From: michal@mpp.edu.pl   
      
   On Thu, 8 Jan 2026 00:55:40 -0000 (UTC), Phil Hobbs wrote:   
      
   > Michal Wlodarczyk  wrote:   
   >> Hello,   
   >>   
   >> I am considering a hobby-level project to build a simple and low-cost   
   >> open-path two (or three) wavelength absorbance meter for atmospheric   
   >> measurements targeting nitrogen oxides (NOx), which exhibit   
   >> yellow/brown color, lying in the visible range.   
   >>   
   >> The basic idea is an active system with a transmitter/receiver unit on   
   >> one side and a retroreflector placed approximately 10–100 m away. I am   
   >> currently considering two wavelengths, one 405 nm and second not   
   >> critical but maybe 530 nm, to allow differential absorbance   
   >> measurements.   
   >>   
   >> For the receiver, I am thinking of using a PIN photodiode placed at the   
   >> focus of a relatively large lens to increase beam diameter and make   
   >> system less prone to air instability, het haze etc.   
   >>   
   >> The main uncertainty concerns the transmitter side. One idea is to use   
   >> high-power LEDs and couple the emitted light into the same optical axis   
   >> as the by reflecting it off plain glass plates (microscope slides?)   
   >> placed in front of the photodiode. LEDs are inexpensive and can provide   
   >> substantial optical power, so emission strength is not the primary   
   >> concern. Probably LEDs are far better than laser diodes in this setup -   
   >> high power, no speckles and paradoxically not being point source.   
   >>   
   >> However, I am worried about stray reflections, surface scratches, dust   
   >> on the glass plates and lens and general backscatter potentially   
   >> overwhelming the weak signal returning from the distant retroreflector.   
   >> A polarizer in front of the photodiode might help, but I am unsure   
   >> whether this would be sufficient in practice.   
   >> In principle, at these path lengths it might be possible to   
   >> discriminate the returning signal based on time-of-flight, but I am not   
   >> sure whether this is practically achievable with LEDs (or lasers) and a   
   >> simple photodiode-based receiver.   
   >>   
   >> I would appreciate any thoughts, references, or alternative optical   
   >> layouts.   
   >>   
   >> Best regards,   
   >> Michal Wlodarczyk   
   >>   
   > Fun project.   
   >   
   > You really have to do some calculations to see if you have any chance of   
   > making it  work, though.  I expect that a much shorter path that you can   
   > shield from stray light would be easier.   
   >   
   > If you go to the hitran database (search for that), you can get   
   > calculated absorption spectra as a function of species mix, background   
   > gas, temperature, and pressure.  It’s pretty slick.   
   >   
   > There are two main types of measurement for this: LEDs + fancy   
   > interference filters, where you find a spectral region with a   
   > concentration of absorption lines; or tunable diode lasers, where you   
   > pick one line and dither back and forth across it.   
   >   
   > Cheers   
   >   
   > Phil Hobbs   
      
   To be honest, the absorption theory itself is the most trivial and   
   probably least interesting part here. The gas I am after is yellow, has a   
   strong absorption maximum around 400 nm, and like any respectable gas has   
   absorption bands whose widths are matched rather to incascendent bulb than   
   single-mode laser. I even do not see a strong need for any filters at this   
   point.   
   Absorbance probably would be in range of 1-5%.   
      
   Sure, one could argue about the detailed structure of those absorption   
   bands   
   and do something more sophisticated, like scanning a tunable laser (as is   
   done for iodine molecules, for example). Maybe. But that is probably a   
   different project. Maybe better...   
      
   In principle, the same approach could also be used to measure ozone, which   
   absorbs red light.   
      
   Yes, at very low concentrations some things might start to break down, but   
   that is   
   part of what I would like to measure rather than assume.   
      
   What really interests me is whether absorbance can be measured over a very   
   long open path using a "single-ended" instrument, with only a   
   retroreflector on the far side, and in potentially rough air.   
      
   One can hope that if the light comes back along essentially the same path   
   as it left, then most disturbances affect both wavelengths in a similar   
   way. If the two channels are strictly coaxial and measured very quickly   
   one after the other, a relative (differential) absorbance measurement   
   might actually be quite accurate.   
      
   The output signal I have in mind is simply the difference in absorbance   
   between the two channels, possibly high-pass filtered with a time constant   
   of minutes, to emphasize moderately slow changes (for example some cars   
   passing through the beam).   
      
   What I cannot really find is any simple (or serious!) analysis of the case   
   where spectrophotometry is done with a retroreflector rather than a   
   separate receiver. Or in situation, when absorbing medium is always   
   present, which makes a true reference channel hard to make.   
      
   Cheers,   
   Michal Wlodarczyk   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)