Question:How useful is this fluid level detector design?

[Grubmeister]

Guys (and gals),
I've been considering a design for a nearly continuous fluid level detector. I'll be honest with you, I've brewed in the past but as of yet, have not finished my herms system. For this reason, what I am leading up to may not result in a workable solution. My interest in a fluid level detector is primarily practical, as I'd like one myself. It's also part of my hobby, thinking up new designs for things I'd like to own. I know there are plenty of fluid level detectors on the market. Some with simple single point detection, others with expensive, exotic designs (optical, sonar, etc.) but I'm looking for something I can build that provides the level of performance useful for homebrew at a construction price that won't break the bank.
There are several cascading questions I have regarding fluid level detectors. These questions are directed primarily at rimsers and hermsers (are those even real terms?), but open to input from anybody. My initial goal is for 5 gallon batches, although since I'm targeting 1/2 bbl kegs, it could be up to that capacity.

Before I start, if after reading this post, please let me know if this has already been done. I'm patient but am not interested in re-re-inventing the wheel. Point me in the direction of a design that performs this function and I'll glady head that way.

I'll start from the top:
1)The general question is how useful would/do you find a fluid level detector in your system?
If you use single point (or just a few point) detector systems, how useful would something with more points be?
Where is the location this helps most?
Where have you found, to your surprise, it does not help or is especially problematic?
2)Do you think a fully continuous detector would be useful or one which provided a few desired measurement points? For example, do you think you need to know the fluid level at all times, or perhaps at gallon increments or just a few predetermined levels that have empirically been shown to be useful?
What accuracy is reasonable to expect (hope for) from a fluid level detector? Is .1" (.1 gallons in a standard 1/2 bbl keg) accuracy "good enough"? Is that serious overkill in accuracy?
3)Is there a shape you would consider to be ideal for a fluid level detector?
Would a 1/2" - 3/4" tube or square pipe be ideal?
If the design were a tube type, what size/diameter would you consider absolutely too large to be usefully implemented?

I'm open to suggestions, assuming they're not harshly worded "forget about it idiot" type comments.
There's no guarantee this is going to work but I think I have some workable ideas. Thanks in advance for all responses. I appreciate them.

-Grubmeister

 

[MacBruver]

If you use single point (or just a few point) detector systems, how useful would something with more points be?

What do you mean by more points? like being able to monitor the BK and the HLT simultaneously?

Where is the location this helps most?

HLT first, BK second

Where have you found, to your surprise, it does not help or is especially problematic?

Not really all that useful in the mlt, except for at dough-in... kind of a waste, there

2)Do you think a fully continuous detector would be useful or one which provided a few desired measurement points? For example, do you think you need to know the fluid level at all times, or perhaps at gallon increments or just a few predetermined levels that have empirically been shown to be useful?

I don't think a "few predetermined levels" would work- there are far too many factors when you consider different recipes and batch sizes that people might do on a single system, let alone between different systems.

What accuracy is reasonable to expect (hope for) from a fluid level detector? Is .1" (.1 gallons in a standard 1/2 bbl keg) accuracy "good enough"? Is that serious overkill in accuracy?

I think that's good. If it was a system that cost some money I'd expect it to go down to 0.1 gal

3)Is there a shape you would consider to be ideal for a fluid level detector?
Would a 1/2" - 3/4" tube or square pipe be ideal?


If the design were a tube type, what size/diameter would you consider absolutely too large to be usefully implemented?

Hard to answer the question without knowing what sort of device it is.

So... what is it? a pressure transducer or something?

 

[Grubmeister]

Thanks for the reply,

What do you mean by more points? like being able to monitor the BK and the HLT simultaneously?

Well no, I had in mind a single device. Each tank would require it's own device.
Of course if I get this right, they would be addressable and make data available to a central control unit. By "points" I was referring to two things. First, the resolution along the tube (for example, .1" increments or .25" increments). I may have used it in another context where I was suggesting fewer level detection "points", for example, where it would be able to detect fluid at the 1" point, the 12" point, etc.
My fingers think fast, just not thoroughly. Kind of like my kids, who take after their Dad.



HLT first, BK second

Makes sense. I suppose most BK's are dealt with manually. At that point it gets messy and having a tube run into the BK, even run down along the wall would be a hindrance.

Not really all that useful in the mlt, except for at dough-in... kind of a waste, there

Now that I'm thinking about it, it would be most useful in a batch sparge run. You're right though, it would be kind of useless during fly sparge, unless you went all the way to maintain a target fluid level, turning on/off either a pump or a valve. I don't have the exerience to know if having a tube down the side or middle of the MLT would invite problems with channeling. There might be other issues I can't foresee.


I don't think a "few predetermined levels" would work- there are far too many factors when you consider different recipes and batch sizes that people might do on a single system, let alone between different systems.

Fair enough. I suspected as much but like I said, I don't have the experience to know. I'm relying on the veterans to tell me if they've empirically identified certain common levels. I guess not.


I think that's good. If it was a system that cost some money I'd expect it to go down to 0.1 gal

I don't think this is going to cost all that much. That's why I was hoping .1" would be adequate.


Hard to answer the question without knowing what sort of device it is.

Sorry, I didn't mean to be quite so cryptic. After re-reading my post, I do sound like I'm hiding some sort of high tech trade secret.
It's the same basic idea as you've seen before with other simple detectors. Point to point semi-sine wave detection but using a larger array of pins to emit and detect the signal. After thinking about it for quite a while I narrowed down the basic problems with this type of setup. To just make an array of the same circuits would require too many parts. It would be huge and really difficult to build. I haven't fully finished my schematic but I'm convinced I can do this with a microcontroller, a few inexpensive components and a decent algorithm. Building the physical tube was the primary problem, which I think I may have solved the other night. I'd really like to prove it first before I exclaim "viola" as it seemed a little too easy when I had it working inside my skull.

So... what is it? a pressure transducer or something?

I'd like that too. Unfortunately, I don't have any actual experience with pressure transducers. Right now that's beyond my abilities. Maybe this one won't work and we'll try that next.

 

[MacBruver]

"It's the same basic idea as you've seen before with other simple detectors. Point to point semi-sine wave detection but using a larger array of pins to emit and detect the signal."

I'm a software developer but that's still greek to me. Based on the rest of what you're saying- it would essentially be a sight glass still? If that's the case, you might want to look into capacitance based sensors. The resolution of those just comes down to your A/D converters at that point.

http://www.discovercircuits.com/DJ-Circuits/capgage.htm

 

[Grubmeister]

"It's the same basic idea as you've seen before with other simple detectors. Point to point semi-sine wave detection but using a larger array of pins to emit and detect the signal."

I'm a software developer but that's still greek to me. Based on the rest of what you're saying- it would essentially be a sight glass still? If that's the case, you might want to look into capacitance based sensors. The resolution of those just comes down to your A/D converters at that point.

http://www.discovercircuits.com/DJ-Circuits/capgage.htm

Gotta get out of here to meet my father in law, who just came into town so I'll make this quick.
the most basic circuit is square wave being generated by a microcontroller. The square wave is capacitively coupled into a stud. The fluid completes the circuit to another capacitively coupled stud. Through a schmidt trigger and back into the microcontroller, where the signal is detected. Terribly simple stuff and not much smarts required by either me or the microcontroller.
Thanks for the link. I'll look at it later tonight when I have time.
-Grub