Hi,
I am looking into how I would go about getting/making a proportional valve. I have started with the basic design of a 1/2" ball valve with a servo, however the valves have on average an approx. breaking torque of 550 oz in.
The highest torque on an RC servo I have found is 430 oz in. and I do not want to even think about how much that one costs. the nice thing about using servos is I could tell the valve to be open/closed or whatever % in between fairly easily.
So, my question...... Does anyone have any ideas on how to make a cheap electronically actuated 1/2" proportional valve?
Or, Does anyone know where I could buy a cheap electronically actuated 1/2" proportional valve?
I guese one thing to note is that the 550 oz in. is breaking torque if the valve hasn't moved in a while so I might get away with manually breaking all the valves before startup and hoping that they don't stick again by the time the are next circled (or just have them cycle open/closed every couple of hours when not in use so they never stick in the first place)
anyway........ any thoughts?
You could use a stepper motor, instead of a servo. Geared, if necessary.
Even the servo could be geared, but the range of motion might not be enough.
What is flowing through the valve? That is, why are you trying to precisely control a valve? There may be alternatives...
Cooling water or wort, during brewing (home that is). The wort (unfermented beer) needs to be drawn off the grain at a certain rate, so the plan would be to have a proportional valve on the outlet of the transfer pump to throttle the draw off (controlled by the change in water level or flowrate). The valve would be a 90 degrees turn.
This is for a home brewing rig that is somewhat a pipe dream, fully automated (to a point), the thing is I need to put a cost estimate together to run pass my accountant (read: Wife) so that she can approve the spending 
One good point is my accountant (read: Wife) loves beer!
Any links to where I can by cheap stepper motors?
This looks pretty good: the 50:1 gearbox brings the torque up to about 375 oz-in, which is pretty close to what you need. If needed, you could do something like a 2:1 chain drive to double the torque again, and exceed your worst-case requirement.
It's an unusually low price for a geared stepper. Even with the shipping, it might be your best buy.
I'd look at getting an easier to operate valve. Instead of a ball valve, get one that operates more like a globe valve (has the round rubber washer inside). Note how easy the bathroom sink valves operate and get ~full flow with an ~180 deg turn. Go to youe local hardware/home improvement center and see what is available.
servos can work if you take the biggest servo an gear it down 1/1.8 or so (servos rearly take full 180 degree, so to turn it 90 deg. you can not have 1/2 gear ratio)
rotary encoder an a dc motor that gives you enough tourqe is the second option.
Industry or store bought:
http://www.valvestore.com/products.asp?dept=1682
This store has electronic valves that take 4-20mA or 0-10V signals, costing from $219
So you have to weigh the costs of diy and store bought
you can try to make pneumatic valve, this is the industry standard as they can close lightningly fast in an emergency
David
Does it REALLY need to be proportional?
Would it work to 'pulse width modulate' it?
That is, use an on/off arrangement, rather than a continuously variable one, and turn the pump on and off relatively fast?
Or maybe use a solenoid and plastic tube as a pinch valve?
Or hack out a "step" rotor for said pinch value that had 5-10 flow rates, use a stepper, and control the pump somewhat.
Huh, even better:
Mount a round disk with an eccentric center hole on the stepper shaft. Put it near a flexible plastic tube in a 'U'. As the disk rotates, it compresses the tube more or less. There you go, proportional valve.
If the ball valve is the only one you can use, you might try lubercating the ball with something like silicone plummer's grease to free up the movement.
Turning selenoids on/off creates shockwave back in the system, it puts a big strain on pipes and valves, they are not ment to be pwm 'ed. the liquid carries its momentum from full speed to full stop, this creates shock in the system and is not good for the pipes or valves.
gentle on/off is the thing if doing precise control.
David
You are thinking on off in fractions of a second. It may not need that. It may be okay to run for a couple of seconds, turn off, wait for a couple of seconds, turn on... It depends on over what period of averaging you can get the flow rate you need.
