Re: CA3140 (or other) op-amp with accurate wide-ranging output?

Jamie <jamie_ka1lpa_not_valid_after_ka1lpa_@xxxxxxxxxxx> wrote in
news:vjU2g.2381$3X4.2368@xxxxxxxx:

Lostgallifreyan wrote:

Hi, I have looked at CA3140 data sheets and seen that the output can
reach close to the positive rail on a split rail supply, and
(allegedly) right down to the negative rail.

This isn't as clear as it should be. First, on a single rail supply,
the offset cannot be trimmed to get a true zero output, you MUST have
some negative. At least, this is true on the two I just tested.
Second, all the specs I've seen assume a symmetrical split supply, or
a low single rail supply.

As I want as much positive overhead as I can get (to measure thirty
watts at 1W/V from a thermopile sensor), my idea is to use a split
rail supply whose total is 34 volts, with just a few hundred mV
(possibly derived by Schottky diode drop) for negative, leaving me a
full 30 volts on the positive side. I want a gain of about 11, so the
input values are within bounds for the full output, and I should now
get the full through-zero adjustment for input offset to appear
corretly on the output.

Will this work without frying the IC? And can anyone suggest some
common, low price and good performing IC's that can do this? Given
this concept it doesn't have to be capable of a single rail supply
with to-zero output, it just has to accept a VERY asymmetrical split
rail. :) I'm hoping the LF411 or LF412 (with external offset trim)
can do this.

you really should have some kind of dual supply to get a true 0
output..
what i did once was to use a simple LM324 to generate a negative
supply
voltage that operated the Vee/Vdd side of any opamp while i used the
already + rail supply for the Vcc.. etf..
with a diode i guess you could lift the common of the supply a bit
and use that as the common for the signal out.., and use the real
common for the rest of the components on board but make sure this
common does not connect with the internal common.! this will give you
about a -0.7 volt effect.
my self, i would just place a dual output DC-DC converter on
board.



I can get the dual supply, that's not the tough bit... I think that DC-DC
converter would be entirely over-egging the pudding. :) I can do the whole
thing with a dual op-amp and a few resistors, one half to make the split
rail out of single rail, the second to do the gain, with offset control
added. I've done this, it works great. (Common rail problem will not be a
problem, I've worked round that one already).

The only snag is I'm not sure I can do it with the mid-rail ground so
asymmetrically arranged as to get fully 30 V up from zero. I can try but
I'd rather not risk wasting op-amps when the question might be better
answered here first. What bothers me is that most op-amps might only be
specified for an output to around 15 volts give or take a couple, even when
like they CA3140, they're said to be able to operate from a >30V single
rail. Getting 30 within +/- 16 might not be at all the same thing as doing
it within +30/-2, and I want to be sure of what I can get away with.
.