PID and SCR Control Explanation as to delivering constant Wattage

Thread Starter

Andrew Leigh

Joined Sep 8, 2008
107
Hi,

Background
In the home distilling fraternity there is great division as to the suitability of PID vs SCR controls for alcohol stills, there is seldom middle ground and members get banned from forums due to heated debate.

I have both a PID (full P,I and D) and a SCR controller, in my limited experience the SCR does a better job. When distilling the liquid (wash), the wash has various volatile components with various boiling points which creates an azeotrope. The azeotrope will have a set boiling point at a specific point, but as the volatiles in the azeotrope are boiled off the boiling point alters and increases.

So the SCR lads argue that constant power is required to circumvent the ever changing boiling point which will require that the PID set point is altered on a continual basis in my opinion. The boiling point keeps rising as the volatiles are boiled off and the azeotrope becomes more water in composition. The other argument is simply that even if you set the PID to say 70C that everything with a BP below that will boil and everything with a BP above that will simply not boil. The other argument is that say 70C requires 1000W to maintain temperature, then the output will be x ml. However, if one could up the power delivery to 1500W, the BP of each component stays the same but the output ml lifts due to the increased energy.

Personal Experience
So my experience with a PID controller was exactly as above. I would constantly need to increase the set point as the azeotropes boiling point lifted. The other irritation was that the PID "pulsed" the heating element. So one could clearly hear the element get noisy as power was delivered and the stream of distillate would begin to flow, the minute power in the cycle was cut the distillate would cease to flow. The negative effect was simply that production output was severely hampered.

While with the SCR Controller I simply dial in the power required to get the distillate flowing as the rate required and I leave the controller alone for the most part.

It is not inconceivable that the P, I and D settings in my controller are awry. I did try it on factory settings, when that was not satisfactory then I tweaked it from other distilling sources.

Questions
1. Is it possible to setup a PID so that there an uninterrupted constant power delivered to the electric element?
2. If so what mode must be set, PI with no D?
3. Can one reduce the "pulsing" to effectively a continuous stream of power?

Any other insights would be most welcome.

Thanks
 
Last edited:

Alec_t

Joined Sep 17, 2013
12,201
as the volatiles in the azeotrope are boiled off the boiling point alters and increases.
I have no practical experience of PID control, but if your PID controller has a programmable ramp function I expect you could set the parameters so that the controlled temperature follows the BP increase. That implies quasi-continuous power.
 

Hymie

Joined Mar 30, 2018
991
PIDs are available with multiple set points with a ramp option between points.
So once you work out the required temperature profile of the wash, you could program in the desired temperatures, dwell times and ramp rates.
You might find that the required temperature profile changes depending on the precise mix and volume of the wash.
 

Papabravo

Joined Feb 24, 2006
16,792
An analogous situation occurs in DC-DC converters. Trying to close the loop by looking at the output voltage has some manifest difficulties. Adding a loop for peak or average current solves a number of those problems. So what happens if you do both? In your case there would be a loop that looked at following a temperature profile and a loop that looked at flow rate.
 

ElectricSpidey

Joined Dec 2, 2017
1,801
Is your PID driving a relay or a solid state device? (not the load)

Because it sounds like your switching rate is too low.

How does your SCR controller know when its time to raise the temp?
 
Last edited:

strantor

Joined Oct 3, 2010
5,633
In the home distilling fraternity there is great division as to the suitability of PID vs SCR controls for alcohol stills, there is seldom middle ground and members get banned from forums due to heated debate.

I have both a PID (full P,I and D) and a SCR controller, in my limited experience the SCR does a better job.
Your use of the terms "PID controller" and "SCR controller" are a bit off. Maybe those terms mean something specific in the world of distillation but in my world (industrial controls) they have little to do with each other.

PID describes a certain kind of feedback loop algorithm.

SCR describes a certain kind of power control device, and most likely implies phase angle control (constant power). I suspect that your specific usage of the term SCR control implies a static, user-set output level with no feedback loop. That does not have to be the case.

alternatives to PID are: on/off control (like an oven), proportional control (like an 80s model cruise control), PI control, PD control, and a host of other control methods.

Alternatives to SCR (phase angle) control are: SSR/SCR (bang-bang) control, contactor bang-bang, Variac voltage control, et al.

You can use any of the feedback loop schemes with any of the power control devices

You can use PI control with a motorized variac. You can use on/off control with a SSR. You can use PID with a phase angle controlled SCR.

Most often PID controllers are designed for (or set up for by default) bang-bang control of a SSR. But it does not have to be so.

