• Advertisement

PID Control of RLM via Power Management

These usually have the means for returning the spirits to the column and usually allow metering of the take off to help in the reflux.

PID Control of RLM via Power Management

Postby zedzedtop » Tue May 01, 2012 2:41 pm

So I had a thought. Say you have a simple RLM boka setup, with a platinum RTD in the vapour under the top slant plate. It feeds the input to a PID controller, which is set to cool to the point you want that vapour to be. The output is connected to the heating element (sort of... wait for it...) Now here's my idea. At first glance, this setup won't work, because when you turn things on, nothing will happen, as the temp is room temp, and the controller is trying to cool. Also, approaching the set point from below, with heating control, would be very unstable, as PID logic isn't designed to react to these types of systems. So here's the mod: The heater element is set to be on constantly, but the cooling output of the PID is rigged up to INTERRUPT the power to the element. With the right parameters/damping/lag set into the PID, shouldn't this work?

Seems like it would work with RLM and VM.

Am I off my rocker?
'Bubble cap plates are better for batch distillation than perforated plates' - Benjamin Franklin

Fortune cookie: Your whole family are well
User avatar
zedzedtop
 
Posts: 2046
Joined: Fri Sep 18, 2009 11:20 pm
Location: NorCal
Equipment type: 25G pot, gas or 4500w
pot head or 4'' 3plate bubbler, cm

Re: PID Control of RLM via Power Management

Postby zedzedtop » Tue May 01, 2012 3:56 pm

I suppose I'm assuming that reduced power will increase purity, and I do realise that might not be the case. With power really petering out at the end of a run, the RLM product output might not see any liquid. Perhap regular LM would be better suited?

How would I test this anyway? I don't think there are 'normally closed' SSRs.
'Bubble cap plates are better for batch distillation than perforated plates' - Benjamin Franklin

Fortune cookie: Your whole family are well
User avatar
zedzedtop
 
Posts: 2046
Joined: Fri Sep 18, 2009 11:20 pm
Location: NorCal
Equipment type: 25G pot, gas or 4500w
pot head or 4'' 3plate bubbler, cm

Re: PID Control of RLM via Power Management

Postby jake_leg » Tue May 01, 2012 5:02 pm

Too much power and the column will flood, too little power and the vapour pressure will drop to the point where the packing dumps liquid back in the boiler. You need smooth control of the power to avoid these pitfalls.

It is easier to keep a stable temperature at the top of the column by control the liquid take-off using a solenoid valve than by controlling the power to the bottom of the column. The usual way to do that (LM-E-ARC) is to use a simple thermostat governed by a temperature probe towards the top of the column.
I remember Aaron Schnell
User avatar
jake_leg
 
Posts: 2815
Joined: Mon Jan 31, 2011 7:46 pm
Location: on the wagon

Re: PID Control of RLM via Power Management

Postby zedzedtop » Tue May 01, 2012 5:36 pm

True, but suppose the max power is the power you'd normally use during a typical LM run, so you wouldn't flood. I was hoping that the PID would take care of things. The ones I have (watlow sd31 series) can work with systems that respond very slowly, where the power source is separated from the sensor with materials of very high specific heat and low thermal conductivity. I figured that if the power was getting too low, the temp would drop at the top of the column. Wouldn't the PID learn via autotune the response characteristics of a still?

Isn't the issue with LM that it is difficult to maintain azeotrope throughout the run, especially at the end? I was thinking that this could automate the end of the run. Theoretically, you can reach azeotrope with no reflux at all, and low % boiler charge, but just with extremely low power input.

Wouldn't the LM-E-ARC as described in rikus book cause more of a surging because of the on-off nature of the solenoid?
'Bubble cap plates are better for batch distillation than perforated plates' - Benjamin Franklin

Fortune cookie: Your whole family are well
User avatar
zedzedtop
 
Posts: 2046
Joined: Fri Sep 18, 2009 11:20 pm
Location: NorCal
Equipment type: 25G pot, gas or 4500w
pot head or 4'' 3plate bubbler, cm

Re: PID Control of RLM via Power Management

Postby squidd » Tue May 01, 2012 11:04 pm

zedzedtop wrote:....How would I test this anyway? I don't think there are 'normally closed' SSRs.


ZZ,

If you input a small DC voltage to the SSR, it would then become NC.

With the Rd parameter set at 1 (cool) the J2 relay on the PID could open this input voltage via another small relay when the set point is reached.

Maybe.

squidd

Edit:

In fact, J2 is a pair of relays. One NC and one NO. So the SSR input could be fed through the NC relay which would open when the set point is reached.

squidd
squidd
 
Posts: 238
Joined: Mon Dec 22, 2008 10:05 pm

Re: PID Control of RLM via Power Management

Postby stillton » Sat May 05, 2012 8:54 am

zedzedtop wrote:Isn't the issue with LM that it is difficult to maintain azeotrope throughout the run, especially at the end? I was thinking that this could automate the end of the run. Theoretically, you can reach azeotrope with no reflux at all, and low % boiler charge, but just with extremely low power input.

Wouldn't the LM-E-ARC as described in rikus book cause more of a surging because of the on-off nature of the solenoid?

what riku does is to change the output. what you would do is to change the vapor speed. that is very sensitive. if you use a small intenal heater it will stop boiling in a matter of seconds. when that happens the tower of vapor will collapse very fast and the equilibrium is lost. on/off regulating the power is therefor not a recommended option. it might work with bain marie or possibly a plated still, i dont know.

the other option would be a SSR controlling the power. but i doubt your pid can control the temperature as stable as 0,1-0,2C that is required. The PID would have to work with a 0.01C sensitivity to even have a few fractions to calculate the PID-parameters. and even if its sensitive enough the reading of the temperature inside the column is very hard to get stable within 0,2C so i doubt the PID can make any valid calculations based on that. but i know much to little about PIDs to know for sure. maybe it can be solved with some other kind of sensor. and/or better software (check out the arduino threads).

i dont think LM or VM is a good option for this. you would not be able to set them in full reflux without lowering the power way down until the vapor doesn't even reach the top of the column.
User avatar
stillton
 
Posts: 345
Joined: Thu Jun 25, 2009 5:39 am
Location: Sweden
Equipment type: VM-E-ARC with Thors hammer


Return to Packed Columns

Who is online

Users browsing this forum: No registered users and 2 guests