Home Optical Spectrometry: The B&W Tek BTC100-2S | Hacker News Reader

Home Optical Spectrometry: The B&W Tek BTC100-2S

sevarg.net

45 pointswalterbell8 comments

It's so cool to see this on HN. I think this is one of the B&W Tek units that came from the NuSkin device which was err... of debatable scientific merit. It appeared on Dr Oz at one point:

https://www.youtube.com/watch?v=mfizniDX3pQ

Loads of these appeared on eBay as well as the complete Nu Skin units with identifying logos removed they used to go for about $15 which was a bargain. I think I picked up 5 or 6. Wrote up my note here:

https://everydaything.substack.com/p/b-and-w-optical-spectro...

Have been playing with one until somehow zapped it by either ESD or common mode voltage from a cheap AC adapter and inadvertently grounding some part of the PCB. Mine is now a stone that will never finish initialization (no LED color change). Haven't got time to look into if it's repairable, but there is so much programmable logic, so not very optimistic.

If you want to get one, I'd recommend protecting it in a shielded box and use a proper power supply.

Theoretically it's not too difficult to build an interface to the linear CCD used in these devices. I put together an interface here:

https://41j.com/blog/2020/01/notes-on-using-a-ilx511-linear-...

And put together gerbers here: https://41j.com/blog/2020/02/ilx511-ccd-interface-r2/

Plan was to eventually design a new board for these spectrometers. But obviously never got round to completing the project (and probably never will).

I was working on something like this for the toshiba tcd2304(?) units a few years ago. I have a working prototype in my closet using a psoc5.

Psoc5 is perfect for this kind of project because it integrates cpld-like functionality with a microcontroller that can also perform serial over USB. You you can get jitter free logic signals to the linear ccd.

The raman spectrometer functionality of these units could be really useful for certain projects. I have half a dozen of them in my basement, and I hacked them a few years ago just for the spectrometer portion. But the filters and fiber laser are nice too. But like you, I never really took it any farther, mostly for lack of a good project to use it on.

I also have the filter and laser part, but the performance of the filter and beam splitter probe is too poor for it to be useful.

At least mine have strong self-fluorescence, making it hard to get a meaningful data as the laser intensity are not that stable. Also the short excitation wavelength of 423 nm led to significant fluorescence in many samples I tried.

Reading out those CCDs needs some signal conditioning and ADC, the time needed to build and debug them by myself worth more than a second hand Ocean Optics or Verity, so I wouldn't bother.

Least squares fitting of template spectrums is, umm, pretty simple for rough and ready results.

Given the 2048 channel spectrum emitted per source sample, someone might want to scan across a line or grid scan an area and really sharpen up the combined spectral data from 10's of thousands of samples by using a technique from gamma ray environment mapping down shifted to the visible spectrum.

https://sci-hub.ru/10.1071/EG02172

https://sci-hub.ru/10.1071/EG998516

Little do people know that these are being pulled from a fully functioning raman spectrometer that operates with a blue fiber pigtailed laser.

I believe these are coming from some medical equipment for detecting carotenoids in the skin.

Nice to see some gnuplot magic at the end of the rainbow

Reply on news.ycombinator.com