A small downdraft skimmer

by Chris Paris for Reef Aquarium Information Depot

This page describes my small downdraft skimmer, which I completed in April 1997. This skimmer has serious problems. You can avoid all the problems just by knowing about them and not duplicating them. I have a photo essay of the changes I made to the skimmer here. The basic design of the skimmer is still okay, although I have converted it to something more HSA-like by replacing the simple tube injector with a Beckett foam head, removing the bioballs, cutting out the portion of the injector tube that is inside the box, and cutting out the baffle (I get fewer bubbles out the skimmer with no baffle than with the one I had).

Perhaps the most important change I made was to replace everything above the foam tube flange. In place of what you see in the pictures below on this page, I have a larger collection cup (8.25" inside diameter), and inside the collection cup I have an inverted funnel acting as a foam riser. The taper of the funnel provides an even transition from wet foam to dry foam, and the narrow neck of the funnel results in high air velocity, which ejects the dry foam quite forcefully. The previous problem of foam getting stuck exiting the foam riser is completely gone now. The large collection cup sits directly on the flange gasket.

If I weren't so busy at my new job, I might be designing and building a new skimmer. Back in 1998, I wrote down my thoughts about skimmer design, and some of the specifics of a new skimmer that I was planning to build. That document describes all of the problems with this skimmer, except for the misplaced baffle problem, which I describe below (I already mentioned that my solution was to cut it out entirely). For your own sake, please don't copy the design that follows without reading the new document.

Click on a small image to get an enlargement of it.

[Thumbnail picture] [Thumbnail picture] [Thumbnail picture] [Thumbnail picture] [Thumbnail picture] [Thumbnail picture]

I wanted to make a small downdraft skimmer that I could run inside my sump, under my tank. I collected the dimensions of several commercial downdraft skimmers, as well as several downdraft skimmers made at home by other hobbyists. I was influenced by all of these other skimmers, but mine isn't a copy of any particular skimmer. I made a few changes that I thought would help make it work better sitting in the bottom of the sump, rather than sitting above the sump, as is the conventional way to run a downdraft skimmer.

Usually my only concern when building a piece of aquarium hardware is that it work well. This time, I wanted to try making something that didn't look homemade. I found that there's a big difference in effort between making a downdraft skimmer that only works well and making one that works well and looks like the commercial ones. I wasn't punching a time clock during the design and construction period, but I estimate that I have about 40 hours of work into this. I could make a second one in slightly less time, but it would still be very time consuming without buying several key tools that I'm missing (drill press, rotary table, tablesaw, jointer).

In the first picture you can see the overall plan of the skimmer. The box at the bottom is about 12" long, 7" wide, and 10" tall. It's made from 1/4" acrylic sheet. The 10" height is about 4" taller than most other small downdraft skimmers. I thought that this would reduce the amount of air bubbles that exit the skimmer through the water output, and this seems to be true.

The slanted internal baffle has the purpose of reducing the amount of air that leaves through the water output. I placed mine with its upper right corner (referring to the picture) one inch below the inside top of the box and one inch from the side wall of the skimmer. The angle of the baffle worked out to about 53 degrees off of horizontal. Most modern downdraft skimmers have a baffle design similar to mine, except that the baffle is at a lower angle (about 45 degrees) and the top of the baffle lines up with about the middle of the foam riser tube, rather than only 1" from its end as in my skimmer. I thought that my design would keep more bubbles where you want them, but after seeing mine work I think I might have been better off sticking with the conventional placement. Still, I don't have very many bubbles leaving compared to some skimmers.

The water exits the skimmer through 2" plumbing. Most small downdraft skimmers use 1-1/4" plumbing, but some reef keepers mention that they have to elevate their skimmers substantially above their sumps, with a siphon tube running between the skimmer and the sump, in order to suck enough water out fast enough to keep up with the pump's input rate. I addressed this problem by using oversized output plumbing. It works as intended.

I purchased the 2" gate valve from US Plastic Corp. for $19.50. The "bulkhead" is just a 2" male thread to slip adapter, which I glued into the hole in the side of the skimmer with Weld-On 40 acrylic cement. The 2" PVC elbow serves the purpose of picking up the water from the point furthest from the bubbles: the bottom of the skimmer, near the inside corner formed by the baffle. This elbow is called a "street elbow," which means that one of its ends has the diameter of 2" PVC pipe, so that you can plug it directly into the "bulkhead." An alternative is to use a conventional elbow with a short section of 2" pipe to join the elbow to the "bulkhead." My solution gives a slightly more compact form factor.

