Ports/Sensors for Automation

grim

For condenser coolant flow you do not need instantaneous open or close, in fact this is an area where PID+Proportional shines, even with slow valves.

Required coolant flow actually changes very slowly over the course of a run, the PID+Proportional are only making very small changes to valve position. For example in my system, when the reservoir is very cold, the valve is at a low open position, as the run progresses and cooling requirements increase, and the reservoir temperature increases, the valve is at a wider position. The move from the low to high flow positions has the whole run to take place.

My trials with the Johnson valves are all about finding a valve that has a very linear response with good control at the low end (necessary for very cold coolant), but with a wide enough maximum flow rate that it can deal with a reservoir whose temperature is significantly higher than at the start.

vooharmy

Could VFD off one pump work by using different line thickness, or splitting the lines to push more water to the product condensor for any given volume of water to the dephlem?

Mickiboi

@grim you may want to consider 2 valves in parallel, one small 1/2" CV 7 and the other one 1" CV 11. You can still control them from the one PID, just use programmable voltage switches on them. Small valve 0-4VDC from PID = 0-10VDC to the valve and the large valve 4-10VDC from the PID = 0-10VDC to the valve. I understand what you are trying to do but I think you will have a great deal of trouble looking for that perfect valve. There isn't one in the Johnson range that will do what you want at 1". Its the low end control on your big valve that you just wont match at low flows, unless of course you use the VFD option mentioned earlier. I can do some simulations of putting the VFD and the Valve in parallel to receive the same signal from the PID. When the valve is only partly open the pump is at slow speed and as the valve opens the pumps speeds up. I have never done anything like that before but I do have access to some pretty good automation simulation software. I can see is its feasible with your 1" CV11 valve. If you can tell me what pump you are using I can get a pump curve for it and see what happens.

You may have better luck with a Belimo CCV but be prepared to pay for it if you want to do it with a single valve. But I can't say if they will have one that will work either.

Lloyd

Yep, I did get a motorized valve as you described and the ~6 to10 second open time is worrisome but I haven't used it yet. Did play with it a bit with a cup of hot and cold water to dip the sensor into. Close time seems to be faster than open time.
Just playing with stuff I know nothing about so disregard me.

grim

VFD won't give you the range of flow you need, I suspect once you hit 10-20hz you will start burning up pumps. Torque will start to fall off rather quickly, and you are going to lose pump head, the impeller will spin and do nothing but heat up.

Also, most pumps are single phase, which means a simple VFD with 3 phase output wouldn't work, VFDs for single phase motors are much more expensive, and in general don't work as well. Expensive valves are going to look cheap after burning up a few motors.

grim

My Johnson valves are 60 seconds from 0-100 or 100-0, and I don't have a problem.

Like I said, over a typical run, 5 hours, my valve makes 1 partial swing. Sure, depending it might tick up or down a position or two along the way, but you are never in a position where the valve is swinging from open to closed, ever, other than making massive changes to the set points during a run, which you really wouldn't do outside of stacking heads, or killing reflux when moving into tails. In either of these situations, letting the valve swing for 5-10 seconds is no big deal.

grim

One of my big pumps is a two speed, I suspect I can use the secondary relay output of my PID to switch the pump between low and high speed depending on the flow range. Trying to control for this is going to be a headache though, because I'm sure it will create a very non-linear/kinked response. But really, not sure it is even necessary.

The experiments with the bigger 1" valve are intended to be the production setup for the 12" crystal dragon product and reflux condensers. Won't get any real world data until those are setup.

We've got a 600 gallon coolant reservoir with a 3 ton chiller, which I'm hoping is going to be big enough for the beast.

Lloyd

We've got a 600 gallon coolant reservoir with a 3 ton chiller, which I'm hoping is going to be big enough for the beast.

Uh yeah, that should cool the 12" column and half the city. Dang you think big. I meter out mains cooling water and you have a 600 gallon reservoir with a chiller.
Do I reek of jealously?

grim

Want a laugh?

We almost picked up a thousand gallon dimple jacket cooking vessel that came out of a Hostess factory. It was used to process the creamy white filling for Twinkies for something like 20 years. The factory closure came out of a labor dispute that culminated in the bankruptcy of Hostess foods which ultimately resulted in the Great Twinkie Shortage of 2012/2013.

