Logyman,
The DC servo motor is a great way to get started, even if what you wind up with is not the most efficient. My first wind gen was a homebuilt I made back in 1998 or 1999, using an Ametek 50-volt motor (4" dia x 7" case length) and several quick experimental blade changes.
I first tried wooden blades, but I turned out to be no woodworker and only made tapered blades with no twist. It worked, though you had to fine-tune the "setting angle". Next I tried blades cut from PVC pipe, pioneered by Randy Young over on the AWEA discussion list. These used judicious layout/cutting procedures (that's a pun, since the inventor was a judge down in south Texas) to get the correct twist from the natural curvature of the pipe. These worked better, though I never did get a 3-blade hub made to my satisfaction. Hard to clamp the individual blades satisfactorily, and epoxy doesn't stick to PVC worth a flip, but it can be done because Randy's been flying servo motors with PVC blades for about 7 or 8 years. I don't know if one has survived that long, or if he's continually replacing components.
I think it was around 2000 that I bought my first AIR303. I borrowed the blades from that, made a new hub from a cast-aluminum pulley, and discovered that these blades worked better than anything I had yet built. Back then, you could buy "spare" AIR303 blades for $25 from Andy Kruse at SWWP, so that's what I did. Since then, "Hornet" blades from hydrogenappliances.com have hit the market, and they're a better choice because they're bigger and longer. Still, as another respondent pointed out, these and similar blades are really designed to run faster than the typical AMETEK motor needs. This finally brings me to the point of my posting - what actually DOES your motor need?
You should characterize your motor as a generator before starting your design.
First, you need a good way to measure RPM. Many folks over on the AWEA list use a cheap bicycle speedometer, which works electronically using a magnetic pickup. You can either calibrate it to read out directly, or you can just come up with a conversion table from MPH to RPM. I haven't tried this, but it really sounds like a good way to go. I borrowed a handheld optical tachometer from a friend at work. That worked well also, you just attach something to the shaft that the sensor will count as it flashes past.
You also need a way to spin the motor while you make your measurements. In my case, I had several of the cylindrical Ametek motors, so I just mounted two of them face-to-face by duct-taping them into the "V" of a piece of angle iron. Then I used several more wraps of duct tape to couple the shafts together. I had a variable-voltage power supply, and used this to energize one motor and drive the other.
Once you get set up, you need to measure output voltage as a function of RPM, for the motor being used as a generator, with no electrical load. Separate the output wires and don't hook anything to them (other than yuor voltmeter). The output voltage should be a linear function of RPM, so that you can come up with a "generator constant" in volts per hundred RPM or such.
Finally, you have to figure out the generator's internal resistance. One way to do this is by using a resistor across the output wires to draw current, while repeating the voltage vs RPM measurement. Choose a resistor that knocks the output voltage down by about one-third. Note - the generator will be noticeably harder to turn, and will slow down under load. That's why you have to measuse the RPM using a calibrated device, and increase the driving power to set the correct RPM. Also, be sure to use a resistor whose power rating is high enough that it won't overheat and change resistance during the measurement.
Here's an example based on one of my own motors:
- Suppose you measure the open-circuit voltage vs RPM, and come up with 4.2 volts per hundred RPM.
- Further suppose you attach a 3.5-ohm power resistor and repeat the test, this time getting 2.9 volts per hundred RPM.
- The internal resistance is then calculated as Rinternal = Rload * [(Vopen/Vload) - 1]
For this example, Rinternal = 3.5 * [(4.2/2.9)-1] = 1.569 ohms
WARNING - Don't use any type of heating element or light bulb as a power resistor for this measurement, as the resistance will change during your measurement!!!!!
You now know everything you need to design with your generator, ie it has an open-circuit voltage of 4.2 volts per hundred RPM, and an internal resistance of 1.569 ohms.
To make use of this information:
Suppose you want to charge a 12-volt battery with 5 amps of current. What RPM will that require? First, you need to know the actual battery voltage. I'd assume around 14 volts while charging. Also you need to know the voltage lost across your diode (always use a series blocking diode to prevent the battery from running the motor when there's not enough wind for charging). For this example, let's assume 0.65 volts (high for a Schottky but low for a P-N junction), but you'll need to actually measure it when you get a diode in your system.
Next you need to estimate the voltage lost across the generator's internal resistance, which is 5 amps x 1.569 ohms, or 7.845 volts. Total voltage needed "internal" to the generator is the sum of these three: 14 + 0.65 + 7.845 = 22.495 volts. 14.65 of this appears across the battery and diode, while the other 7.845 is lost across the internal generator resistance.
Anyway, in order to deliver a charging current of 5 amps, you'd need 22.495 internal volts, which is now easily calculated by dividing by your open-circuit generator constant. RPM = 22.495 / 4.2 = 5.36 hundred RPM = 536 RPM.
If you spin the generator at 536 RPM, it will put out 22.495 volts with no electrical load, and will put 5 amps into our theoretical battery/diode combination (if you maintain speed when it's loaded).
Now all you need to do is repeat these steps for the actual generator you have in hand!
