..talking to myself..
I requested a sample price of the 641nm diode and the Mitsu sales rep emailed me back
Good afternoon Steve Please supply the following information and we will be happy to supply you a sample free of charge. - Program name - Estimated annual usage when in production - Target date for prototype - Target date for production - Application Sincerely,
That capacitor idea is seriously nice. I use the 2.5V types directly across diodes, but this idea is very good. It places a lot of demand on the switch, but a changeover reed relay might do it if you limit input current surges. Not too sure on the lifetime though. Probably ok with a large value cap because it won't have to switch faster than a few Hz. (Better that than too slow, you want to reduce transients when switching caps).
Also, a resistor between the switched caps and the last one before the LM317. Those caps have a high ESR so you'll see transients on the LM317 input when you switch in a full cap. A resistor in series before the last cap will slow that down so the LM317 should smooth it out ok.
Check with a scope. And double check. It might not work as you hope.
Good idea. Switching high current DC will arc and carbonize the relay contacts. This is draft 1. Thanks for the input. All ideas are welcome. That's why I put it out here. The caps I've found (Newark electronics)
claim to have very low ESR. I'll order the parts, dust off the 'scope and experiment with a cheap LD 1st
<edit> the resistor in the LM317 is in the wrong place. I'll post Rev2 soon. I may even go to solid state instead of the relay, but I like the relay (will be noisy tho) because of the zero volt drop vs 0.7V or worse in semiconductors. I stay away from reed relays usually except for very low current applications. I'll probably go with an industrial duty DPDT rated at >10 amps or so.
The rate wont be a few Hz, It should be switching only once every 5 seconds (0.2 Hz); increasing relay life.
Yes, nothing works like you think it will in the electronic design world..
Anyone who designs and builds something and it works flawlessly the 1st time is either extremely lucky or a super-genious engineer
There's nothing stopping you from using a PBS to do that, but a really good dichro would accomplish the same thing. (Don't know how expensive a "reflect 660, pass 635" dichro would cost, but I'm thinking it would be about on par with what new PBS cubes cost.) It would need to be a custom dichro though; not sure who you'd need to contact for that. (Doubt that Edmund's would have it.)