Post by thread:: [OT] disk drive spindle motors.

My name is Bill, and I'm a dumpster diver. [Hi, bill!] I take apart old disk drives with my kids, mostly for the magnets. (Cool magnets in the head positioner.) Seems a shame to waste the spindle motors, though... Now, a spindle motor is a brushless multiphase (usually 3phase) motor, not unlike a DC fan, right? Except that fan motors have a simple little one-or-two transistor circuit to drive them, while the disk motor probably has at least one dedicated IC, a bunch of power transistors, and probably the direct involvement of a microprocessor, so that the speed can be aequately regulated... If I don't CARE about speed accuracy, is there a simple circuit I can use to drive a spindle motor? Thanks Bill W

William Chops Westfield wrote: > I take apart old disk drives with my kids, mostly for the magnets. (Cool > magnets in the head positioner.) Seems a shame to waste the spindle motors, > though... Now, a spindle motor is a brushless multiphase (usually 3phase) > motor, not unlike a DC fan, right? Except that fan motors have a simple > little one-or-two transistor circuit to drive them, while the disk motor > probably has at least one dedicated IC, a bunch of power transistors, and > probably the direct involvement of a microprocessor, so that the speed can > be aequately regulated... If I don't CARE about speed accuracy, is there > a simple circuit I can use to drive a spindle motor? Sure, but the fan's are two phase drive, the disk motors are three. To get an idea of what waveforms you need you probably should look at the drive signals from a spinning drive. Look at CURRENT too (small sense resistors are usually found in the source leads of the drive FETs. Use a 12c508 to generate 3 square waves with 120 degree offsets for the 3 leads your motor likely has. You'll also have to ramp the frequency so that the motor will track the drive rate and not slip. The problem will be that to get any kind of torque out of the motor you'll need to buffer the outputs with push-pull drivers (all those power transistors are there for a reason. Spinning several hundred grams at 3600/5400/7200RPM takes POWER). If you want a non-PIC solution you asked the wrong list <G>.

Use a 12c508 to generate 3 square waves with 120 degree offsets for the 3 leads your motor likely has. You'll also have to ramp the frequency so that the motor will track the drive rate and not slip. That's what I thought. I was hoping for a magic analog transistor circuit that did all that automagically (as well as producing real sine waves instead of square waves, maybe.) At least some of the drives have an identifiable "multiphase motor driver" CHIP, but it seems it is driven by a microprocessor (you supply a pulse stream and it convertts to multiphase/bipolar/etc, but you still have to ramp up the pulse rate and such, I think.) The problem will be that to get any kind of torque out of the motor I don't THINK I care much about torque. That's another design point - don't need much torque, don't need accurate speed... If you want a non-PIC solution you asked the wrong list <G>. Maybe, maybe not! BillW