Questions
1. Is it possible to setup a PID so that there an uninterrupted constant power delivered to the electric element?
2. If so what mode must be set, PI with no D?
3. Can one reduce the "pulsing" to effectively a continuous stream of power?
1. YES. Get a PID controller with an analog output, and an SCR controller with an analog input.
2. PID. And it needs to be tuned properly. The default PID values will NOT work. And they aren't something you can guess at either. Your controller needs to be tuned for your specific application. I recommend getting an auto tuning controller, fill your still with water, and run an auto tune on it. It will probably take at least an hour but once it's learned the nuances of your application, temperature control from then on will be rock steady.
3. YES. See #1
 

Thread Starter

Andrew Leigh

Joined Sep 8, 2008
107
Is your PID driving a relay or a solid state device? (not the load)

Because it sounds like your switching rate is too low.

How does your SCR controller know when its time to raise the temp?
The PID drives a SSR.

The SCR does not know, One would set it high and allow the liquids to boil off as they reach their boiling points. The output of the SCR will such that all volatiles can boil off at one setting. The difference is that at higher wattage the rate of boil off is simply higher due to the energy being supplied.
 

Thread Starter

Andrew Leigh

Joined Sep 8, 2008
107
Your use of the terms "PID controller" and "SCR controller" are a bit off. Maybe those terms mean something specific in the world of distillation but in my world (industrial controls) they have little to do with each other.

PID describes a certain kind of feedback loop algorithm.
Indeed this is how I understand it.

SCR describes a certain kind of power control device, and most likely implies phase angle control (constant power). I suspect that your specific usage of the term SCR control implies a static, user-set output level with no feedback loop. That does not have to be the case.
I understand this to be the case, basically a fancy variable power supply with user set input.
Was not aware that it could be any thing else.

[quote}alternatives to PID are: on/off control (like an oven), proportional control (like an 80s model cruise control), PI control, PD control, and a host of other control methods.

Alternatives to SCR (phase angle) control are: SSR/SCR (bang-bang) control, contactor bang-bang, Variac voltage control, et al.[/quote]
My only experience with the PID has been as an off control which does not suit the application.

You can use any of the feedback loop schemes with any of the power control devices

You can use PI control with a motorized variac. You can use on/off control with a SSR. You can use PID with a phase angle controlled SCR.

Most often PID controllers are designed for (or set up for by default) bang-bang control of a SSR. But it does not have to be so.
So this is the missing link, my ignorance of how these things work.

1. YES. Get a PID controller with an analog output, and an SCR controller with an analog input.
2. PID. And it needs to be tuned properly. The default PID values will NOT work. And they aren't something you can guess at either. Your controller needs to be tuned for your specific application. I recommend getting an auto tuning controller, fill your still with water, and run an auto tune on it. It will probably take at least an hour but once it's learned the nuances of your application, temperature control from then on will be rock steady.
3. YES. See #1
The problem is that I do not want to control temperature as the liquids boiling point is constantly changing. I want to be able to set the PID to drive a SSR such that together they become a variable power supply with no on/off operations (not talking about fast switching).
 

strantor

Joined Oct 3, 2010
5,633
The problem is that I do not want to control temperature as the liquids boiling point is constantly changing. I want to be able to set the PID to drive a SSR such that together they become a variable power supply with no on/off operations (not talking about fast switching).
What measurable variable would you like to use as feedback to the controller if not temperature?
 

Thread Starter

Andrew Leigh

Joined Sep 8, 2008
107
I should have mentioned that as a temperature controller the PID works a treat. It has a resolution of 1°C and the overshoot is very small as she approaches temperature. The problem has been the on/off, and probably the coarse resolution of 1°C does not help. So it is a case of now you get distillate now you don't. Then you bump her up a degree or two and the cycle repeats.

The PID unit (Rex C-100) has a self tuning function with an output of 3 - 35v d.c. is this the analogue output of which we speak?
The SCR devise is a self contained Thyristor controller which has a variable voltage output controlled through a potentiometer.
 

Thread Starter

Andrew Leigh

Joined Sep 8, 2008
107
What measurable variable would you like to use as feedback to the controller if not temperature?
Agreed hence the problem. I could set the temperature for say 94°C but would need the output to be continuous so would need to be able to set the PID accordingly.
 

