> You can buy UV photodiodes from hamamatsu but they are probably expensive.
> Is the wavelength short enough to erase an eprom ? Program it and see how long it takes to erase -
> you ought to be able to automate this cycle.
> 
> I believe UVB changes the ionisation voltage of neon, so maybe the frequency of a NE-2 relaxation
> oscillator may depend on UVB exposure, assuming enough  would get through the glass.
> 
> Photoelectric effect ? detect photocurrent with a very sensitove op-amp?

Since this is a one-off, even a fairly expensive photodiode wouldn't be
too bad a thing.  A previous respondent indicates that the erase/write
cycle ages an EPROM, and that measuring that time might not work
due to changes over time.

These lights don't emit SOLELY UV-B, but the emission of UV-B is
the critical function in their replacement.  They continue to emit visible
light for a long time after they stop emitting enough UV-B.

I'm thinking that the UV-B present would be enough to erase an EPROM,
if slowly.  I believe these lamps are intended to pretty precisely mimic
natural sunlight, so I'm not expecting complete erasure, just a steadily
increasing number of erroneous bits.  When a certain time passes with
the error rate being below a certain threshold, assume that the UV-B
emission is too low and the bulb must be erased.

I like the idea of the NE-2 oscillator.  I'll look into that.  Thanks!

Mike H.
_______________________________________________
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist