| T.R | Title | User | Personal
Name | Date | Lines |
|---|
| 2046.1 | you'll get a charge out of this one | DOGONE::WOODBURY | | Tue Aug 20 1991 11:02 | 15 |
| Typically, you will want to use an op-amp circuit with very large
gain, and electrodes (two [differential inputs to op-amp] on the
chest for the pulse and a reference on the foot if you want: that's
what we used to do in biology lab). From this, a voltage converter
to get to TTL and input to the rs232 or Centronics port for an
interrupt.
The major concern here is that you are wiring yourself to a machine
which is plugged into the wall socket with 120v-AC. It can hurt when
you do things like this. I wouldn't play around with it unless I
ran the analog section on batteries and used an opto-isolator before
the TTL section. I bet this is the reason you don't see these things
in all the catalogs....
good luck
|
| 2046.2 | | FSDB00::BRANAM | Steve Branam, DECcallserver Project | Tue Aug 20 1991 17:25 | 11 |
| For real isolation, there's the method CIC (???) uses in their heart rate
monitors: the sensor is on a chest strap and transmits a weak FM signal to the
receiver, which straps onto your wrist like a watch and has an LCD display.
It has a range of about 3 or 4 feet. Both parts are battery powered
and completely safe (also no annoying wires running from the sensor).
So now all you need is an FM receiver (and decoder?) to catch the signal and
turn it into a digital pulse to the PC...
These monitors are available in many bike mags and catalogs for $130-150.
SDB
|
| 2046.3 | | DRIFT::WOOD | Laughter is the best medicine | Tue Aug 20 1991 19:23 | 9 |
| CIC also makes a version of their heart watch which can upload the
heart rate samples to a PC. The watch records the heart rates at 5,
15, or 60 second intervals. After use, the watch is plugged into an
interface that is also connected to a PC via a serial port. The data
is uploaded where it is saved and graphed.
All you need is $$$
John
|
| 2046.4 | Polar Vantage XL HRM looks like this: | FMCSSE::ELLISON | Rick CXN2/35 DTN:523-2917 | Tue Aug 20 1991 23:23 | 73 |
|
The following is pretty far removed from the thrust of the discussion, but
it does describe the Heart Rate Monitor referred to in Note 2046.3 by
DRIFT::WOOD.
An elastic strap which goes around my chest. There is a hard plastic
section which has two "snaps" on it. This part is positioned over the
heart. I have not investigated closely, but suspect the hard material
is highly conductive and there are probably two halves - one for each
snap.
A transmitter which snaps to the chest strap. This contains a battery
which I replaced after a year - only because I had to replace the
battery in the watch. I supect there is some circuitry which senses
being connected to the snaps (perhaps some "leakage" between the two
hard plastic parts).
The watch which is the reciever and local storage. It seems that
there's about 3k bytes storage (2hours 40 minutes at 5 seconds per
sample; 8 hours at 15 seconds and 30 some hours at 1 minute).
Although, perhaps unrelated, the watch has a metal (stainless steel I
suspect) back. The watch will receive the signal from the transmitter
without being physically on my arm (I've wrapped it around the tube of
the frame of a stair climber with no problems).
The computer interface. This connects to a serial line (RS-232). Data
Terminal Ready is required from the host (or a convenient 9 V battery)
to enable the interface. The data is transmitted at 600 baud, 8 bits,
no parity, 1 stop bit.
The chest transmitter emits a barely audible click in sync with my
heart beat - the room must be quiet, and it took me several months to
become aware of it. My guess is that the transmitter is a very simple
electro-magnetic pulse generator without any of the refinement of a
frequency based "radio". Part of my reasoning is that another watch I
have has a auto restarting count-down timer which creates an audible
beep upon expiration of the timer. I was able to trick the HRM into
thinking my heart rate was 60 beats per minute by setting the timer to
a one second period. It'd be interesting to see what a spectrum
analyzer had to say about the matter.
The watch makes a connection to the computer interface in what appears
to be a similar "wireless" manner - since the cover of the interface is
plastic and there are no jacks or other direct electrical connections.
I've not detected any "clicking" from the watch when uploading, but my
Mac-II has a very noisy fan which may mask it. I've also not tried to
see what the watch's transmitting range is.
