On Wed, May 20, 2015 at 1:57 PM, Jerry Biehler <[email protected]> wrote:
> Would anyone be interested in helping me build a controller for a 
> spectrometer? I have a Spex 1681B, it is pretty simple, step/direction pulses 
> control the scan with limit switches and then an analog signal from the 
> photomultiplier tube. Hardware I can do, software, not so much. Figure use 
> one of their edisons.
>
> http://www.horiba.com/fileadmin/uploads/Scientific/Documents/OSD/1681340E340S.pdf
>
> -Jerry
>

I'd be up for helping you out, but I can't this weekend... too much
going on. What did you have in mind other than telling your hardware
to scan the monochromator and (presumably) grab ADC readings?

I had been working on this a while ago, for the CCD array spectrometer
I was working on... but I didn't have much direction other than for
data collection and doing pH experiments (I was in a chemistry class
at the time I started writing the software):
https://github.com/nmz787/open-spectrometer/tree/master/desktop-software

Plotting the data is achievable with many different solutions, and
since writing the above code, I feel matplotlib isn't a great choice.
It looks OK, but not amazing, and I've heard the library isn't
particularly well written, and I know it's not great to try embedding
into other GUIs.

There's also Public Lab's Spectral Workbench, but I don't really know
what it can and can't do... I feel like it's pretty tied into their
spectrometer, and I don't know if they have things like beer-lambert
calculations built in.


What really interests me is some kind of machine learning on spectral
data, to match a given spectrum to a library of data. This is where
the big money spectroscopy companies shine, and what really
differentiates them compared to other hardware producers.


Anyway, I want to get together with you to talk
SEM/High-Voltage--low-noise stuff sometime anyway. I've been thinking
of a way to do EDX too, but using a PIN photodiode instead of a SiLi
detector. In theory it should work, but with less resolution, and I'm
guessing probably less signal-to-noise (so would require long scan
times, another guess). I was planning on using an LPC4370 with it's
80MHz ADC on it... but not sure that will be enough bits, and it's a
good amount of work to get up and running (I only started about a year
ago and then got busy and distracted with how hard it was to really
get some steam up with it).

I'm out in Hillsboro most of the time, if you're ever out this way,
drop me a line.
_______________________________________________
dorkbotpdx-blabber mailing list
[email protected]
http://music.columbia.edu/mailman/listinfo/dorkbotpdx-blabber

Reply via email to