Different sinusoidal valve idea. With a bad pic!

[BrainBandAid]

You may have seen this incredibly awesome closed loop sinusoidal valve in one of the TOTM's, I don't remember which one...

http://img.photobucket.com/albums/v312/brainbandaid/313_sinusoidal_side.jpg

This one uses one suction from the tank, and splits it to three different returns. The white plastic "valve", which spins slowly inside the acrylic tube, routes the pump discharge to one of the returns at a time. It has three holes bored into it at different radial angles, and along with the small motor, times the discharges along kinda a sine wave without turning the pump on and off.

My tank is (will be :rolleyes: ) an in-wall design, with the display viewable from either side, from two different rooms. What I would like to do is setup a similar device, but instead of one static suction and three dynamic returns, make it so that in one position, the pump takes a suction from one side of the tank, and discharges on the other. Then as the valve rotates, the flow slows (being recirculated through the valve/pump) then the side that was the suction is now the discharge and vice versa.
I think this would make a nice even back and forth motion across the whole tank.

Please excuse my pathetic Paint drawing, but I'm still learning Sketchup. Here's a basic idea, if you can understand it... It's definitely not to any scale.

http://img.photobucket.com/albums/v312/brainbandaid/closedloop.jpg

The holes drilled in the valve are 90 degrees from each other, two in each position. On the pump side, the suction and discharge are always the same. As the valve rotates, there would be a time when both the suction and discharge would actually be short cycled together, thus recirculate. The diameter of the holes vs. the diameter of the valve body would determine how much would recirculate. This recirculation would smooth out the transition from suction to discharge, evening out the flow.

It's probably not practical for a normal tank, but on mine the only place I have to drill the bulkheads is on the sides, which are well hidden inside the wall. (The tank will be 48x18x30 BTW. Don't say it. I know, I know... )

I'm thinking UHMW rod for the valve, clear PVC for the body, with regular PVC fittings for the rest. Probably somewhere around 2-3", with 3/4 or 1" plumbing for the suction/discharges. Around 1200gph.
Doesn't need to be all pretty acrylic. I haven't worked out the seal issue yet, but the modified magnetic pump impeller idea seems pretty good.

I'm still not quite sure how the duel suction/discharge nozzles would look like, maybe vertical spray bars of some type...

So, good idea? Bad idea? Suggestions? Will it work?
Practical or just cool looking?

[BrainBandAid]

Anyone?

[zapata41]

looks prety cool, should work once you figure out how the pump will like having the inlet and otlet blocked at the same time.

just thinking out loud now....

how will the plumbing like the surges in the flow from the change in directions of the water flow. i think you should make it out of acrylic since making a sealing surface would be much simpler since you can bond acrylic to itself and the machine it much easier than you can with pvc.

what about having two bulkheads on each side instead of one. making the intake and return seperate pipes would simplify what the in tank plumbing would be.


seems like something pretty cool to builb. why not go with a bigger pump like a dart or something.

Tim

[BrainBandAid]

I like the two bulkheads on either side idea... less clutter around the valve for starters...

I don't think there would be too much surging in the piping, for the same reason the inlet and outlet won't be blocked at the same time:
As the valve body rotates, the closed holes would be connected to flow before the open ones fully close. My picture should probably have bigger holes to show that...
I think this would make the flow short cycle through the pump instead of going to the tank. That's what would make the flow sinusoidal and avoid surging.

Thanks for the feedback!

[Fluxion]

I have been thinking about doing the same think on my tank for quite a while. I am currently building my second sinusoidal device and my third may be just as you described. My only problem is lack of space. My stand is full and I only have 6" behind the tank. I have ordered some small timing motors to use in my experiments. I bought them here: http://surpluscenter.com/sort.asp?UID=2006110921284176&catname=electric&keyword=MTSM

I have been using the 3/4 rpm version with a magnetic drive.

[BrainBandAid]

Great link, thanks! (And your link for machinable wax on your other thread; that's looking like a better idea the more I think about it)

I was thinking more of around 6 to 8 rpm. For my design, one revolution is one back and one forth motion. Basically one wave. I sat around for a while, trying to imagine what a nice wave frequency would feel like, and that's what I came up with...

You may have seen this wavemaker (http://www.wave2k.com/index.htm) before, but in the faq section, they state one wave every 6 seconds, or 10 waves per second. This would be 10 rpm for my idea.
They've got a pretty cute movie download of fish playing in the surge.

How is your magnetic drive working? Or are you still building?

[Fluxion]

I have a 2 way valve up and running on my nano here at work but once in awhile it stops turning the valve. I haven't figured that out yet.

I think I will add more magnets to the one I am building now. The magnets I have are very difficult to separate but when you put about 1/4" of pvc in between them they don't have much pull.

I spent about an hour measuring my tank at home last night and I don't think I have the room for all the plumbing to use the opposing inputs. I haven't given up yet, I'll figure something out.:D