Description
I call the water vortex generator a centripete because the centripetal movement of the water is utilized. It is a very simple construction. Ordinary handtools only are required. The main vessel consists of a vertical 400mm long plexiglas tube; 100mm in diameter. 

At the bottom a disc is glued to the tube; at the centre a hole is made to accomodate an outlet tube with 10mm dia. 20mm above the bottom disc a hole is made in the main tube for a second outlet tube with the same dia. This hole and tube should be offset to the left about 30deg from the radial. 60mm below the brim of the main tube a 10mm dia hole for an inlet tube is made which should be offset about 30deg to the right to give the water a rotational impulse. (I used three inlets spaced at 120 deg intervals but that extra complication is not necessary for the generation of the vortex. The inlet speed of the water, however, will be three times as high which may affect the result of the separation). For convenience, I also fitted an overflow outlet some 20mm below the brim (not here) . I used short brass tubes for the different inlets and outlets with plastic tubing of suitable length forced on. The brass tubes were glued on to the main tube and bottom disc with "Super epoxy".
The main tube should then be mounted on a support to accomodate ample space for the central outlet at the bottom. I arranged the plastic tubings to allow for a water flow of about 3lit/min in each outlet. The contaminations was injected into the main flow prior to the inlet by a shunt arrangement to the main tubing from the tap. When the water is turned on I let it climb to the spill over outlet. A wonderful looking vortex is formed very quickly.
The photo shows a close up of two of the intakes to the centripete and the centre of the vortex.
 

 Results of experiments
Separation
97 documented experiments have been carried out to investigate the possibility to use the cycloidal water vortex to separate coarse grain coffe from water. As an average 97% (minimum 95%) of the coffe which was injected into the flow prior to the vortex was separated out through the centre outlet.

The intensity of the pollution ranged in different experiments between an estimated 4.000 and 15.000 particles per liter (ppl) in the inlet flow. The outlet flow, however, was evenly divided between the centre and the periferal outlets. This means that the content of particles in the clean water (peripheral) outlet went down to 240 and 900 ppl respectively (6%) as an average. The density of the coarse grain coffe used for the first time corresponded to a vertical velocity in still water ranging from about -3 to +8 cm/sec. The majority of the grains was lighter than water (+value).

The photo shows the lower part of the centripete. The coffe is drawn to the centre of the vortex. The outlet at the back of the centripete is not operative. The cleaned water outlet can be seen at the lower right part.

The separation of oil was also very succesful but I did not have the laboratory means to quantify the contents of the two outlets. Looking at the process it was evident that almost all oil from the inlet was sucked into the center of the vortex and drained through the centre outlet.

It should be noted that the experiments were conducted with a centripete not optimized for the purpose and carried out with a single unit. A cascade of units in series would naturally further improve the results.

Drinking water
When I treat clean tap water in the centripete for drinking purposes, I close the outlet at the side of the tube; only the centre outlet at the bottom disc is used. Several people handy with a divining rod have told me that there is a very strong force field around the centripete out to about one meter and an even stronger field above the funnel of the vortex spreading out upwards.-- An American scientist recently mentioned on the vortex-L list on Internet that he had made a theoretical study of a water vortex. He ended his message stating: "I have noted with some interest the concepts of Viktor Schauberger in which he proposes that energy may be produced through the negative entropy generating effect of a water vortex. According to my analysis such a vortex may possibly produce negative entropy". --Be that is as it may, we always keep several liters of swirled water in our fridge and have done so for years. My wife has learnt to operate the centripete to make certain there is an ample supply when I am away from home. 

I have recently designed av very simple PET centripete for the treatment of drinking water.

Check also
http://home.5.swipnet.se/~w-58759/index.html
and 
www.newphys.se