In the fourth picture you can see the holes at the bottom of the bioball tube. In every downdraft skimmer that I know about, the bioball tube goes completely down to the bottom of the skimmer box. This is to keep the bioballs in the tube. In order to let the water exit the tube, you have to have one or more holes in the bioball tube. These are traditionally located near the bottom, facing away from the baffle. A problem with some skimmers is that these holes aren't large enough or numerous enough. The result is that water backs up in the bioball tube and threatens to come out the top. It seems that these skimmers will pull in the most air when there is no standing water in the bioball tube.

I wanted to be able to run a lot of water through my skimmer, while retaining a relatively short bioball tube (it needs to fit under my tank stand). My water escape holes are not actually holes, but slots. There are three slots, each 1" tall and encompassing about 120 degrees of arc. The slots are separated by 1/2" vertically. If you're bored, calculate the surface area of my slots and compare that to the number of 1" holes you'd have to have to equal that surface area. I'd like to know the answer, as I didn't bother to do the calculation.

The top of the bioball tube is 23-1/2" above the bottom of the skimmer. The injector cap raises the height to 26". The bioball tube is 2.5" inside diameter by 2.75" outside diameter extruded acrylic. The fifth picture shows a view of the injection mechanism. The bioball tube is topped with a 4-1/2" diameter flange. This provides for a secure mating with the injector cap, and also allows me to extend the length of the bioball tube in the future, by building a length of tubing with a flange at each end.

The injector cap has a flange on the bottom to mate with the flange on the top of the bioball tube. Sitting atop the flange is 2-1/2" (or so) of 2.5 ID acrylic tubing, which is permanently sealed on top with 1/4" sheet acrylic. That 2-1/2" length of tubing has two connections on it. One is a 3/4" ID by 7/8" OD tube going straight through the wall. This serves as the air inlet to the skimmer. The other is a 90 degree elbow made with the same 3/4" tubing. This serves as the water injector. I made the elbow myself, by cutting a length of tubing at 45 degrees on my bandsaw, sanding the cut ends on sandpaper glued to a sheet of glass, and gluing the pieces together with acrylic solvent. The injector is simply this 3/4" tubing. I bring water to the skimmer by 3/4" flexible tubing, and the water is injected into the skimmer through the 3/4" ID acrylic elbow. There is no reduction. I've not tried any other injectors.

When I first started using the skimmer, I was driving it with a Quiet One. It's not a pressure pump, but this isn't a pressure application---the pump just has to lift water about 20" above the sump level, and release it through a 3/4" tube (the same diameter as the pump outlet). There's less pressure to the pump than if I were using it as a tank return (since the tank is higher than the top of the skimmer). But I wanted more flow, so I built a new injection cap that has two 3/4" injection tubes. I have two Quiet Ones injecting water now. One of them is also the circulation pump for the tanks, so only about half of its capacity goes into the skimmer. I would guess that the total water flow is about equivalent to a Gen-X, but the Gen-X might be more.

The foam riser side of the skimmer is completely made from cast acrylic. The bottom tube is 6" ID by 6.5" OD by 6" tall. It is glued to the top of the skimmer box over a 6" hole. That is, the diameter of the hole in the top of the box matches the inside diameter of the bottom foam riser section. There is a flange, to enable disassembly and cleaning, just above this first section of tubing. The flange is made from two rings of acrylic. Both rings are 3/8" in thickness and have an outside diameter of 8-1/2". The ID of the bottom ring is 6", to match the ID of the bottom tube, and the ID of the top ring is 4", to match the ID of the thinner riser tube.

To clamp the two halves of the flange together, there are 12 holes around the perimeter of the flange. 12 is probably overkill. The holes in the top half are 1/4" diameter, and in the bottom half there are 1/4-20 machine screw threads. I drilled the two halves of the flange in a single operation, with the two halves tacked together with hot melt glue (with the protective paper that comes on acrylic still in place). The holes take 1/4-20 by 3/4" nylon hex head screws, purchased from US Plastic Corp.

