Some of marine pics might help Vaughn, most of this comes from there:

Older:
http://www.reefcentral.com/forums/sh...hreadid=966407

General google search:

DIY needle wheel impeller - Google Search

So you can use various materials, DIY impellers are cheap and you can always add the old blades back or buy a new impeller, so you can play around for peanuts really.

Another idea is to take impeller blades off a needle wheel and slip those on to the magnet impeller housing of your powerheads etc.

There must be 400 ways to dissolve CO2, but each has a trade off.


When you approach how you want to add CO2 to the tank, you need to sit down and figure out what you want and like.

and what you do not.

Bad:
I hate cleaning disc.
So cleaning little is good.
So is reduced flow over time.
In the tank.
Need flow, and place it in the direction of flow


Good:
Lots of mist I can add inside or dissolve mostly prior
Fast response time between adding and effects in the tank
Out of the tank
Reliable
Active, no need to place in the direction of flow, already has flow
Small
In line


I can all sorts of stuff inside the tank with a powerhead etc.
But I do not want something in the tank.
Nor something large

So stepping back for a minute and thinking about the big picture helps.
Now you can think about a design that will suit your needs/goals.

Some might not care about a powerhead in their tank, it's for a plant farm, they just want something easy and cheap.
Not everyone is like that though.

My issue is developing a good method like this that works in line with a cansiter filter that folks can DIY.

Mazzeis are great for larger tanks, but they do need pressure, pressure I'd rather not give up if I can help it.

I do not mind some loss.

So you can loop a smaller external pump with a return line and add a needle wheel impeller.

That would fit the bill.

Here's a small pump(you can look around for others):
Hydor SELTZ L25 Pump 265 gph (Saltwater Aquarium Supplies > Water Pumps > Submersible > 0 - 500 GPH )

External operation, fed the CO2 into the loop prior to the needle wheel pump.
The in/out of the flow for the needle wheel should not affect the overall flow of the CANISTER AT ALL.

It just grabs water in the line and adds it back at a certain rate, the pressures should be the same, actually slightly higher at the end closest to the return.

This circumvents any Venturi issues for smaller tanks and cost a little less.
It also has no flow drop or pressure drop associated with it.

Just buy an external small pump, mod the impeller, and you are done.
Add in line in a loop in your canister set up, use those in line quick discconnect they sell for you canister to Tee off and control the flow to optimize the CO2.

Main return line_________________
I I
X----------------

The solid line is the return line, the I's are the Tees.
The X is the pump with the needle Wheel, and the other I is the return to the main line.

It's fairly easy to add a tight fitting seal from air line to canister tubing for CO2 injection port. Cut the hole extra small and heat a bit, then work it into the canister return tubing very tight, this should be water proof to about 15psi.
Some places sell small Tee's with 1/2 ID and air line size Tee ports.
Adding some E6000 glue over this can help and that stuff is good for sealing up pesky leaks. If you glob the E6000 on there, then wrap the glue in plastic wrap to get a good seal, this can help.

Add a check valve to the CO2 line obviously.
Rex Grigg sells very nice brass check valves among other things that might be of use to folks.

This is what I plan on adding to one of my tanks that lacks a sump.
The cost of the Octopus is that of most powerheads and it has a nice needle wheel already, so for the $ and the application, I have a hard time beating it.
Sump flow rates are also much more consistent than canister filters and a good surface skim is always nice.

So for some tanks, it works well in line, some, I can add to the return port.
Easier for the sizes I deal with than mazzei.
Mazzeis are more useful I think, given the trade offs, for larger tanks or folks who like to mess around.

I can still get similar if not better results with needle wheels and use less electric and flow loss.

If you are not using say 20-40% of the pump flow, then mazzeis are good for larger tanks.

I think it is important to note, that one is not better than the other really, they just have trade offs.

Listing the advantages and disadvantages fairly to anything is the key to good advice.

Regards,
Tom Barr

MaxiJets can supposedly be run inline, and they're cheaper than dirt. I've never used one, so don't take my word for it, but supposedly they have an O-ring sealing the volute to the motor block. Dunno if the volute retention mechanism is all that great.

Dirt in this hobby can be expensive, though.

Today I melted holes in the blades of the impeller of my little, cheap Catalina powerhead - my hands aren't steady enough to put more than 4-5 holes per blade, and they are kind of randomly spaced. Here is how it looked when I got through:


When I had the powerhead back together I installed a small diameter CO2 line into the suction area and used an airline reducer fitting to get it back to normal airline size. I stuck the assembled unit in a plastic bin of water and ran it - normal flow of water, no suction on the CO2 line. I blew into the line gently and it filled the container with very small bubbles.

It is now in the tank, with the CO2 line hooked to the vent valve on top of my external reactor. About once every 15 seconds a big bubble of CO2 goes to the powerhead and it generates a mist of CO2. I had to rush out to refill my CO2 bottle then, so I'm still not done adjusting the bubble rate to get my drop checker just green. Interesting idea!

