designed and built), and do my own experiments, build my own devices.

In an earlier post I said that I liked to work with professorial types. While working, the professors I built devices for were so foreign to my skills that I was like a magician. One professor from Brazil said, "Machines are afraid about Maritimer!" So when I see the people on this forum are so expert in what they know, but also less knowledgable in what I know, I want to step in and fill in the blanks.

In working with these expert professors, I learned to listen to them, and question them so that I could fit their knowledge into mine. Ignorance and knowledge unite with knowledge and ignorance to produce new, but unknown, knowledge and hypotheses.

My grade 11 French teacher was a chemist at Labatt breweries in Montreal. He left because he didn't even drink alchohol; he wanted to do something more suited to his personal disposition. But he told us about fusel oils and hangovers, so I've always been very uninterested in whiskey and dark drinks. In Montreal, I always bought Alcool (pure neutral) at the SAQ. When we moved to Nova Scotia, I found that white rum and gin were accepable substitutes.

Well, that's what a good dose of azeoptrope will do!

M
lampshade
Rumrunner
Posts: 541
Joined: Tue Jan 24, 2012 3:19 pm
Location: Southern California

Re: A digitally-controlled-power power supply

Post by lampshade »

sambedded wrote:
Maritimer wrote:I found that my algorithm for controlling the power using a random function unfortunately did not work very well.
I'd recommend you to use Bresinham algorithm. It will evenly spread your "on" semi-periods across control interval.
Could someone explain how the Bresenham algorithm applies? That algorithm was developed to optimally identify which pixels on a computer screen are illuminated to render a straight line. By optimally, I mean uses the arithmetic operations that result in the quickest computer processing time. I don't think Maritimer's computer processor is over-taxed -- finding the solution for the number of ON cycles vs. the number of OFF cycles only has to be done once for each change in power setting.

~
~
In Flanders fields the poppies blow * Between the crosses, row on row, * We are the Dead. Short days ago * We lived, felt dawn, saw sunset glow, * Loved and were loved, and now we lie * In Flanders fields. -- from a WWI poem
Post Reply