You are right about using a differential, followed by non-inverting stage
for the job with this op-amp ...

The solution to my problem as it happens is to do with gain as much as the
maximum voltage swing I could expect from an LM358.  Somebody has mentioned
that if my negative is at 0, then I could expect clipping above 3.5V.  This
is indeed what I saw, clipping, and is what gave me the clue, and after
further simulation, have concluded that the gain is pretty much limited to a
0-3.5V swing for an input range of 0-400mV (can't be bothered working that
out).  I could settle for that, or as I have now tried, use both op-amps in
the package to get a swing from 0-4V which is what I was after.

Thank you for all the advise.

PS: Must get my hands on some better op-amps ?  Any suggestions as to a
suitable one for this job ?

Ian.



-----Original Message-----
From: pic microcontroller discussion list
[mailto:PICLIST@MITVMA.MIT.EDU]On Behalf Of Paul Hutchinson
Sent: Friday, 21 February 2003 1:51 pm
To: PICLIST@MITVMA.MIT.EDU
Subject: Re: [EE]: Differential Amplifiers


Yes, there are limits to how much you can amplify a signal with an op-amp,
even in the theoretical ideal op-amp. In any real world op-amp there are so
many limiting factors that without knowing the exact circuit topology and,
input signal you're using, it is nearly impossible to figure out which
limitation(s) you're hitting.

But since this is the PIClist I'll take a guess :-).

A limit I've seen people trip over often, is the magnitude of the power
supplies. If your input times your gain exceeds the power supply then the
output will be clipped to the supply level. In practice no real op-amp can
drive it's output all the way to its supply rails. So, you'll see clipping
at some value less than the PS voltage.

By the standards of 1974 when it was introduced, the LM358 was a "HiGain"
amplifier. By today's standards it's a very inexpensive commodity amplifier.
I only use it and, its quad brother the LM324, for signals from DC to ~100Hz
and with gains of < 20. If you really want 100x gain on this series of amps
I'd go with a 10x gain differential stage followed by a 10x non-inverting
stage.

If you're interested in learning a lot about op-amps I highly recommend
Walter G. Jung's excellent book "IC Op-Amp Cookbook". If you just want to
get a specific circuit working let the list know more details and I'm sure
you'll get lots of favorite op-amp recommendations ;-).

Paul

> -----Original Message-----
> From: pic microcontroller discussion list
> [mailto:PICLIST@MITVMA.MIT.EDU]On Behalf Of Ian McLean
> Sent: Thursday, February 20, 2003 8:21 PM
<snip>
>Do op-amps have a limit on the amount they can amplify a signal ?  What I
am
>finding is that if you increase the difference in resistance in the divider
>in a differential amplification setup, at a certain point, the signal
>becomes non-linear at the peaks.  Increasing the reference voltage into the
>op-amp does not seem to help.  At a certain point in the divider value, the
>output simply shifts further up from zero, rather than increasing the gain,
>and the peaks are flattened out.  Note that this is not the same as getting
>the low point to zero on a non rail-to-rail op-amp by providing a negative
>reference - I am already doing this.  I am having a great deal of trouble
>getting much more than about 20x gain from an LM358.  I would like 100x
>gain, but this seems impossible to achieve with this op-amp, even though it
>is marketted as a "HiGain" op-amp.
>
>I am pretty sure of my observations.  I am not proto testing it, but SPICE
>testing it with CircuitMaker, so I can see a sine wave amplified output
>compared to a sine wave input - a good way to graphically see with a SPICE
>transient analysis what is going on.
>
>PS: I am good with digital electronics, but I wish I was better at analog
>...
>
>Rgs
>Ian.

--
http://www.piclist.com hint: To leave the PICList
mailto:piclist-unsubscribe-request@mitvma.mit.edu

--
http://www.piclist.com hint: To leave the PICList
mailto:piclist-unsubscribe-request@mitvma.mit.edu