I also found that a Rio2100 pump only costs about $35 and you can get a needlewheel for it for $10 - another relatively cheap way to get a needle wheel pump.

i used the bioring mod on my otto powerhead, with the co2 line attached to its venturi (similar to rio's). it seems like the co2 is mistier. it's been running for only 1 full light cycle, so can't tell if it's helpful to the plants.

however the flow appears to have slowed somewhat and the noise of the otto has gotten louder. i'll have to trim the bioring down a little more to reduce the noise. if that doesn't quiet it down i'll try to drill small holes in the impeller.

After reading this and the links Tom posted I noticed one of the designs used a mess, so Iwent out hunting for some supplies and found that it seemed kinda of pricy at the moment(extra tools). So I sat on the couch with an extra pump I had and got a spong out and tied it to the impeller, Got a bucket out and stuck it in and worked pretty good till the pump couldn't get enough air and wold lock up. So I sat back on the couch and tried tieing a plastic scrubby pad to the impeller, tested in the bucket and worked. The only flaw was the scruby seemed to be falling apart the longer I ran the pump. So again back to the couch and found a old algea scrubber (one of those blue ones) and tied it to the impeller, fishing line was hard to get through this so I used a thumb tack to get a hole, this worked really well didn't fall apart. So I got out my Maxi-jet 900 and tied a algea scrubber to the blades on the impeller, just adding the algea scrubber to this decreases the bubble size by over 50%. This may not work for all pumps if their isn't alittle extra romm in the impeller space but dosn't seem like alot of scrubby is needed. I cut the scubby in half and just trimmed around the impeller. I have some expereince with fans and being unbalanced and the disc and scrubby seemed that if you didn't get the disc glued on pefectly at a right angle with the magnet you could through off the balance, so I think this way might be a bit better with no drilling or modifing the impeller. Not really sure what the size of holes in the scrubby would have on the bubble size though. If anyone wants me to post pics just shout.

That's the ticket. I must say that few holes doesn't make for a huge difference, but you should get a bit of mileage out of it. I'm experimenting with another impeller right now. Call me crazy but i am actually going to use some stainless steel needles!

Ah, a needlewheel made with needles?

If my hands were steady enough, and if I had a Dreml tool, I would try slitting each paddle from the OD to the hub, making each one into 2 or 3 narrower paddles, staggering them first 2 then 3 paddles. But, neither condition holds, so I may decide to melt some more needle holes in them instead.

I would like to see photos of the rotor modifications being made and tested. It is an interesting subject.

Drilling holes works well, better than making the disc mesh, it's fine if you can get the right sizing, but the drilling of a pre made impeller is very easy, and it fits well and do not reduce the flow nearly as much.

I do not like disc diffusers as they are inconsistent and have sometimes a lot of backpressure, even the ADA, so on one tank, the CO2 does not start or finish for 1-2 hours.

Unless I use a timer separate for the solenoid, I cannot get the right CO2 at the right time.

So I have issues with the disc as much as equipment inside the tank.
I have some disc that work well, but only a few.
Cleaning weekly etc did no good.

I have a couple that are decent, but I must have went through 20.

The pore size, backpressure etc, are not consistent.

Regards,
Tom Barr

I must say that in my thinking of mesh types, I envisioned an impeller where I took the blades and wove some stainless steel scrubber material (you know, like for washing dishes in your kitchen sink) through carefully cut holes or slots in the impeller blades. Now that would surely make for some serious bubble slicin' and dicin' action! But here's the problem with that and the mesh. Little holes and fiber fill, any kind of mesh, any kind of membrane with little passages like sponges... all these things have a dreadful little unfortunate quality when it comes to the needle wheel concept -- they clog up. That means you're going to have to keep an eye out for reduced flow on a regular basis, and probably clean out the needle wheel impeller about once a week, maybe two. The only way I can see getting around this is by using some kind of very fine pre-filter for the power head (the via aqua 1300 has just such a thing), but I have to clean out those power head sponges every three weeks anyway. Getting gunk into your impeller housing such as hair or sinewy plant matter could really do some damage I bet. You'd get that stuff wrapped around and around and either eventually snap a weak impeller shaft or surely start bringing the impeller blades off balance...scaping...popping sounds...grinding...damage...weeping and wailing and gnashing of teeth, hellfire and brimstone...ad nauseum. Bah. So when you take this into consideration you go dang, that meshy stuff would really tear into those bubbles, but wow what a problem that could create. A maintenance nightmare. Is it worth it?

My staggered needle idea would be less prone to clogging and still have the benefit of speedy fine bubble crushing. The idea is to get a rigid, small diameter material arranged around that impeller shaft. The smaller, the better bubble crushing edge. The more numerous, the better the overal bubble crushing. It seems there's just an inverse relationship (a trade off) between how many rigid, small-diameter obstructions you can put into that impeller space, and how much clog/cleaning you're going to get/do. Mesh puts a heck of a lot of small rigid obstructions into that space. Needles not as much, but close enough, and without the quick-clogging property of mesh.

I will make sure I post pics of my endeavors going forward, and their results.

I want a dremel tool too. Dang. Those things are so cool.