OK, so it might be a little excessive for a stir plate, but I wanted to share my little project with you guys and provide an overly detailed explanation.
- 80 mm computer fan
- Neodymium magnets from a hard drive
- 1-1/4" non-stainless steel fender washer
- 12 V, 1 A power adapter
- 2.1 mm ID, 5.5 mm OD barrel power connector
- Panel-mount fuse holder
- 1 A slow blow fuse
- 1 kOhm linear potentiometer with switch
- Knob for potentiometer
- 500 Ohm, 20-turn trim potentiometer
- TIP31A transistor
- TO-220 heatsink
- 10*uF, 25 V electrolytic capacitor
- 15 kOhm resistor
- Blue, high-intensity LED
- Panel-mount LED holder
- Perf board
- 10-24, 1-3/8" pan head machine screws (x4)
- 10-24 machine nuts (x12)
- #10 washers (x12)
My stir plate was basically constructed using parts that I had on hand or could grab quickly from Radio Shack. I started with a typical 80 mm computer fan that was spec'ed to run 500 mA at 12 VDC. I made the decision that I didn't want to have any "dead-time" from the point where I first started turning the knob to adjust speed, so I found the fan's "cut-on" point. This is simply to voltage at which the fan will start to turn and would be the starting point for my knob. More on that later. I then measured the fan's current consumption at a couple of different voltages and found that it consumed 80 mA at the cut-on voltage of 2.6 V, 175 mA at 5 V, and 470 mA at 12 V. Once I knew the details about what I was trying to control, I came up with a circuit.
The schematic is pretty simple if you're handy with electronics.
Power is supplied via a "brick" or "wall wart" style 12 V, 1 A transformer feeding into a barrel jack. I added a 1 A slow blow fuse (remember, the theme here is over-engineered) for protection in case something lets out its smoke in the middle of the night. Power is controlled by a switch built into my 1 kOhm potentiometer. The 1 kOhm potentiometer controls the base-to-emitter current into the TIP31A transistor and, therefore, the collector-to-emitter current that feeds everything to the right of the transistor including the fan motor and the nifty blue LED that is used to indicate the speed of the stir plate via the brightness of the LED (the relatively large resistor of 15 kOhm was chosen to give the high-intensity blue LED a little less blinding capability). The kickback diode represented by the dashed line was determined to be unnecessary for this motor as shutting off the fan doesn't cause any significant voltage spikes (I would suspect the same for most small fan motors), so it was omitted. The 500 Ohm potentiometer is used to set the minimum bias point of the TIP31A transistor to avoid the dead-time issue that I mentioned earlier.
Since my fan's cut-on level is around 2.6 V and 80 mA, it doesn't spin when I provide it with anything less than that. If I didn't include the 500 Ohm potentiometer, what would happen as I ramp the 1 kOhm potentiometer, and consequently the base current of the TIP31A transistor is that the motor would start to sweep from 0 V and 0 mA up. Until the motor saw 2.6 V and 80 mA, it woud sit still. Since I didn't want about a third of my knob rotation to be dedicated to this pre-cut-on region, I use the 500 Ohm potentiometer to set the minimum level of bias to the transistor. The potentiometer is a trim-style which means that it take about 20 turns to sweep from 0 Ohms to 500 Ohms. To tune the circuit to the fan's cut-on point, I click the circuit on by just barely turning the 1 kOhm potentiometer until the switch clicks. At this point, the 1 kOhm potentiometer provides the full 1 kOhm resistance. I then begin turning the 500 Ohm potentiometer until the fan just barely starts to twitch, then just far enough to make it start turning. At this point, I've provided the transistor with the proper level of bias so that my 1 kOhm potentiometer starts sweeping right at my cut-on point and continues on up to full power, and my circuit is fully tuned.
To tidy everything up, I built up my circuit on perf board and stuck it inside a simple enclosure with the 1 kOhm potentiometer, LED, fuse holder, and power jack all mounted through the sides. The fender washer was superglued to the fan to provide a magnetic barrier between the neodymium magnets and the magnets and coils inside the motor. The hard drive magnets were placed on top of the washer and their placement tweaked to provide as little vibration as possible while running the fan at full speed. The magnets were then glued to the washer. The fan was mounted on four screws that pass through the bottom of the enclosure. The height of the fan was set using nuts to get the magnets just below the top side of the enclosure.
Note that you're not done until your workbench is a mess.
So, now comes the big question...how well does it work as a stir plate?
Here it is stirring a 1 L flask of water at its lowest speed setting.
Here it is at its highest setting. I'm thinking that this speed isn't going to be all that useful for 1 L yeast starters, but the extra power might be appreciated with a larger stir bar in a 2 L, or larger, flask.
Finally, here's a video of me sweeping the stir plate speed.
I hope that this is at least interesting, or as a best-case, useful to someone on the board. Feel free to let me know if you have any questions or thoughts on the project.