How to Turn On and Off a Stream of Gas?

Yes, experiment!!!

Find out how much or little pressure it takes to pinch off your hose. Find out how long it has to be open and closed. Find out how large or small of an orifice the valve has to be. It would be nice if you can use standard units of any kind, but any repeatable values will put you onto the road to doing real engineering. (closed = 32 pennies, open = 5 pennies)

D*mn, you beat me to it...

A good way to open/close a pipe without too much power is to have it looped back on itself, a little pressure/movement will kink it shut.
Del

That would work if the wheel was in constant motion, but it's only doing about one half turn each four minutes. It's the motor for a solar tracker.

The mechanism is super low resistance, and about 50 ml of eths should be enough to turn it over, tho I'm working to 100ml to be safe. Each 'squirt' of ethanol is about 4 ml, so I'll need 25 bubble out / bubble collapse cycles to get enough in the wheel cup to turn it all at once.

Are you dumping the ethanol directly on to the wheel? If you used a reservoir to collect the ethanol pumped with a small valve to control flow over wheel you could get constant rotation.

Constant rotation isn't what I'm after. Since it's tracking the sun I want it to kick in only when the sun has advanced by a degree or two. The device corrects once, then does nothing until the sun's moved off again.

However if the sun goes behind a cloud or something similar the device will correct constantly til it's caught up.

Could you do anything for opening & closing with a clock pendulum? Where's Rube Goldberg when you need him?

Maybe, but I'd like it to be as set-and-forget as possible, and having something which swung day in day out for months might take more precision construction than I'm hoping for.

Worth thinking about tho..

How about solar panel to collect electricity, store in a battery, a timer to control valve events, and an electric solenoid valve matched to battery-panel voltage?

The whole thing has to be as absolutely basic in the construction as possible. It's an open source project meant largely for the developing world so people can generate their own energy.

Most if not all materials should be something you could scrounge from an average scrap yard. No tools more advanced than a hack saw and a drill.

SAF,

Could you possibly provide a rough sketch of your system?

Sketch:

The feed alternator would go at the top of the boiler.

Something I'm thinking (not sure if this makes sense)

The slidey thing is a cut in half syringe. The rest is just two springs, a ball bearing, two bolts, one nut and some sellotape.

(Plus pipes, a cork, and a T junction.)

This is to provide a threshold on the opening and closing of the valve, it has to build up a pit of pressure, then opens fully, then the pressure has to drop a bit before it closes again.

I've been thinking of this for the last few days but couldn't think of a way to seal the cylinder til I thought of using a syringe.

Thoughts?

Other ways of doing it?

Very low pressure poppet relief valve. Would open upon reaching the pressure setting of the valve and bleed off the pressure. Resets itself when pressure is bled off.

You could also use the weight of the fluid. A cup with linkage to close a valve. First pump would fill the cup the weight would shut the valve. Hole in cup to drain on to the water wheel. As the fluid runs out the cup raises valve opens for next pump. Would have to balanced pretty well.

Also thought about the old pressure cookers. Weight on the port to hold the pressure back until it over comes the weight. Pressure bleeds off the weight resettles and resets itself. Can be finely tuned by drilling holes in the weight.

As I understand it, a poppet valve is a spring valve? And a pressure cooker / weight valve is the same except uses gravity instead of a spring? I tried that, but as I say, the spring vrs the pressure just reached equilibrium and I got a steady stream of gas. There needs to be a threshold built in.

The on/off action is needed for the pump to work pretty much at all, so to reach a cup in the first place to turn off the valve, the valve is already going to have to be working to drive the pump. If that makes sense..

You have a threshold with this type of a valve. What you imply that you require is a hysteresis of this threshold so that the energy difference from the upper and lower thresholds provide the power for moving your wheel and opening your valve. Now as you've noticed the poppet and gravity valve do a very good job of limiting a maximum pressure your chamber can produce. This makes them a very good safety valve with an easy to calculate threshold that you should include in your design to prevent a runaway system from becoming dangerous.

See what a great advantage knowing the magnitude of the forces can be in doing a design.

As poppet relief valve was thinking more of the type used as safety relief on compressed gas systems. Like the ones on compressor tank, when they open they bleed off a lot of the pressure before closing. They usually allow a 5 to 20 % drop in pressure before resetting. They do make them adjustable if you can find one at the low pressure that you need you can adjusted it for just the right amount bleed off to max the pumping action.

This looks like it would work, but it would be pressure-dependent.(cycle time influenced by pressure).

How accurate is your heat source?

Barometric pressure may even be a factor.

Have you tried your relief valve design on a regulated compressed air supply?

There are some pneumatic valves that use low pressure to shift a sealed valve. These might be usable in your design. ARO, and Clippard come to mind. Have you looked at them? I will look into some of them and send you anything that looks good. How much pressure do you have to work with?

http://www.clippard.com/store/display_details.asp?sku=2050

http://www.ingersollrandproducts.com/_downloads/lit/0615-M_Valve.pdf

Both of these companies offer pilot open, spring close valves that can be used to replace your plunger valve. Both also offer pneumatic amplifier valves that would allow a small pressure to open the valve. Once the pressure is bled off, the spring would close the valve again, containing the pressure from the boiler until it again reaches sufficient pressure to cycle the valve again. Both companies have products that operate in the millions of cycles reliability range. If you want me to suggest a particular valve, send some pressure and volume figures. Good luck.

What about a float valve from an older toilet? Without the float ball to add leverage, the slight pressure should push the valve off of it's seat easily. Then re-seat as pressure drops. By working with the arm length, you should be able to vary the amount of pressure buildup.

There has to be a threshold, if the valve can ease open and shut an equilibrium forms between the pressure pushing out and the spring / gravity pushing in. It naturally settles to somewhere in between with a steady flow of gas.

It needs to snap open, snap shut.

I tried a thing with a rectangular flap axled halfway along it's length in a square pipe. The theory was that the back of the paddle was happiest in the middle of the tube (like a fin at the back of a surfboard), while the front was happiest pressed against the tube's side (like a fin at the front of a surfboard). These two opposing equilibriums would mean that the flap couldn't stay shut (the pressure at the back would open it) and couldn't stay open (the flow would catch the front and slap it shut). So it would have to somehow alternate between the two.

It actually worked, but did what I was hoping it wouldn't, which was flutter at about 10 hz, rather than once or twice per second which I need.

Still pretty cool tho.

Anyway...

http://customer.honeywell.com/Honeywell/UI/Pages/Catalog/ComponentCategory.aspx?Catalog=Buildings&Category=Electromechanical_32708&ChannelID=%7b02CD9124-96F3-4A04-8EA7-1777CCCAD5B3%7d