Guy Sirton wrote:
> > > The ECG signal is usually amplified with an instrumentation
> > > amplifier.  Since most of the 60Hz signal (50Hz around here) is
> [snip]
> > The fact that it would never eliminate completely the 60Hz is from the
> > Analog Devices book:
> > "An ideal operational amplifier responds only to the difference voltage
> [snip]
>
> This is exactly why you should use an *instrumentation* amp and not
> just an op-amp.  Also known as a differential amp this is a special
> configuration that responds to differential voltage a lot better than
> just an op-amp.  Instrumentation amplifiers also usually exhibit a
> much higher input impedance.
>
> Guy

Yes, and the main reason for an "instrumentation amplifier", and we use
them here, like the LT1101 (cost $12.00 in single quantities, 10 or 100
x gain selectable), is that you can control the gain without interfere
with the inputs, since the inputs are isolated from the feedback net.
This is also why they have very high input impedances.  Some inst.amp.
are better than others, but they are always expensive.  Some cheap
solutions can be built using regular op.amps, but it needs a good unit,
with low VOS, temp drift and high CMR, always inside the operational
VCC(s) range.  It is possible to do it with two or three op.amps.  In
some not so critical applications, we use the LT1413 to build simulated
inst.amp, since it has a good low VOS, temp drift and high VCC.