January 2005, Issue 174

Light-to-Frequency Conversion (Part 2)
Pulse and Oxygen Content

SERIAL DATA

The sampled value is the main attraction here. By properly shielding the TSL230R from extraneous light (the biggest culprit being 60-Hz light sources), a nice waveform can be produced from the sampled data. Pay attention to the old adage: Garbage in, garbage out. Bogus data will surely cause frustration no mater how good your algorithm.

A number of flags enable various data output. Although the sample values are valuable, they aren’t too helpful unless they’re displayed in a format you can easily recognize. I found that using HyperTerminal to grab the data and Excel to view the imported data as a graph was extremely helpful in planning my algorithm without having to write an application in, say, Visual Basic to grab and display samples either as text or graphs. I’m sure National Instrument’s LabView would’ve been a great tool to use here as well. By adding other values like MaxCNT and MinCNT to the output, I was able to see exactly how my algorithm was using the data and fine-tune it (see Figure 1).

Ultimately, the basis of this project comes down to displaying two variables, the beats per minute and oxygen saturation level. I’ve covered how a beep fires off each time a maximum peak is established. Text messages to display the beats per minute and O2 are chosen using PB4 (after reprogramming its function).