It wasn't just a mash tank, it was a piece of history! Twinkie turned whiskey, could have been a great marketing angle.

In the end, they wanted a little bit too much for it, so we passed.

Lloyd

I'd kill small innocent children for a Twinkie right now.
Its been that long since I had a Twinkie!

You may have made a mistake.

But that's nothing to the massacre I'd commit to getting a decent pizza.
And getting authentic Mexican food would warrant genocide.

Somehow I'm feeling both hungry and mean :))

Sunshine

So, first of all: Lloyd, for some reason I am happy, that you are so far away from us now. I mean, my children are still small and mostly innocent.

:@)

Second, I don't know what a Twinkie is, but the next time, we ship a parcel to you, I will remember, that you like sweet things. I have something in mind for you, that is truely Austrian and very sweet, too.

@grim: A mash tank with history could really be something special for publicity. Perhaps you can negotiate a bit?

Smaug

Lauren had her first Zinger this past week end. Was better than I thought it would be.

vooharmy

I had to Google Twinkie, frankly I am shocked! People complain that Vegemite is gross!!

MickyMoonshine

Hi Lloyd Is there a chance of getting a RC with the sensor port already in for water temp sensor?

punkin

Could possibly make it a standard fitting with plug on Supercondensors?

Lloyd

@punkin said: Could possibly make it a standard fitting with plug on Supercondensors?

Best idea yet. 4" super and on everything larger than 4". It may take awhile to propagate throughout the entire lineup.

vooharmy

Should start with the 5"!! :))

Mickiboi

Spoke with my pump guy today about running motors at low speed continuously. Single phase out of the question, 3 phase no probs. I have a Danfoss VLT 2kW drive in my shed, he is going to give me a 750 watt 3 phase pump because I buy so much other shit of him. Time for some testing. Told him what I wanted to do and he said go for it no reason it wont work. My problem is I don't have a still yet so I have to make do some other way. But funny thing is he said, "You know I always wanted a still".

Putting 2 valves in parallel works on the simulator. 2 X 1/2" valves of the same CV using voltage switches set at 0-5 = 0-10 and 5-10 = 0-10. Looked good in the program. Works out that 2 X CV7 = 1" CV 10. Linearity is also pretty good.

Lloyd

Please speak English @Mickiboi, your last post made no sense at all.
x= M7/47 carry the one....
Dumb it down a little please for Punkin's sake!!

grim

@Mickiboi - My test rig for the big valves is a number of containers and heating elements. Goal was to try to simulate the internal volume of the big 12" CD deplegmator and the heating load. Approximately 5 gallon container with an immersion element, larger 30 gallon container to simulate the coolant reservoir (or using mains water directly).

Slowly turn up the power to the element to approach the set point, once the pid nears setpoint it begins control, I can easily raise or lower power to test the responsiveness of the valve and the control setup (which is actually a bit fun).

Problem is my small element isn't large enough to simulate the heat load to the big 12" depleg. At one point I was simply pouring in near boiling water to simulate the load (to get the big valve to open near 100%).

It's not a pretty test rig, but I wasn't looking to spend a ton of money on it, so I repurposed all sorts of plumbing I had laying around. While it's not cheaper by any means, I seem to get a better feel for these things once I see them in action, versus trying to worth through the mathematics behind them. Ass backwards and expensive, I know. But until you are hands on, you don't really get to understand all the "real-world" problems.

Mickiboi

Sorry Lloyd. @grim knows what I am talking about, its a very hard to explain. Next time I will send something like that in a PM to him.

Drunkas

@Mickiboi why could you not just stay with a single phase pump and just have a pressure buy pass back to your tank. Might be cheaper than using 3 phase

Lloyd

@Mickiboi said: Sorry Lloyd. grim knows what I am talking about, its a very hard to explain. Next time I will send something like that in a PM to him.

Oh my no, please post it all here, Micki, for those that understand such things.
Just poking at Punkin because he and I are simple county boys that are better at making stuff work than understanding the guts of electronic do-hickeys.

