Alice Campbell wrote:

> Thanks Chris,
> Yes, it (they) are bridge style, some have + swing, others both + and
> - swing.  I'm trying to follow your description with my feeble brain,
> but im not sure what a servo setup is.

By servo, I mean using an op-amp to achieve a DC value.  Seems like your everyda
y
application, but sometimes a control loop is not exactly stable, and you have to
 use
some RC tuning to 'slow' the op-amp down to the point where it centers your sign
al at
the proper DC level.  Refer to composite amplifiers (can't claim to have built t
hem,
but see applications from time to time).  In this case, you drive the bridge min
us leg
until one of the sensor return lines rests at ground.  Refer to Linear Tech's
application note 43, page 43-4.  The seventh application circuit is the one I am
refering to.  But, this circuit has an opamp that requires negative volts.  So,
instead
of having a negative supply well under the op-amp's desired output, why not gene
rate a
negative votlage generator circuit controlled by PWM.  Instead of a Flyback styl
e, you
could use a 'buck boost' or a 'cuk' converter.  Ref Linear Tech's AN19, page 13-
15.
This way, you could take the same op-amp in AN43, and make it do the PWM.  Just
RC
filter the feedback value, and the op-amp becomes very unstable.  It then oscila
tes
with the PWM frequency slewing back and forth to satisfy your 'servo setpoint'.

It's great stuff, good for days or even weeks of bench experimenting.

>  Because of the millivolt
> levels  out of the bridge, i still need some amplification to get the
> signal into the range the adc0831 can read.

If you do the ground centered bridge, you can do a straight non-inverting amplif
ier.
Nice and easy, one point adjust for the gain.  BUT, and it is a big one, if you
will be
going negative (bipolar), you must then setup the amplifier to drive your signal
centered about mid-rail.  Wait a minute.  If you have other circuitry and a proc
essor,
you must have a 5V rail.  The above discussion then refers to a +5V rail and a b
uck
boost that creates around -5V on the transducer.  But back to the amplifier.  Yo
u can
center the reference point on a non-inverting amp about the mid-rail.  Then posi
tive
and negative transducer readings will read above and below the mid scale reading
.

> yes,
> i will put in a better adc once i understand what im doing.  I rather
> suspect that the wobbly voltage will turn the readings to mush but
> its easy enough to put together on the breadboard and see.
>
> I like the idea of using the pic to generate the voltage for
> the bridge, and maybe the amp, yet still let the rest of the circuit
> use lower voltage.  i am already controlling power to the peripherals
> with the pic and i still have 2 free pins on port b.  anything to
> squeeze a bit more out of the batteries...
>

The sad part in all of this is that the efficiency of the DC/DC converters is no
t all
that great.  Especially for low currents, where most of your current is Quesient
current.  The Capacitor switcher circuit generating -5V may be a good candidate.
  I
think that they have pretty good efficiency.

>
> alice
>

Chris Eddy