From: Tony Williams
Subject: Re: current source for array of infrared leds
Date: Fri, 06 Dec 2002 13:14:30 +0000 (GMT)
References: <email@example.com> <4PMG9.firstname.lastname@example.org> <email@example.com> <firstname.lastname@example.org> <email@example.com> <firstname.lastname@example.org> <email@example.com>
NNTP-Posting-Date: Fri, 6 Dec 2002 13:15:24 +0000 (UTC)
User-Agent: Pluto/1.14i (RISC-OS/3.60)
In article <firstname.lastname@example.org>,
> > > Tony Williams wrote...
> > > > Wavelength drift can be pure hell if the IR is going
> > > > through water, because water has a series of deep
> > > > (steep-sided) attenuation notches.... smack bang in
> > > > the 850nm to 950nm region. Been there. :-(((((
> A quick look at the data shows pretty low absorption from 840-880nm.
> The main peaks appearing from 880-990 and in particular at 900 and
Yes thank you.... those numbers do jangle slightly, because
(afair) I had IR LEDs at (a nominal) 890 or 920... both of
which gave instability problems.
> In general the low absorption windows also correspond to the
> wavelengths used for optical fibre comms. Removing OH impurities was
> one of the key challenges in fabricating low loss fibres. In fact many
> good texts on fibre optics also include some data on OH absorption or
> at least provide a reference.
Thank you.... On the bookshelf, right in front of me.
"Optical Fiber Communications", by John M Senior,
Prentice Hall, ISBN 0-13-638248-7 or 0-13-638222-3.
Page 67, Fig 3.2, gives a graph of absorption due to
water (in a glass fibre) over the range 600-1400nm.
The graph I remember was much more jaggy though,
perhaps it is softened by the attenuation being a
log scale on Fig 3.2.
Either way, an IR LED, drifting its way up and down
those jaggies, did make life uncomfortable.
The turbidity meter first acquired a temperature
stabilising oven for the IR LED, but I later
sidestepped the problem by shifting to a tungsten
bulb, running at dull red, to give a wide band.