Ports/Sensors for Automation

vooharmy

After many years of research and thought I am now finally pulling the trigger on my "Dream Still". There has been a lot of talk about automation especially from OD and Lloyd and this is something of great interest to me. I would like to be able to easily add this to my still in the future and wonder to myself if there has been any consensus on the placement of these sensors and ports. I am looking particularly at a 5" still and wondering to myself what changes, if any would be needed to take advantage of automation in the future. The big one that jumps out to me is the placement of a thermowell in the dephlegmator, possibly before and after the dephlegmator and at the parrot. I am also wondering if there is a benefit to placing thermowells in some or all of the tees themself? I know some people have done a lot of research on thermometer and PID placement and I am keen to hear any thoughts on this. Automation is something that I would look to include in the future but I want to be prepared for that inevitable upgrade!! If you have any advice please chime in.

punkin

I certainly don't have much knowledge of this. I would think though that a thermowell straight after the condensor will give you the same information as one inside it.

Also that a thermowell in the reducer above it will give the same (almost) reading as one in the top tee.

olddog

Plus one in the dephlegmater, about 2/3 from the bottom.

Lloyd

@olddog said: Plus one in the dephlegmater, about 2/3 from the bottom.

Yep, that's the only one that's really needed and works better than a sensor in the vapor path. The rest of the sensors don't do much except provide information.
I switched my PID sensor from the vapor path to the dephlem as OD suggested and it added much better stability - maintaining the vapor temp within a fraction of a degree.

vooharmy

Ok, wells that's great information to know! I was also wanted to know if there will be areas of heat difference in the 4" 4 port super dephlem, compared to the 5" 2 port super dephlem. Thinking to myself if the heat is well balance as it rises up, then temps will be all the more stable. I tend to over think things so I wouldn't be surprised if this was an irrelevant detail!

Lloyd

On my personal 4" super dephlem I wound up using one port for the sensor and another for a drain.
I had it configured at one time with duel Ins and Outs for the water but I didn't see any difference - but that's probably because its so tall. It seems to get enough internal turbulence with just one In as the water bounces around the pipes.

vooharmy

It was actually observing your super dephlem that put the idea in my head. The in and out are on opposite sides and the column is tall. I wonder if the temp will hold as steady if the in/out are on the same side? It seems like a trivial detail but it might make a difference... Or maybe not! It's all guess work for me.

grim

Much of this came out of my own experience, but quite a bit of this was hashed out during the 12" Crystal Dragon design sessions:

I would think these 5 would give you the most information and control:

1. Reflux Condenser Coolant Temp (through the wall, in the RC)
2. Product Condenser Coolant Temp (through the wall, in the PC)
3. Vapor Temp (above the RC)
4. Condensate Temp (Alarm/Safety - in or before the parrot)
5. Boiler Temp (Only really useful for monitoring heat-up speed)

1 and 2 would serve as inputs to PID controls for either solenoids or proportional valves

3 could be used as a control input but puts you well beyond simple PID control as there are multiple variables that impact vapor temp, otherwise fed to a temp display

4 could be input to a simple temperature controller with alarm, or more complex to cut off boiler heat should condensate temp rise above safe temperatures

5 could be a control input to a ramp control PID if you wish to vary the heat up time (perhaps you have immersion elements and are using a wash that contains small particulates or a higher than usual sugar content, maybe you want to ramp slow to reduce the watt density of the heaters). Otherwise right to a temp display.

There are good reasons to control 1 and 2 with a PID (Solenoid or Proportional) - If you are using municipal water, there is good opportunity to reduce water usage by maintaining precise temps. If you are using a recirculating coolant system, your control mechanism will need to increase the flow rate over the run to keep temps consistent, because your coolant reservoir will rise in temperature. In the latter case, it makes using a recirc system significantly easier.

grim

In terms of approach, I'd always recommend RTD sensors in direct contact with whatever is being measured over thermocouples. RTDs are much easier to use, much more forgiving. RTD used to be very expensive, but now they are nearly the same cost as TC.

Lloyd

Its more forgiving than you think.
While folks talk about optimized this and optimized that most know its not the pool cue but the one that holds it. Meaning its either in the hands of a master or in the hands of a learner - and most of us are somewhere between the two extremes.
I have long held that a newbie get a firm footing by running the simplest of pot stills to get a grasp of distilling before moving on to more advanced equipment like plated stills. They can be had for cheap and if you go modular then your investment will not have been wasted as the parts can be reused into more complex stills.

grim

@Lloyd 100%

There is always the risk here of a kind of Illusion of Control. A belief that more sensors, more accurate sensors, controllers, equipment, and the like will result in a better product. We all know that isn't the case.