The best way to lay out and drill the holes for the screws would be to use a rotary table or dividing head and a drill press. I don't have either of those, and they're expensive. The holes don't have to be laid out perfectly evenly around a perfect circle, though if they aren't you'll be able to fasten the flange together only one way (the one where a hole in the top half lines up with the hole in the bottom half that it was originally lined up with when the holes were drilled). I wanted my holes to look like they were laid out perfectly, even if I didn't achieve the precision that I could have if I dropped $300 on a rotary table. My solution was to write a short Postscript program (Postscript is a sort of page description language that some printers use). The program draws a figure like the one at the front of this paragraph. It is parameterized by the OD of the flange, the ID of the flange, the diameter of the circle through the middle of the screw holes, and the number of screw holes. I run the program by printing the program with certain parameters plugged in, cut out the OD circle, and glue it with rubber cement to the acrylic ring(s) that I want to drill. The Postscript program is here. I still can put my flanges together only one way, though the error is small so I could get around this by drilling the top holes slightly oversize.

The 4" ID section of the foam riser is about 7" tall, and goes through the bottom of the collection cup. There are two ways to construct this. One is to have a single 7" length of tubing, and a hole in the bottom of the collection cup that matches the 4.5" OD of the riser tube. In this approach, the weight of the collection cup is supported by the glue joint. The other approach, and the one that I took, is to make the 7" riser in two sections. The bottom section is 4" tall, with the bottom of the collection cup glued to the top of this first section. The diameter of the hole in the bottom of the collection cup is 4", which is the ID of the tube. Then a 3" section of the 4" ID tube is glued on top of the bottom of the collection cup. So the bottom of the collection cup is sandwiched between the two pieces of the 4" ID foam riser.

The wall of the collection cup is a 4" length of 6" ID by 6.5" OD cast acrylic tube. The lid to the collection cup is watertight and airtight at the pressure that exists inside the skimmer. It is capped by a two piece acrylic flange, similar to the one that allows the foam riser to be taken apart for cleaning. The flange's OD is 6.5", the OD of the collection cup (thus this flange doesn't overhang its tube, as the lower flange does). The top half of this flange is a solid acrylic disc, not a ring. It's sealed with eight 1/4-20 nylon screws.

Scum drains out of the collection cup through a 3/4" ID by 7/8" OD acrylic tube that goes through the wall of the collection cup, at the bottom (it doesn't come through the bottom as in other downdraft skimmers because the close clearance between the OD of the riser tube and the ID of the collection cup doesn't leave enough room for the relatively wide scum drain I chose).

All flanges are sealed by 1/8" thick neoprene gaskets that I cut myself, with ordinary household scissors, from bulk sheet material purchased from MSC. I had to cut holes in the gaskets for the screws to go through. I did this with a custom made manual "punch" that I made on my lathe (it's just a thin walled tube that I press down onto the rubber sheet and twist with hand pressure).

Notice that the skimmer has only four orifaces. On the bioball tube we have the air input and the water input, on the collection cup there is the scum output, and on the box there is the water output. This means that all air must enter through the single air input tube, and leave through the scum pipe (except for the tiny amount that leaves in the form of bubbles in the water output). This enables the user to filter all of the input air by connecting an air filter to the air intake. This also virtually silences the air sucking sounds and the gurgling sounds that exist inside the bioball tube. This also enables the output air to be filtered through a carbon filter, which can reduce the stink of a jar full of skimmate. Finally, it enables collection jar designs that block off or put back pressure on the air output line when the collection jar is full. This prevents air from entering the skimmer (or reduces the flow greatly), and that kills the foam in the skimmer, which prevents it from overflowing the collection jar.

You can make an auto-shutoff collection jar. AETech (maker of the ETS) sells one, as does Marine Technical Concepts. It's a sealed cylinder with one input (for the skimmate hose), and one output. The output just has a floating check valve in it. They do this with a ping pong ball captured in a tube. When the skimmate rises to the level of the ping pong ball, it pushes the ball up so that the ball covers the air exit hole in the top of the collection jar. This more or less completely seals off the air exit, which means that no air can enter the skimmer, which means that the foam should immediately disappear altogether. I've been meaning to build one like this, but I have more important projects in line first.

Other DIY skimmers

Add or View Comments

Chris Paris
Last modified: Thu Jan 18 21:29:55 Eastern Standard Time 2001