Papabravo

Joined Feb 24, 2006
16,792
It is hard to see a path to a solution with a statement like: "set the PID to drive a SSR"
The PID takes analog inputs and has analog outputs and they operate continuously -- no on-off behavior.
The SSR (Solid State Relay) is a digital device with a digital input and a digital output. It has no continuous behavior, just on-off behavior
I understand the goal of having a continuously variable power supply. I'm going to suppose that the power we are talking about is AC power. If that is not the case then it changes things a bit.
To make a continuously variable AC power source you must be able to take power from the mains and change the voltage or the current or both. If you had a variac transformer with a controllable mechanical arm that could rotate the dial that might be a feasaible, but impractical solution. The SCR (Silicon Controlled Rectifier) is a device that modifies the AC waveform from the mains to basically change it from a sinewave to something else with a lower power. Using this method you chop up the sinewave so it only has non-zero voltage and current for some portion of a cycle. In this way you can vary the power delivered to the load from 0% to 100%

In my reading of the situation the details about what distinguishes the two methods is murky. I see no reason why you cannot combine PID feedback of the power level, flow rate, with SCR control of the AC power.
 

strantor

Joined Oct 3, 2010
5,633
Agreed hence the problem. I could set the temperature for say 94°C but would need the output to be continuous so would need to be able to set the PID accordingly.
You can absolutely do that. Like I said, you need a PID controller with an analog output and a phase angle fired SCR with an analog input. Look at the links I provided.

They make a Rex 100 with analog output but it's a different model than what you have.
Screenshot_20210227-130956_Drive.jpg
The control output digit would need to be an 8.

Your problem of the boiling point changing is probably a direct result of the bang-bang SSR control scheme. If you went to a PID phase angle fired SCR scheme I think you would be fine.
 

Thread Starter

Andrew Leigh

Joined Sep 8, 2008
107
It is complex and I do not fully understand it. Perhaps explaining the desired output would help explain the issue.

My el cheapo 4000W Thyristor controller works a treat. I turn the potentiometer to full, the liquid will begin to boil at some point, at this point I back off the power until I get a constant stream of distillate about the thickness of a pencil lead (industry standard ;)). If I want to have a higher spirit resolution (alcohol %) I then back off the power even more until I get a drop for drop output. My output is infinitely variable (up to the max output capacity). This is the desired result, so right now I have no issues.

However I started with a REX C-100 PID which drives a SSR. I am still unable to duplicate the output of the Thyristor controller. Others say that a PID (including the one I have) is not only suitable, but better than the SCR for the process. Nobody can tell me why, nobody will give me their parameters (some have identical units) for me to see if they work with my PID. So I have to assume that they are operating the PID using temperature as a control variable and setting the temperature continually starting at 78 and then ending at 94. I also have to assume that they are also experiencing the on/off nature as they are all driving SSR's.
 

ElectricSpidey

Joined Dec 2, 2017
1,801
Try setting the PID output to 10 CPS and see if that works for you. (or up to 60 CPS if the SSR and PID can handle it)

(as long as the SSR can handle it)

Be sure the control output is voltage pulse.

In the end the PID switching speed connected to a SSR may be too slow for your needs.
 

Thread Starter

Andrew Leigh

Joined Sep 8, 2008
107
You can absolutely do that. Like I said, you need a PID controller with an analog output and a phase angle fired SCR with an analog input. Look at the links I provided.

Your problem of the boiling point changing is probably a direct result of the bang-bang SSR control scheme. If you went to a PID phase angle fired SCR scheme I think you would be fine.
The problem with the boiling point is a function of the solution (liquid) which is a combination of many components each with a different boiling point. As these components are driven off the boiling point of the solution intrinsically alters.

When you set water on a stove in a pot at 1kw and it will boil at 94°C (well certainly where I am from), if raise the power to 2kW the water will still boil at 94°C but there will be more steam due to the increased power. The increased steam for us is increased volume output. So you can control the output of water vapour by simply applying more power.

If you put a pot of alcohol wash on the stove and set it to maintain the solutions boiling point, 78°C. The lower BP components will boil off as the their BP is reached which in turn alters the BP of the solution. As the lower BP components have boiled off the average BP now increases. However the stove is set to 78°C and will maintain that temperature. The result is that the process has now stalled as any substance with a BP of say 80°C will not be able to evaporate. So one now needs to lift the temperature to say 81°C, and the cycle repeats.

If one was to say set the PID to reach 94°C to avoid all these adjustments then the amount of energy required would increase the output past the pencil lead trickle, the condenser would not have sufficient capacity to cool the distillate and the alcohol % would reduce. All undesirable attributes.

It is for this reason that some argue that controlling the power into the still is better than controlling temperature. The PID lads insists that PID's are best but they are not technical enough to explain it to me who probably would not understand.
 
Top