David Lions wrote:
..............
>Does anyone know of a straight-forward theory page on ECG circuits on
>the web?  I am doing a quick one-off ECG at the moment.  Using this
>circuit: http://www.wavetek.com/products/AppNoteArbs.html
>


Don't have any ECG sites to point you to, but I looked at the
wavetek site, and have a couple of comments:

1. First off, they are using a DC-coupled integrated amp with
[apparently] high-gain. This may not actually work very well because
the skin electrodes typically have contact potentials in the 0.5-0.7v
range. The ckt will immediately saturate unless the contact
potentials are very close and track together over time.

2. ECG signals are much larger than the EEG signals Lance was talking
about - low mV range vs low uV range, respectively. His problem is
much worse in terms of gain and noise rejection, but yours is worse in
terms of low-end frequency cutoff. He can probably use 1-2 hz low-end,
you will need somewhat < 1 hz.

3. A fairly straight-forward AC-coupled instrumentation amp will
generally work pretty well for ECG measurements. Unlike wavetek's
ckt, you can generally use a series-C, R-to-gnd differentiator
ckt on the front-end of the amp, with RC time-constant of 10 sec
or longer.

4. You will note that, in wavetek's Fig 2, this waveform probably
represents an amplifier with low-end cutoff >1 hz or so, rather
than DC-coupled, like in Fig 1. A real ECG signal typically has a
DC "plateau" in the S-T segment. To see it, you have to have a
very low low-end cutoff.

5. You can generally knock together a basic ECG amp in a short
while, compared with Lance who will have to fiddle around much
more.

6. NOTE - You do have to be careful when connecting anything to
the human body that you do not inadvertantly electrocute the
person being monitored. Commercial ECG/EEG/EMG systems all have
means to electrically isolate the high-level electronics
- optoisolators, transformer-coupled power, etc.

hope this helps,
- Dan Michaels
Oricom TEchnologies
http://www.sni.net/~oricom
==========================

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