Software wise, there was not any Mac software available when I bought
mine; so I got the IBM PC version. Some experimentation with a PC
enabled me to deduce the encoding of the data from the watch. I wrote
a program on the Mac to translate the binary to something humanly
readable, but have yet to make headway on fancy graphing and so on. If
you're interested I'll be glad to share what I know about the encoding
of the data stream - it's very straight forward.
About a year ago someone and I played phone tag for a round or two but
never connected, and with work pressures, I dropped the ball. Perhaps
this reply will re-stimulate.
To come back to the orginal flow of building your own HRM, the opto
isolation idea can probably be easily implemented. Radio Shack had a book
with a lot of electonic basics with simple experiments to show how it
works. There was a section on electro-optical stuff. The title escapes
me. But book was 8.5" x 11", about 1/4 to 3/8" thick, and the font for the
print was "handwritten". Recently, I've not seen that book at my local
R-S, but several physically smaller ones by the same author with the same
handwriting.
regards,
Rick
|
| 2046.5 | Build HW to Get Data; Use SW for Crunching | GIAMEM::JERICKSON | John Erickson, DTN 232-2590 | Wed Aug 21 1991 14:26 | 71 |
| A very interesting discussion going on here!
I have a few comments, based on my experience using a CIC/Polar
wireless HR monitor, and [before that] considering building my
own wireless monitor.
* Some research I did into back issues of IEEE biomedical
engineering literature showed that one of the most vulnerable
elements in an EKG monitoring system is the integrity of the
_electrodes_. Having read numerous discussions about this
made me wary of both building my own and even purchasing a
commercially available system; having used CIC's system for
a year I have _tremendous_ respect for the simplicity and
elegance of their design. Those Finns are clever, fer sure!
* Seriously consider using a wireless system, both for
_safety_ (isolation) and for _convenience_. I have converse
with folks that have used wireful (as opposed to wireless)
monitor; their feeling is that these are incredibly
inconvenient.
The description of the "simulated" system is quite similar to
what I wear on my _wrist_ during my daily running and biking
training --- the point being, one can purchase a system which can
do the basic stuff described in .0. If you're gonna do it
yourself, perhaps consider the following:
(1) Purchase an inexpensive wireless HR monitor. Even
consider contacting CIC/Polar or a place that sell _and_
services these units, and ASK TO PURCHASE A REPLACEMENT
TRANSMITTER.
(2) Be clever and design a receiver. Find a local DEC lab
that has a spectrum analyzer, or try to get access to one
through local amateur radio operators. Also figure out a
circuit which will calculate the HR based on the number of
samples counted during some fixed amount of time.
(3) Design an efficient, CMOS-based circuit that stores
these HR sample values.
(4) Here's where it gets interesting: also design-in
circuits which sample _cadence_ and _speed_; have your
unit store them bundled with the HR sample.
(5) Create a simple interface with which to read the
information out at some later time. If you have a parallel
port on your PC, this is _really_ easy; if you want to go
RS232, include one of those CMOS, 68xx-family
microprocessors with the built-in serial lines in your unit.
The sample storage buffer can be part of the 68xx's memory
map, and PROM-based commands you design will allow
information to get sucked out under the control of the PC.
(6) Put in all in a nice little box which attaches to your
handlebars. Go out for a ride and fill up memory; when you
get home, plug in the serial line from your PC, switch modes
on the little box, and activate the "download" part of your
PC-based program. This puts it into a structure which your
analysis program can then play with, or alternately into a
data file which you can do _all_sorts_ of things with ---
like reading into a program like, say, MathCAD (From
MathSoft) for numerical analysis and interpretation.
In other words, I advocate off-loading the data accumulation task
to a remote hardware unit, and suggest using the "host" PC for
wicked-excellent data analysis tasks.
Later! And good luck!
John
|
| 2046.6 | Try electro_hobby note | WMOIS::MALLETTE_P | | Wed Aug 21 1991 14:55 | 7 |
| To base noter,
You might try the notesfile "CSOA1::ELECTRO_HOBBY" Its frequented by
the electro-inclined that may be able to help you out quickly.