If the man is making beer (or ale or whiskey or whatever...brewing) he is thinking precise control, dumping wart is just a matter of a solenoid (on/off) but temperature control is a heat exchanger poblem and the temperature must not swing much for a good batch.
there are arduino micro-breweries
http://www.arduino.cc/cgi-bin/yabb2/YaBB.pl?num=1241337653
or the same at this link
But he doesn't use valves to control the flow, but pumps.
more specific: large peristaltic pump
its worth a peek there
David
If you use a geared motor to turn the valve, you will still need feedback to know when it gets to the position you want it. What you can do, in that case, is hack a servo; basically, you cut the connections to the motor in the servo and connect them to the inputs of a larger current handling h-bridge driver, then route the outputs of that h-bridge to a larger motor (like a windshield wiper motor, or some other large DC gear motor). You take the driver gears out from the motor to the potentiometer gear, but leave the potentiometer connected to the output shaft. You then need a linkage from shaft of the new motor to the output shaft of the hacked servo (a simple lever can do, or you can use gears, belts, chains, string, whatever). You can also remove the potentiometer from the servo, find out its value, and buy another similar (but larger) potentiometer and solder extended wires from it to the servo-driver board. You can also drop the servo entirely, hook the h-bridge to the Arduino (via a couple digital i/o pins), and the potentiometer output to the Arduino (via an analog input pin), then write code to turn the motor in the proper direction until the potentiometer reads the value you want.
Look up "homemade high power servo" - also look into forums detailing how to build radio-controlled PowerWheel cars; some take the route of a geared-down servo drive, others custom build their own servo systems like I described above. Also - check this site out:
Don't worry - it won't bite ya! It's an old site called "Omniscience Futureneering" - they have this project:
http://130.94.182.150/mowers.htm
That describes the servo mod I was talking about above - there are other sources on the internet as well, but they were one of the first, I think (well, maybe the Seattle Robotics Society got there first - who knows!); check out some of their other projects as well - there's some fairly bizaare and fun stuff on there (I think this site has been on the web since it started, or close to it)!
Another possibility to look into is homebrewing a peristaltic pump...
Finally, if you can find a food-grade DC-power pump somewhere, you might think about doing PWM with the pump, instead of opening and closing a valve (brtech noted this earlier)...
Hope this helps, and good luck with your brewing!
As to the lawn mower projects, I tried to duplicate the below H-bridge and did not have success (damaged a P channel MOSFET). As to large servos, bottom is one of my "to do" monster servo projects sitting in the project corner.
Thanks all!
Some good leads to follow in all this:
Paul and Ran - Steeper mottors look good and at ~$10-$20 might be alright to at least get one to prototype but still might want to have feedback from the valve to insure position (maybe).
ArduinoM - pneumatic valve would be sweet, but it would also require a compressor, etc so maybe not feasable for a small rig (if I ever progress past homebrewing maybe).
brtech - It is a flow control issue not just "so much volumn in so much time", to fast and the grainbed compacts and doesn't let any water through / too slow and you are sitting there all day. The tube pinch valve does sound like it might work but I think I'll keep with the ball valve for the time being.
cr0sh- great info for getting feedback for the valve position. I had a project to install a plasma cutter in the last place I work and they got their positional "home" by driving the motors against the hard end stops and detecting a high current on the motor, aka driving but not going anywhere. But that is probably not the best route for this project.
Thanks for all the input
If you don't mind the higher expense, you can get one of the various geared down servos like below.
still might want to have feedback from the valve to insure position
Un;ess your stepper is straining enough to miss steps, you could get away with just using an opto-interrupter to detect the "fully closed" position of the valve, and counting steps to estimate how far the valve has turned.
Disk drives with steppers to position their heads have done this for many years.
you could get away with just using an opto-interrupter to detect the "fully closed" position of the valve
Yeah, pretty much one way I was thinking and just setting the 'home', could also have one on each fully open/closed so it doesn't try to overdrive the valve. The opto-interupt would be a nice way to do it