Six of 'em, to be precise:
Four different types. I'll address them one by one, starting with the smallest.
Japan Servo Co. Type BH55FT8-02 / 8131A / AX050249B
24V, 2A, 8W
Only one of these motors. It's the only one that has two discrete SOICs, part numbers LB11923 7UKM (the 3-phase brushless motor driver IC) and 74HC4052A, which appears to be an analog mux/demux chip.
As I'll touch on later, these motors don't like getting taken apart; I can't get the rotors off without getting a bit destructive. However, these vias might be leading up to a few Hall effect sensors:
Note the beefy coil connector too--three of these spaced at ~120 degrees, as one would expect. I didn't get a shot of the underside of the board; it's got a lot of discrete components, and the white connector is labeled as follows: 24V, 24V, P.G., P.G., S.G., 5V, LD, S/S, F/R, BRK, CLK, H/L, and ENC.
Shinano Kenshi Co. DR-6236-174 / GX060015F
24V, 2A, 18W
Two of these fellers. The single driver IC on one is impossible to identify; it looks like the label has been etched off or otherwise obfuscated. The other one looks like this:
SK3002 747 T68 doesn't bring up any matches. Hell, how do you guys find datasheets online? I keep coming up with really sketchy looking sites.
Since I had two of these (and they're smaller than the other pair) I decided to putz around with one of 'em. I used a bit of 6-32 hardware to build a little stand so the motor could rotate freely on the bench. I built up a breakout cable using one of the original connectors from the copier (thank goodness I kept the wiring harness) and fed the motor 24VDC and 5VDC; when I pulled the START/STOP pin high, the motor started twitching. It seemed to twitch once per rotation and would do this maybe 10 or so times before shutting down.
I cobbled together an Arduino driver that allowed me to send a 5V 50% duty cycle square wave to CLK; when I did this and hit the START/STOP with a HIGH signal, the motor responded by spinning at [what looked like] maybe 500-1000 RPM or so. Sending 5VDC to the CW/CCW pin, as expected, changed the direction; I think HIGH = CW, but I could have it backwards. Either way, I didn't figure out what LD or BRK did, and I didn't test CLK above 1 kHz or so.
I also put a bit more effort into pulling the rotor housing off one of these units, starting by removing the label:
Looks like they assembled the motor and used a press to push a serious divot into the shaft, spreading it out and sealing the whole deal up. I don't think this was a malicious move to keep would-be hackers out; it's just a cheap and effective way to securely attach the rotor (which is presumably lined with magnets) to the shaft. Either way, I didn't want to drill it out so I tried the other side:
Kinda a funky looking snap ring. I don't have any fancy snap ring tools so I donned safety glasses and popped it off with a screwdriver; however, nothing seemed to come apart despite some pretty serious encouragement.
One other thing to note on these two motors:
So it's pretty tough to see, but the two lines between the rotor housing and the circuit board appear to be some kind of component. There are a number of identical ones all around the circuit board, too. Maybe a bunch of Hall effect sensors? Man, that would be sweet.
Japan Servo Co. Type BH60FT22-04 / 8218A / AX060310C
24v, 2A, 22W
Two of these guys too. Beefy, with a machined mounting bracket instead of the stamped piece on the previous pair:
Same driver chip as the smaller Japan Servo Co. unit, but no mux/demux visible (I suppose it could be hiding under the rotor):
Nidec 58M404A010 / AX06 0311B
24V, 3A, 40W
This one is pretty beefy.
Driver chip is on the back; it's a SOIC, and quite difficult to read. I couldn't get a good picture of it (well, I struggle with good pictures in general, but that's another matter...) but I did make out 62FV towards the end of the part number; any ideas? In any case, it's also got a heat sink mounted device which appears to be a bundle of FETs:
So that's where I am to date. I'd love to put together a decent driver library for these motors; something that's portable to various -Duino/etc platforms that you can feed direction, speed, acceleration, etc and know that the motor will do what it's told. @Angeliki Beyko suggested I swing by Twin Cities Maker for the Twin Cities Robotics Group meetup; I brought one of the 18W units to show and the kind folks there had a lot of great suggestions, including:
- use a polarized light source to get a better image of the ICs
- try to reassemble one of the copier's circuits, power it up, and 'scope the unknown signal channels leading to the motors
- rip one apart and figure out if it's got things like Hall effect sensors
- use the hardware but build my own driver circuit for the motors
- keep sending pulses to the motors to see how they respond
Anyone have any ideas? Who's played around with brushless copier motors?