Note the following things:

• In the injector tube (at left), the Beckett air-water injector has a flange near its bottom. The flange prevents air from cycling between the injector's output side and its air intake holes; once the air gets sucked through the injector, it stays there. This was supposed to increase the efficiency of the air sucking, but frankly I'm not sure it makes any difference. I don't know of another skimmer that uses this technique, and there are plenty of other successful skimmers out there.
• The injector tube stops at the top of the skimmer box. There are no bioballs in the tube, nor indeed is there anything that would keep the bioballs in the injector tube if they were present in the skimmer. Water simply drains through the injector tube and into the skimmer. As you will see in the final photos of this series, the air-water mixture backs up in the injector tube all the way up to the flange (discussed in the previous bullet), so there is no water freefall in this tube. It's more like a pipe containing a mixed air-water fluid.
• The collection cup has a very wide diameter, and it incorporates an inverted funnel as the upper stage of the foam riser. The collection cup just sits on the flange that in normal downdraft skimmers separates the lower from the upper foam riser sections. There are no mechanical fasteners connecting the collection cup to the flange of the skimmer. There should be, but I didn't have a way to put them into this retrofit job. I run the skimmer in the sump, so minor leaks are not a problem. It just so happens that I haven't noticed any leaks anyway.