Re: [CAD_CAM_EDM_DRO] 3 phase steppers (was: Motor current measurement)

Nicolas Benezan wrote:

>Hi Jon,
>
>yes, sounds reasonable. But what strategy is used to regulate the
>current? For a 2 phase drive it's easy because the windings have
>separate terminals. You can simply turn on the corresponding transistor
>and wait for the current to rise to the nominal level and then turn off.
>
>At a 3 phase drive I could do the same for 2 terminals but the current
>of the 3rd is always the sum of the 2 others. So depending on the angle
>of the current vector I have to consider which terminals are regulated
>and which is "floating" and have to switch the current sensor inputs
>accordingly. Sound a bit tricky to me. I thought there is some trick to
>simplify this.
>
>

Yes, but the key is that the currents are NOT independent. All you have
to do is regulate TWO
of the lines, the third must follow. You only need 2 current sensors.
One scheme is to put a
50% square wave on one line, and then apply varying pulse widths to the
other 2 lines. Take a
look at the ML4423 chip (oddly, no longer made, but the data sheets are
still on the successor
companies' web site years after the part was discontinued). Just do a
Google search for ML4423
and you should find it.

Jon