This page
Other pages
Account
Administration
I continue to work on the photometer project, this weekend I worked with my expert friend Steve Newsome a bit.
Some updates as I learn more about the problem:
- The standard for measuring luminosity, ISO 17514
requires that you illuminate the dial for a certain amount of time with a specified amount of precisely defined light.
According to this excellent page, those specs are
- D65 (daylight) lamp
- 200 lux for 30 minutes or
- 400 lux for 20 minutes
According to Steve, the D65 is most closely matched by a xenon lamp, which I dislike. I'd prefer to use LEDs, mounted in a collar around the to-be-pictured fixture.
From looking at this graph from Kronometric's writeup, I'd say that more than 1,000 lux is useless:
(Image credit: Kronometric)
He also lists excitation as 200-450nm, and there are definitely LEDs in that range. Most affordable seems to be either 365 or 375nm; the 200nm ones are tens to hundreds of dollars each. (And hazardous, too, you can easily fry an eyeball with deep UV.)
- Amazon has a 20mW 5mm for 2.48. Sold by Parts Express, who's product page is here
- 20mW
- 3.5 to 4VDC forward drop
- 30 deg viewing angle
- 20mA current, max 30mA
- Preferred site Sparkfun has some too but its hard to compare specs. Theirs (0.95 each)
- 395-400nm
- 3-3.6V drop
- 100-180 mcd
Designing a circuit
Right now, I'm thinking of most-simple: eight LEDs, with a simple switch to power them, nothing fancy at this point. I found this awesome LED design wizard which suggests this circuit:
That's conservative, power-wise, but should provide good lifetime.
I should probably note that my 5-LED UV flashlight will charge up a dial to crazy brightness in less than 30 seconds, so 8 higher-power LEDs should saturate the hell out of a dial.
I'm bummed to not have the ISO spec for scientific comparisons, but hey, maybe someday. I can, I think, measure to my own satisfaction how much UV illumination it takes to saturate and just standardize on that.
Progress!
