Re: SLA7062 Sync Mode (was New Article bla bla)

Jeff,

Great post! nicely put!

Thank You,

Ballendo

--- In CAD_CAM_EDM_DRO@yahoogroups.com, "jeffalanp" <xylotex@...>
wrote:

>
> Hi,
> Typically the 7062 runs each of the 2 phases asynchronously to

one

> another; each phase is "doing its own thing". The combinations of
> various PWM frequencies generates noise in the audible range, as

each

> phase is turning ON at different times. This noise (often

referred

> to as squeeling, grunting, hissing, spitting - and perhaps other
> uncouth sounds) is a combination of the two high pitched PWM
> frequencies to make a lower pitch sound.
>
> There is one place this will not happen: when only one phase is

ON

> (or alternatively when one phase is OFF). In this case (when one
> phase is OFF), there is only one PWM frequency (the one ON), and

it

> is generally above the audible range. At this point having the

PWM

> be "synchronous" doesn't matter since there is only one ON anyway.
>
> At all other step angles the frequencies can combine to make
> noise. There is however one place that you can reduce noise

(other

> than 0-100), and that is at 70.7. In this position, both phases

are

> trying to deliver the same amount of power, thus should be just

fine

> with the *same* PWM ON/OFF times. In this case, rather than

letting

> the circuit run the PWM so they are "doing their own thing", the
> start of the PWM signal will be synchronized for both. They both
> turn the switches ON at the same time (and should therfore be

turning

> them OFF at close to the same time as well). The 7062 PWM circuit

is

> a fixed-off time circuit, so after a "fixed-time" the circuit will
> try to turn the switch ON again. In synchronous mode, this is

only

> allowed to happen when BOTH are ready to be turned back on. At

70.7

> the time when both are ready to turn back ON will be very close to
> one another, so minimal (if any) power will be lost.
>
> The problem for steps other than 100/0 or 70.7/70.7 is that if

both

> are started synchronously (at the same time), one will reach its

turn

> OFF level sooner (less current to build up). The other phase

which

> has to build up more current will remain ON longer. Now, the one
> that turned OFF sooner would generally wait its fixed-off time,

and

> turn back ON, but if it now has to wait until the other phase is
> ready to turn ON (has reached *its* peek current and elapsed its

off-

> time), the first phase continues to loose power through decay as

it

> waits to turn back ON synchronously. This loss of power leads to

mis-

> position since it is no longer giving the correct amount of

current

> for that particular step angle.
>
> Thus, too much decay while waiting for the other phase to be ready

to

> turn ON synchronously --> loss of current --> mis-position.
>
> Jeff
>
>
>
> --- In CAD_CAM_EDM_DRO@yahoogroups.com, John Dammeyer <johnd@>
> wrote:
> >
> > Hi,
> >
> > I'm also a bit confused by the SLA7062 data sheet. All they say

is

> that
> > noise is reduced when the unit enters sync mode. It doesn't

really

> say what
> > sync mode does.
> >
> > The way I reduce current in the ELS driver which uses LMD18245

is by

> > shifting the reference voltage from 5 volts to something like
> 1.25V. This
> > causes the motor current sensing circuits to start chopping at
> 0.75A instead
> > of 3A (when configured for full drive current).
> >
> > The side effect of this is that the motor now becomes quite a

bit

> noisier
> > since the chopping happens differently from the full power

mode.

> The
> > LMD18245 is also a 'fixed off time' chopper.
> >
> > The only other way to reduce current would be to change the

current

> through
> > the individual windings to perhaps reduce one winding to 0

current

> and the
> > other to some low level that causes less noise. However, when I
> tried that,
> > (before using the 'reducing the reference voltage' approach), I
> found the
> > motor always moved so I gave that up.
> >
> > John Dammeyer
> >
> >
> > > Keeping track of steps
> > > is a logic
> > > function, and is (or at least should be) unaffected by the
> > > PWM operation.
> > >
> >
>