jw@mck-csc.mckinsey.com (Jeffrey Weiss) (03/29/91)
In article <159639@felix.UUCP> dennisg@felix.UUCP (Dennis Griesser) types: >OK, so let's talk detector quality. Specifically noise immunity. > >Like half the people on the net, I have been experimenting with the IR >detector module made by Sharp, sold through Radio Shack. There are occasional >bursts of optical noise that fool the receiver module into believing that it >has seen an IR signal. I can deal with this. The incidence is low. > >The big problem is electrical noise on the power supply! ... >Any ideas, gang? I use the Sharp module also, in a remote control I build for my lights. Maybe it's just asking too much for these things to be as sensitive as they are (mine works at about 25 feet) and to be completely immune to spurrious detections. To avoid the problem, I use a Motorola chipset which encodes (and decodes) my signal into a complex pulse train. In fact, I think the encoder sends the signal twice, and the decoder must get the same signal twice in a row. Even under noisy situations, I NEVER get a misfire. Before finding the Motorola set, I looked at offerings from TI and Signetics, and their approaches were similar. I think simple signals (e.g., single pulses) are always going to be error-prone. PS: My sense is that the Sharp module is very similar in performance to what is typically found in consumer electronics (even looks the same as what's in my VCR). I have found difficulty, though, in finding IR LED's that equal the light output efficiency of typical remotes, though. The best I found is about 60-70% as good (distance-wise) as my Philips remote control. -- jw@mckinsey.com (...mit-eddie!mck-csc!jw) Jeffrey Weiss