Control 3phase cd drive motor - no sensors

[Grossel]

Hi.

I've had some thaughts about this recently, but is unsure.

If I had three H-bridges, in theory I could easilly drive a 3phase stepper (I call it a stepper just for short). And if all H-bridges only could provide low and high voltage out, then I could run it and speed is totaly dependent on clock frequenzy.

But if I had a H-bridge that could have a off-state/high z/not connected output, then I could on each phase do this:
* When a phase normally would go from positive to negative, the driving H-bridge instead goes in off-state.
Question is: If I was able to measure this voltage on the non-connected stepper wire, what kind of waveform could I expect?
Also, when torque increase, I guess that this waveform would be compressed to some degree.

Goal here is to find out how I can make a cd-spinner motor run like a normal motor without using hall sencors.

Thanks

 

[ronv]

It's done all the time in disk drive motors, but you need a micro to get the motor started in the right direction to start with (until it generates enough back emf). Might google disk drive motor driver.

 

[Grossel]

Might google disk drive motor driver.

Heh , tried that thought
I fail finding any sources describing this. All sources I manage to find depend on hall sencors.

Want to made (or at least a working plan) a diy version, so no chips
Background knowledge is all.

Back emf is what you call it? Well, there should be ways to sense that.

 

[ronv]

Sorry, It's been a long time...... Here are some links that might give you some background. Here is my best recollection on how ours worked. We had a DSP and a dedicated chip similar to the one below.
The first problem is getting the motor started without knowing which phase it is sitting on. As I recall we pulsed each phase and measured the current. The one with the highest current was the one that had the most iron under the winding. From there we could start the motor with the dsp by assuming which coil to energize based on where it was. We could then "manually commutate" by forcing the proper sequence at a known speed for our motor. After about a revolution there was enough back emf in the undriven phases to sense the location. They make a nice sine wave as the magnets pass. During the startup the back EMF was used for velocity sensing, but once it got to about 90% the sectors written on the disk were used as the velocity info and the back EMF only for commutation.. Sorry I can't be of more help but maybe this will give you a place to start. It's a complex system but it eliminates the cost of the hall sensors and in disk drives cost is king.

Brushless DC electric motor - Wikipedia, the free encyclopedia
**broken link removed**

 

[Grossel]

They make a nice sine wave as the magnets pass[/url]

Hi. Does this means you have TWO phases where their driver (H-bridge) isn't connected to gnd nor Vcc?