In searching out information about water filters on the Web, I came across Doulton filters, which appear to be pretty decent. They have single-pass efficiency of better than 99.7% for particles of 0.3 to 0.5 micron, and are better than 98% efficient at removing particles of 0.2 to 0.3 micron -- just about the kind of performance I'm looking for, and I have the luxury of letting the dye solution make multiple passes through the filter, which is dandy. When I went to find out how they work, I discovered that the fundamental material is essentially bisqued porcelain. Well, la-di-da -- I can do that!
I went to the lab and found a teacup with a tiny crack in the bottom that occurred during bisque firing. I put RTV (silicone rubber "bathtub caulk" stuff) over the crack, even though it didn't go all the way through, and went looking for a suitable filter vessel at a nearby thrift.
While I was there, I came upon some decorated flowerpots. Hmmm, sez I. Terra-cotta. Porous. Cheap. Universally available. HMMM!!
...So I bought one, a small undecorated one. Set me back a whopping 33 cents. Of course, there's this hole in the bottom; but we can take care of that.
I also bought A Vessel:
Almost three bucks, an outrage! Should have been two. Oh, well... it was the right size. The flowerpot sat happily within it.
I also had to find something to serve as a lid. Thought about various things, and finally compromised -- this is probably too thin and wobbly, but it will do for an initial test --
Cute, eh? Twice as expensive as the flowerpot, dammit! You pay for cute! All I need is the lid, too... sigh.
I then went to the hardware store and got what I figured would be the requisite fittings:
This gives me a valve on the inlet, and quick-connects at both inlet and outlet. It doesn't show you the three washers I had to add in order to get things to fit correctly, because at the time I took the photo I didn't realize I was going to need any washers.
I figured that the flowerpot was probably kinda filthy, and who knows whether there might be nasty soluble materials in the clay, so I tried to scrub off the price (grease-pencil, ick!), and I boiled the thing in distilled water for about 20 minutes. (I know, not enough; but this is just a test, and it probably won't work anyway, so I'm not as concerned as I would be for a production unit.) I pulled it out of the water and dried it somewhat (a few seconds in the microwave helps drive off some of the water, especially when the object is nearly at the boiling point anyway), took it upstairs, and put a little piece of stainless steel over the hole with RTV --
I then hacked holes in the lid of the tin and attached the fittings. (You'll see them in the next-but-one photo.) At that point I considered the constraints, and decided that a flat bottom with sharp corners (you probably can't see them; they're around the little raised area in the center) was not conducive to good behavior under applied pressure, so I took off the handle, washed the exterior, flipped the vessel upside down, wrapped packing tape around it to form a collar, and poured a quantity of 30-minute epoxy into it, so:
At this point I must interject a comment.
It turns out that epoxy, in its curing process, tends to be somewhat exothermic. When you are only doing minuscule amounts, this is ignorable. I once had the privilege, however, of watching a surprisingly modest quantity of some fairly special epoxy in a little styro coffee cup literally burst into flames (in someone else's lab -- permit me to assure you that I was merely an observer). I'd forgotten about that; and when, some few minutes after I poured the material into the mold, I noticed that I could smell it; that it seemed much less viscous than it had been only moments earlier (the bubbles were rising to the surface and bursting, which they hadn't been doing previously); and then that there was a delicate plume of vapor rising from it, I became slightly nervous.
Ahem.
I opened both front and rear doors to my residence, and allowed the chill winter wind to carry off at least some of the effluent vapor. I also watched carefully to be sure the stupid thing didn't set fire to my nice chair. As luck would have it, however, the epoxy just quietly hardened, and after a suitable delay I carried the vessel upstairs to attach the remaining bits to it.
Here's the completed filter. You can see the flowerpot hiding inside, and the little piece of metal that I've used to block off the [former] pouring spout --
Now, it is a well-known fact that RTV doesn't stick particularly well to glass. I also doubt that it will stick particularly well to crappy paint on thin metal. Thus and therefore, I shall probably put some cable-ties around this device to prevent it from placidly coming apart as soon as I push any solution into it.
Even so, I would guess that if I put much pressure on it, there will be various leaks. That being the case, I'm probably going to run it on a siphon, with the source only a few feet higher than the sink. There is also the fact that it is likely to be dog-slow, so I probably get to let it sit at least overnight and possibly longer. If it succeeds, I have ideas for designs with higher throughput that will handle decent amounts of pressure, and can be added to the recirculator. ...But that's a big "if". We'll just have to see.
Your forbearance in this matter is deeply appreciated.
Thoughtfully yours --
jon
