Re:Water Flow and DYI CO2 reactor

> From: ibi007 at lion_connect.more.net (Allen Sandifer)
> Date: Mon, 14 Aug 1995 16:30:08 -0500
> Subject: Re:Water Flow
> John let me know how you DYI defusser works out. I have been using a little
> defusser that is made by Sander. It works by either flowing water across the
> bottom of it or by pumping air to it which causes the water in the defusser
> to circulate. I am still not totaly happy with it. I have been consider
> trying to pump water through it from the filter or reverse the action by
> putting an air line in the intake if the filter and the attach the other end
> to the outlet of the defusser. I am thinking that this would create a vacum
> and circulate the water through it. 
> Allen

My DYI CO2 reactor works very well. As I described briefly last time, it
is made from part of a typical gravel vacuum tube. The design is pretty
much based on a number of people's as described in the archive and else
where. I think mine is also functionally similar to a commercial unit
(Dupla?) I saw in a local store. 

Take the fat end of a gravel vacuum tube (I had a spare Lee's brand unit),
and fill it with bio-balls (or glass marble balls, the Fluval pre-filter
ceramic bits, or even gravel). Measure the distance between your intended
CO2 reactor location in the tank and the end of your Fluval (or any
canister) outflow spray bar, and cut the attached silicon tubing to
length. I was lazy and left the gravel vacuum the way it was (10' dangling
behind the tank). Connect the tubing end to the spray bar end after you
drill a hole in the spray bar end cap. Cap the bottom (the opening end) of
the gravel vacuum tube using a PVC pipe cap or similar.  You may not want
to glue it if you wish to change or clean the bio-balls inside later. The
cap I used is a perfect force fit. 

Now drill a hole in the center of the bottom cap, and push the airline 
tubing from your CO2 source through the hole and inside the tube. Drill a 
few more smaller holes in the cap to allow water to escape. Basically you
have the following:

   Silicon tubing came with gravel vacuum

                          | w| <- Water from your canister filter
                          |w |
                          | w|           The 'O's
                          |w |           are bio-balls.
                      ____| w|____
                     /      w     \      The 'w's are water dripping down
                     |     w      |      from your canister filter SB.
                     |      w     |
                     | O O O O O O|
                     |O O O O O O |                                            
                     |~~~~~~~~~~~~| <- air (CO2) above, 
                     | O O O O O O|    water below
                     |O O O O O O |
                     | O O O O O O|
                     |O O O O O O |
                     | O O O O O O|
                     |O O O O O O |
                     | O O O O O O|
                     |O O O O O O |
                     | O O | | O O|
                     |O O O| | O O |
                    ||_____| |____||
                    |______| |_____| <- Cap
                       ^   | |  ^
                           | |  drill a few holes in bottom for water to
                           |^|  leave and into the tank.

IMO, this is a quite efficient setup since water first drips through an
almost 100% CO2 atmosphere and then becomes in contact again with more CO2
bubbles fighting their way up through the water and bio-balls before
finally leaving the reactor at the bottom. 

This thing sits at one corner of the tank behind some tall H. corymbosa. 
You can see through the clear PVC and count the number of bubbles when
setting your manual flow rate. You can also use the amount of "air space" 
above the water line to estimate how much CO2 you have in your system. I
have it set at about one bubble every 1.5 second, and my pH is fairly
constant around 6.9, down from 7.9-8.2 without CO2 (10-12 dH KH). Your
results may very. 

 John Y. Ching (jyching at watnow_uwaterloo.ca)    |
 Pattern Analysis and Machine Intelligence Lab	| 
 Department of Systems Design Engineering      	| 
 University of Waterloo, Canada            	|