I think we were both beaten by our parents as small children for taking the TV apart to see what made it tick. Never quite figured it out but I will tell you it is best not to poke around on the inside of one with a butter knife even if unplugged. (The scar is almost not visible now).

fadge

@Drunkas said: Mickiboi why could you not just stay with a single phase pump and just have a pressure buy pass back to your tank. Might be cheaper than using 3 phase

I have to use a pressure bypass when starting my unprimed submersible pump otherwise it cant get the lift/flow going. Once it takes off then I can wind the bypass back to either off or trickle, is that sort off what you mean with keeping the pump rate up a bit, to bypass a little back so it does not have to slow too much?

fadge

Drunkas

@fadge said: fadge

Yes

Mickiboi

@Drunkas We are playing around with a different idea. @grim is trying to get his flow right using a 3 way bypass valve to keep a constant vapour temperature for his 12" CD. Which means it has to go from no flow to full on and everything in between using the 0-10VDC modulating signal from his PID in a linear fashion. I just suggested the VFD and 3 phase pump as it would be cheaper than the 3 way valves as they cost a bit. I haven't tried this out but I have a pump coming and will couple it to a VFD and put it through UZGins dephleg and see what happens. I have no idea if it will work or not but my pump guy says no problem. We shall see.

The CV of a valve is a constant that all valve manufactures use to specify flow rates for a given pressure. If you put 2 X 1/2" valves with a CV of 7 in parallel and then stage one to come on after the first one is fully open, this is equivalent to a 1" valve with a CV of 10. @grim tests showed the 1/2" valve was fine for low flow when it first cranks up but as the heat load increases it runs out of flow. When he uses the 1" CV11 he doesn't have the control when the valve first opens that he has with the 1/2" so it probably lets too much water through and hunts everywhere, never settles down and over condenses. Also the larger you go with a valve the less linear the flow tends to become which means control around the centre position of the valve travel becomes unpredictable to the PID. The PID is expecting that its change in output (Valve signal) will have the same effect on the input (Temperature) regardless of what the output value is. My suggestion was to parallel 2 X 1/2" valves as the simulation software showed that it should work and give him a bit more linearity although he would lose a very small amount of flow at the top end, maybe about 3-4 litres a minute as he will lose a CV of 1.

Mickiboi

Sorry put a wrong figure in there. It should be a 1" with a CV of 7 due to the smaller area. DOHHHHH

Drunkas

I think I know what your talking about. I under stand that the 1" is letting to much water flow and how its making it hunt. Whats the limit on @grim heating. Could it be as simple as running the still hotter so it can handle the high water flow.
If I am still not getting this then dont worry about me and I will just sit in the corner and not say any more.

grim

The goal in the selection would to find a valve where the maximum flow required by the condenser would represent the 100% open position, and no flow required at 0% open.

It's not a matter or works or doesn't work, I think you could get a pretty wide set of valves to work just fine, but not optimal.

For example, the bigger valves with the higher cv - on a small still you would probably be using the 0-25% range, any more flow and it wouldn't really matter, you don't need it.

Too small of a valve and you might not be able to flow enough coolant, or in a deplegmator, perhaps limited in how low you can drop the temp. It's a bottleneck.

The more linear the response (1 unit of additional opening represents 1 unit of additional flow), the easier it is to control with a PID (usually the curve is nonlinear anyway, and sloped something like a 1/4 circle arc).

In reality, there is probably a good set of valve sizes and cv's that would work just fine depending on application. Now, there is a big difference between municipal water under pressure and a recirculating coolant, probably a bigger impact on valve selection than the size of the still.

DocPorter

Has anybody looked into these valves for a proportional valve?

Series ZC Characterized Ball Valves @ Plast-O-Matic Valves, Inc.

You can get different V notches in the ball to make the flow curve more linear.

They look to be reasonably fast acting (6 second 0-90°) and are good up to 158°F. And come in both 2-way & 3-way:

CATALOG EBVA / TEBVA MULTI-VOLTAGE ACTUATOR WITH FAIL-SAFE AND 4-20mA DIGITAL POSITIONER OPTIONS (PDF)

If only I could find a price...