Regs,
PM
|
| 2046.7 | more of my 2 cents worth | FMCSSE::ELLISON | Rick CXN2/35 DTN:523-2917 | Thu Aug 22 1991 00:14 | 97 |
| re: <<< Note 2046.5 by GIAMEM::JERICKSON "John Erickson, DTN 232-2590" >>>
<<<deletions/Rick>>>
* Some research I did into back issues of IEEE biomedical
engineering literature showed that one of the most vulnerable
elements in an EKG monitoring system is the integrity of the
_electrodes_. Having read numerous discussions about this
made me wary of both building my own and even purchasing a
commercially available system; having used CIC's system for
a year I have _tremendous_ respect for the simplicity and
elegance of their design. Those Finns are clever, fer sure!
RE>
RE> The place where I mail ordered my HRM was Creative Health Products (the
RE> US distributor for the Polar line and others). They are very helpful
RE> people. One thing they (or the HRM instruction booklet) point out is
RE> that some people have problems getting good contact from the normal
RE> elastic strap based electrodes. They say that the disposable EKG type
RE> electrodes can be used instead with a dab of conductive grease instead
RE> of the elastic strap. Those electrodes bear a resemblance to the clear
RE> suction cups used to hold the stuffed Garfield cat's paws to the inside
RE> of your car's back window.
* Seriously consider using a wireless system, both for
_safety_ (isolation) and for _convenience_. I have converse
with folks that have used wireful (as opposed to wireless)
monitor; their feeling is that these are incredibly
inconvenient.
RE>
RE> much better said than I did.
<<<deletions/Rick>>>
(1) Purchase an inexpensive wireless HR monitor. Even
consider contacting CIC/Polar or a place that sell _and_
services these units, and ASK TO PURCHASE A REPLACEMENT
TRANSMITTER.
RE>
RE> The CHP will send you a good catalog of their product line. You can
RE> call them at 1-800-742-4478. Their ad says "guaranteed lowest prices -
RE> 90 day price protection guarantee" and "all products sold with 30-day,
RE> money back satisfaction guarantee". Their range of HRM products is very
RE> broad. I haven't looked at the catalog recently, but I think I
RE> remember "spare parts" in there.
(2) Be clever and design a receiver. Find a local DEC lab
that has a spectrum analyzer, or try to get access to one
through local amateur radio operators. Also figure out a
circuit which will calculate the HR based on the number of
samples counted during some fixed amount of time.
RE>
RE> For some reason that I can't remember any longer, I came to the
RE> conclusion that my HRM must maintain some kind of a sliding window of
RE> the average heart rate of which it dutifully records the current value
RE> once every 5, 15, or 60 seconds. Assuming that to be true I then
RE> decided that it must not count the number of beats in the sample
RE> period, but rather count the time between each beat which it
RE> manipulates into the sliding average. That explaination made more
RE> sense to me than anything else I could think of (but then lots of
RE> things that make sense to me have no relationship to reality).
<<<deletions/Rick>>>
(4) Here's where it gets interesting: also design-in
circuits which sample _cadence_ and _speed_; have your
unit store them bundled with the HR sample.
RE>
RE> The HRM interface I have is compatible with a bike mount receiver
RE> (instead of the watch) which collects heart rate data and the normal
RE> bike cycloputer stuff like wheel revolutions and crank revolutions
RE> (from which can be calculated speed, distance, and cadence). I don't
RE> have one of these.
(5) Create a simple interface with which to read the
information out at some later time. If you have a parallel
port on your PC, this is _really_ easy; if you want to go
RS232, include one of those CMOS, 68xx-family
microprocessors with the built-in serial lines in your unit.
The sample storage buffer can be part of the 68xx's memory
map, and PROM-based commands you design will allow
information to get sucked out under the control of the PC.
RE>
RE> my HRM uploads its heart rates as an unsigned 8 bit byte indicating beats
RE> per second from 0 to 250 or so - the remaining 5 or so values are flags
RE> for events like "start of session", "interval split", "change to
RE> alternate set of rate limit alerts", "end of session", and "end of
RE> file". Each event is followed by some defined number of bytes
RE> containing encoded information like the current average heart rate at
RE> the time of the event, elapsed time since the start of session, or time
RE> of day when session started, and other stuff like that. The basic
RE> heart rate data are just one byte for every n-second sample interval.
<<<deletions/Rick>>>
RE> Let us know what you come up with.
RE> regard,
RE> Rick
|