I modified my ASM G3 (Sedra 5000 pump) to a meshwheel skimmer this weekend. Here's what I did:
1) Cut off all "needles" from the impeller except for 4, 1 each at the "12:00", "3:00", "6:00", and "9:00" positions. (My thinking is that this will help prevent "slippage" of the mesh as the impeller turns.)
2) Cut a 3 squares of Enkamat PF4 mesh material, large enough to cover the impeller wheel. Each is 1 layer.
3) Trim the mesh down to the impeller wheel.
4) Weave some fishing line into the mesh to attach. (Some people use zip ties, I thought this might work a little better).
The impeller did not immediately want to start turning, which makes me think I may not have trimmed it back enough. However with some fiddling I was able to get the impeller to turn when the pump is powered on, so I left it. I wonder if 2 layers may have been better than 3 for this reason (well that and one other reason, more on that below the pictures).
I can safely say that this mod does make the skimmer produce more foam, so I am pleased with the modification (and relieved that I didn't wreck an impeller for nothing - as replacements cost around $50). In 24 hours my skimmer has produced a full-on inch of skimmate, which is hands-down the best it's ever performed.
However, I did some before and after airflow measurements with a 0 to 25 lpm range airflow meter from Muis Controls, and the #'s don't show a difference. Before with the "stock" impeller I read about 5 lpm nominal (with spikes to 10 lpm), which is exactly the same readings as with the modified impeller. So I'm somewhat at a loss to explain why people on RC claim improvements from 5 to 12-18 lpm, however I suppose there are many factors/variables at play, such as venturi design, pump model, how the pump is plumbed, etc. Based solely on the visuals and not the measured airflow numbers, I know the skimmer is producing more foam than before, so I'm pleased with the mod.
However, I am left wondering if this may also be an argument to trying 2 layers over 3, perhaps the mesh density is a little too high to prevent an "even better" air suction.
Over time I may experiment with other air injection methods to see if I can improve on the air flow #'s, but for now I'll leave that as a project for another day.