I've always looked at the benefit of these items in improving consistency of product, but not necessarily quality of product. Which means if you don't know how to use the pool cue, all that this technology will provide you with is the opportunity to make a consistently BAD product!

The big fear I have is that those who don't understand the process well might assume that these kinds of systems allow for a rig to run unattended, which we all know is never the case.

vooharmy

This will be my third still, the first two I built myself so that I understood the process. I am fascinated by the science and artistry of distilling and god knows I like a good drop! This is my first plated still, the other two being a pot still and a VM/LM boka. For me automation is in the future, something I will use in the future to bring consistency to the recipes that I know and love. In the meantime I want to ensure that anything that could be easily completed now and would be costly in the future are done. I can't thank you enough for your detailed and insightful posts, I will definitely take everything that has been said on board.

CothermanDistilling

FWIW - I am happy with my BCS-462 that uses thermistors... I measure: boiler liquid, below first plate, above top plate before dephleg, above dephleg, cooling water in, cooling water out of dephleg.

I will someday monitor cooling water out of product condenser, but that is with a danfoss valve and simple SD thermometer now, I will try the 2/3 dephleg thing someday... I may do a custom 8" dephleg that has the tubes separated to provide room for a thermostatic valve sensor

Like Lloyd and Grim say, you don't need lots of sensors to make good product.... but they are fun and educational..

@Lloyd - maybe it is time to look at new simple thermometers... something a step up from the current simple ones, with the ability to calibrate them and also a bracket to mount right to the location or some other part of the still... I would suggest classic, elegant, robust as the theme... then of course I also want a really tricked out thermometer, imagine the pebble watch as a thermometer, ePaper display could show analog or digital, bluetooth connects to a controller...

grim

My vote is Triclamp RTD sensors - Keep it simple, keep it modular.

Folks looking for lower-cost or more simplistic designs can opt for the lower cost sensors:

Tri-Clamp Liquid tight RTD Sensor @ Auber Instruments

Folks looking at scaling up eventually, or going pro, can look at higher quality sensors:

Sanitary @ Omega

brewsmith

So I have now read through a few of the threads where defleg temperature monitoring has been mentioned. I will be pulling the trigger shortly on a 4" regular defleg, as I am running 5500W to a 15.5gal boiler and a 4" plated/packed column and it doesnt seem like the price for the super would be justified for my current serup. As such, does anyone have any input on the best way to rig a 2-port defleg for monitoring.

I am thinking about installing a tee on the output, threading in a thin rtd probe from auber designed to minimize flow restriction in pipes through the tee into the defleg, then having the third port of the tee angled down for the water exit. Does this seem like it would yeild an accurate reading compared to those installed in a dedicated port? It seems with all the water mass and turbulent flow that there can't be THAT much variation, but then again I have not yet used a SD defleg, so I dont have any real world experience.

grim

Dedicated ports for the condensers are only really important once you get to very large sizes, because in larger systems you need to deal with significantly more lag time, the volumes are higher, and the systems are much slower to respond. It becomes a real challenge when you are dealing with both long lag times, and inconsistent lag times, as your controller will hunt, overshoot, undershoot, trying to compensate. You will drive yourself crazy trying to tweak program settings to correct for this.

Also, the larger the difference between your coolant temperature and condenser set point, the more important it is to locate the probe as close to the coolant as possible. Especially if your controller is programmed to allow flow to reduce to zero.

grim

Easy way to manage this is to just put a bypass on your control valve, something like a small needle valve, and make sure you keep it cracked enough so that there is always some baseline level of flow (zero flow not an option). In this scenario, even if your sensor is down stream of the output, you can be reasonably sure that the difference in temp will be relatively stable. We don't care what the actual number on the display reads, only that it's consistent. Who cares if by locating it 2 inches downstream it reads a half degree lower. Once you know where it needs to be, that's the target number. Going to be different between setups anyway, no magic numbers.

Or, if you use a proportional valve, and you've got a nice PID, just change the output control setting from 0-20ma to 4-20ma, it's a neat little trick that can keep the valve cracked a bit. I suspect you can do with with PWM and solenoids, but it's going to depend on the specific controller.

