Converting Signal from Polar Heart Rate Monitor

With such low frequencies, I'd be looking at timing the time between pulses, rather than using an analogue frequency-to-voltage converter. Not sure how accurately you can do that under Windows. When I last did any Windows programming, you could only be sure of about 50mS resolution. It'd be easy with a micro-controller, though, if you have the facilities. ... Johnny

Let an MCU do the job, or install some unix.

With an MCU you can just time the square wave and send the data over builtin uart. I think even a 3 USD mcu can fix this.

schreef in bericht news: snipped-for-privacy@b28g2000cwb.googlegroups.com...

Matt,

There's no use in converting a digital signal to analog only to convert it to some other taste of digital again. I see two solutions that might help you:

1 - Use some micro with a build-in UART and a MAX232 like 5V(logic) to RS232 level converter. The X-tal clock of the micro should be accurate enough for your application but that can hardle be a problem. You can make the micro count clockpulses between two rising edges of your heartbeat monitor, convert the result to an appropriate ASCII and send it to the serial port. No need to say the bulk of the work wil be programming the micro. 2 - Convert your 3V square wave to RS232 level and connect it to an input of the serial port. No need to use RxD. DSR or CTS can also be used. Now write an interupt routine that acts on a change of the input used. Increment a counter and check the time. After 60s you process the counter contents and reset it. You can vary in different ways. Use two interrupt routines, one for the clock and one for the heartbeat. Use more counters to provide an update on every heartbeat and so on. This solution places the burden on the PC-programmer.

petrus bitbyter

The frequency of the pulse rate monitor is INDEPENDEDNT of the pulse rate.

It is the envelope or the PULSE REPITITION RATE that you need to measure.

Tutorial at

--
Many thanks,

Don Lancaster                          voice phone: (928)428-4073
Synergetics   3860 West First Street   Box 809 Thatcher, AZ 85552
rss: http://www.tinaja.com/whtnu.xml   email: don@tinaja.com

Please visit my GURU\'s LAIR web site at http://www.tinaja.com

Not

Hi Don, I only made my suggestion based on Matt's info of a 3V square-wave at ~ 1Hz, implying that the pulses were synchronised to the heart-rate. And that: "The frequency of the signal varies according to the users heart rate." Incidentally, isn't the "pulse repetion rate" equal to the frequency?

After having a look at your web-page, things are a little clearer. A couple of questions, though, if you don't mind:-

1./ Do the "six micropower pulses" at (B) in your diagram represent 6 pulse events, (heartbeats)?

1. What does the output directly from the "conductive pickup pads", before signal conditioning, look like?

(I've had a bit of a look on the web, but couldn't find exactly what I was looking for.)

... Johnny

Note most 232 receivers do not need the signal to go beneath ground. In fact, they are designed to be somewhat TTL compatible. I never bother with 232 converters for simple hacking purposes. Note that the 232 converters with charge pumps put jitter on the pulse. Not significant in this application, but it can be around half a microsecond, maybe more.

Also, as windows progressed, the interupt timing got much more poorly controlled. You really need DOS to get good timing on the handshake line. Again, probably not a problem in this application. There is a commonly used program called slicer.exe that is used for detecting the presense of signals on the handshake lines.

It is a good diagnostic to see that the line is wiggling.

amplifiy it (until it meets RS-232 spec) and feed it to the carrier-detect pin. the carrier-detect pin can trigger an interrupt/signal/event so the rest can easily be done in software.

you may need isolation in there too, an opto-isolator can easily be driven from a 3v signal (just pick the apropriate resistor)

--

Bye.
   Jasen

millisecond timing is easy in windows, microsecond timing is trivial in linux. 10ms precision would be overkill for an application like this.

--

Bye.
   Jasen