I finally finished "fine" tuning my diy overflow project, which I had been working on for the past few weeks (a few hours here and there!)
I had set out to build a continuous siphon gravity overflow, rather then spend $50 - $70 for a pre-made injection molded model.
I built the skimmer and overflow tanks using food storage containers which are made from some sort of clear plastic. I picked them up at a local discount store (a 6 diff sizes pack for $8), and used a smaller and larger sized for the skimmer and overflow respectively.
I used a Plexiglas scrap bent into a U shape to bind the two containers together. At first I tried gluing the bracket onto the containers, but that proved to be impossible, as neither super glue nor two-part epoxy bonded strong enough for comfort. So I opted for tiny stainless steel nuts and bolts, and copious amounts of silicon caulk
The siphon tube (U-tube) is built from 3/4" sch 40 pvc
pipe and elbows. I drilled top of the pipe and epoxied a check valve into the hole, to allow for starting the siphon.
Originally, I had designed the overflow to use 1/2" schedule 40 pipe for the stand tube, and 1/2" tubing to connect it with the sump, using mating 1/2" threaded plumbing parts and an o-ring to form a bulk-head in the overflow. The tubing ran into a 1/2" pvc
ball valve, and then into the spray bar of my diy'd trickle filter (another post). In the bottom of the filter is a maxijet 1000 (~250 gph
) pump, returning water to the tank via more 1/2" tubing.
This however, proved to be impossible to adjust. The pump was draining the filter faster than the overflow was filling it, and throttling the pump using valves resulted in a slow flood of the filter.
So, I took the overflow apart, and reamed the stand-pipe hole larger, to accommodate a ¾” bulk head, which consisted of a ¾” socket to ¾” MPT, an o-ring, and a ¾” FPT to ¾” socket. I then ran ¾” pipe through a few 45° elbows into a ¾” street to ½” fpt to connect with the valve on the filter. All this pipe is white schedule 40, although I may switch to gray schedule 40 (electrical conduit) for the pipe and fittings, since it’s less visible behind the tank, if/when I build another one of these.
After adjusting the height and design of the stand-pipe, I was able to operate the filter with the pump being supplied as fast as it drained (equilibrium).
The optimal design of the stand-pipe involved cutting it to be even with my ‘target’ water level for the tank, and then cutting four ¼” ‘teeth’ into it, which seem to avoid air-lock better than the smooth round surface of the pipe by itself.
I had originally intended to pre-filter in the overflow by using a sponge fitted over the stand-pipe. However, in a ‘power-failure’ simulation, I discovered the sponge was acting as a siphon, allowing the water level to be drained much lower than the stand-pipe’s opening. This is too bad, since the sponge was serving two purposes, first, it filtered large junk from the water, so trickle filter would seldom need maintenance on its bio-media, and secondly, it acted as a muffler to quiet the gurgling noise the stand-pipe makes. This aspect of the overflow will need more research; perhaps a coarser sponge would prevent siphoning but still make a decent pre-filter.
A second aspect that needs further refinement is the main siphon connecting the skimmer to the overflow. I had originally thought ¾” would be sufficient, since it can move a LOT of water under pressure. However, the main siphon is an equilibrium seeking siphon; hence it has very little pressure, so even the smallest obstruction causes noticeable reduction in volume. I think my next version will include 1” or 1½” main siphon (u-tube) and stand-pipe, to insure speedy delivery of water handle a more powerful pump.
Sorry for the long post … pictures may be coming soon, or I might just wait for version 2 of this project, since I don’t yet have digital.
Thoughts and comments would be appreciated. Questions are always welcome!