Flutterby™! : Physiology geekery 2006-08-23 18:39:31.343872+00

Physiology geekery

2006-08-23 18:39:31.343872+00 by Dan Lyke 8 comments

Anyone out there have experience with heart rate monitors?

Yeah, I know, what I really want to do is build a couple of ZigBee devices, probably using the Freescale MC13191 transceiver, one for wheel revolution pickup, one for pedal cadence (although I could probably coalesce these into one device on my rear triangle that, later, also picked up crank force, although I'm not sure how to do that yet; piezo pickups in the cranks seem like a lot of machine shop cost, and looking for vibration in the chain seems like quite an engineering task), and once I got that far something that'd transmit physiological information to a Linux box on my handlebars seems like it'd be trivial, but...

Sometimes you've gotta look at the time available and say "ya know, I'd really rather be riding my bicycle rather than diddling with hardware I'm not getting paid for."

So what I want is something that tells me, when I bonk on climbs, whether I'm up against blood flow or nutrition.

I don't know much about HRMs in sport, but I do remember being surprised when I got a HRM watch for my dad. His rate jumped all over the map as he stood, sat or even moved. I realized that the difference was in the way the system tracked heartrate. When I worked in nursing home I took pulses all the time, but counting beats for 30 seconds and multiplying by 2 gave us an average heartbeat, while monitors track the interval between beats and extopolate it out to beats per minutes. I discovered that the actual rate changes slightly all the time. i suspect that for different uses different sampling techniques may be needed.

#Comment Re: Ooh bike geekiness made: 2006-08-23 19:53:16.739899+00 by: soopahviv

I'm all for the geekiness of building your own, but in the interest of maximixing riding time, perhaps an off the shelf device is the way to go? I got the Polar s725x a little over a month ago, and I love the hr/speed/cadence/altitude graphs it gives me after every ride. I think they have an extra thing you can buy for power, too.

Dunno how an hrm would help with determining blood flow vs. nutrition as the cause of a bonk. Maybe add some sort of glucose meter to the setup??

My thinking on HRM for blood flow vs nutrition on bonking would be that if my heart rate is reasonable and I've bonked, it's got to be nutrition. Crude, but less crude than I've got right now, and given that I'm as strong as most of the folks I ride with in the sprints, but get dropped on the climbs, I'm looking for solutions.

I'm sure I'll end up with a glucose meter and skin full of holes before too long, though...

It's nice to have someone with an actual recommendation. It looks like there's minimal Linux support for the S725, which may or may not apply to the S725x, but <whine>I suppose I could use the Windows machine for that</whine>. It sounds like the chest strap is a "send it back to Polar when the batteries fail" deal, but that's just the way it is.

Sigh. Now I need to justify $500 for a proprietary bike computer (computer + cadence + interface dongle), although the up-side is that they've got the power measurement too.

Oh: Petronius, thanks for that insight. I was thinking on yesterday morning's ride about an optical sensor on the bars that I could stick a thumb into to get away from the whole EKG chest strap, but averaging the sample over a fairly large time appears to be a necessity.

Note: Pulse oximetry via a non-invasive finger sensor's been around a long time. You get pulse rate, plus o2 levels, a really usefull thing. If your heart rate and o2 levels are decent, it's fuel. o2 levels down.. breathing issues? It's been a long time since I played with that world.

Well that's an easy circuit to build, it's just a matter of getting a finger in the right place.

Hmmm... I need to start by bundling up and selling a lot of stuff around my workbench.

Dan, while you CAN build the circuit, I can probably scrounge a pulse-ox finger probe. The decent ones use several specific wavelengths of light to shine through your finger, reacting to oxygenation levels in the blood. I was alos amazed how cheap and plentiful they are on the open market: like this one, which uses an 8 second average for heart rate...

Aaaand, the specific wavelengths are 650nm and 805nm, fairly easy LEDs to come up with, the harder challenge appears to be figuring out the sensitivity of the phototransistors.

(Yes, I know, fun intellectual exercise, but too damned many projects to not buy stuff off-the-shelf.)