CothermanDistilling

@brewsmith - here is what I did, this pic is actually the product condenser, I have the same fitting welded to a 4" regular dephleg:

I used some swagelok weld-on 304L 1/4" compression fittings and a 1/4" thermowell with a teflon ferrule, then slid a low-cost, easy to replace, bead type thermistor in it (BCS uses thermistor)... I have them in kegs, tc fittings, etc..

part number is Swagelok 304L-400-1-4W ( I got on ebay for cheap, 14 for $35) have to drill out to 1/4"

thermowell was brewers hardware

ferrule was grainger 1VCZ6

I see you are in the south, if that is south USA, I am near Tampa, FL, if you wanted welding like this done or wanted mine and I could go to a 4" super, let me know..

punkin

Well there ya go. The only thing i was right about was that i don't have the knowledge to be able to comment. :)) :(|)

captainshooch

I am looking for one more 4" standard deph. If anyone wants to upgrade to the super deph shoot me a pm. @CothermanDistilling modified his nicely which could be better used by someone looking to do some automation or I would offer to get that one from him so he can get the super.

CothermanDistilling

captain, you can have my regular dephleg if @brewsmith does not take it, and I will get a super... also another reason for you to come visit the distillery.... ;-) Oh, I just priced out copper sheet at Alro metals, it was $840 for a 3'x8' sheet of .0625", $12/lb...

captainshooch

Sounds like a plan :D
Sorry, did not mean to hijack the thread..

CothermanDistilling

back on the thermowell port subject, i am thinking about trying to spinweld some thermowells on with my mill/drill, set up a jig to hold the condenser or whatever, set up argon tube to fill the condenser, set up another argon tube to surround the outside, have a hole in the condenser and chuck the thermowell in the mill, start it up, press down till glowing red, shut off mill motor, maybe have good weld in a fraction of the time... obviously I will not start on a condenser, but a piece of scrap pipe..

brewsmith

I am certainly interested in that defleg @CothermanDistilling, that is a nice little mod. My only concern is that I am using RTD probes for all of my sensors and I don't know if I have a probe that will fit it. It looks like that's a 2 wire thermometer installed? I guess I may be able to pull the guts out of one of my current RTD's and install it in the thermowell...anyone tried this?

CothermanDistilling

The thermowell is a short (1.5" maybe) section of 1/4" OD SS tubing with one end rounded and welded shut. it is sealed by either an o-ring or a tapered teflon ferrule

In that pic, it has two thermometer wires in it, it has the SD digital thermometer probe and a bead type thermistor in it... (the SD probe is a bit behind the bead and is used for a snug fit, so a jewelers screwdriver gently pushed both in to get good contact between the bead and the thermometer wall. It seems to have great response in my opinion) if your RTD has a 1/4" OD, you just replace the thermowell with your probe, if it is smaller, you can just place it inside the thermowell...

punkin

This is @Lloyd's setup mate for the same thing.

vooharmy

@punkin @lloyd Thanks for the picture it really does highlight how adaptable these modular systems are! Can you tell me how deep the probe is going into the dephlem? I have a picture of my thoughts, Below with A) and C) being the same.

olddog

It does not have to be deep, as long as it can sense the temp of the coolant

OD

Lloyd

The probe goes into mine about 1".

vooharmy

I was thinking that was the case, the coolant tubes are rather close together but there is plenty of access if the probe only needs to be inside 1" or so. Thanks for the information.

Mickiboi

Setup I did for @UZGin. Control panel, thermo wells and 3 way valve. Dephleg in/out temps were done with Swagelok 6mm ID S/S to suit the thermocouple diameter, soldered to the coupling using Eutecrod 157. Vapour temp probe in the reducer, TIG welded the locknut and sealed the thermocouple thread with food grade Hi temp sealant, this exposes the thermocouple tip directly to the vapour to reduce thermal lag in the system. Boiler temp probe uses same Swagelok pipe but TIG welded one end closed and TIG welded the locknut to the plate, this is not for control, its just there to monitor the temperature, I used a PID controller for the boiler temp because I have boxes of them. Control is all 24VDC low voltage. 3 way valve is a 0-10VDC modulating DN15 with a CV of 5.0. I used 4 pin plugs for all connections so you cant blow anything up by plugging the wrong component into the wrong socket. There is plenty of room left in the control panel for a VFD drive for a pump and heatsinks for SSR control if and when he chooses to add more control. Works well. Panel is Ctick and COC certified for commercial applications. Valve control has a manual and automatic setting allowing you to dial in a valve position or set to operate from temperature controller.