B&W-tech Spectrometer & 473 module: Setup+Mods+Info

[CynicalBrad]

Re: B&Wtech 473nm unit / Spectrometer Mods & Info

I'm going to try to get some raw data ɗumps from mine later to try to see what is going on.

 

[Cyparagon]

Re: B&Wtech 473nm unit / Spectrometer Mods & Info

The cheapest we could get it is for the price of an unaligned version of their specrto, which isn't as good as these probably are. I looked at their specs for all their products and was not impressed.

What ARE you talking about, sir? It's the SAME THING.

From the science-surplus site: "We offer reconditioned compact fiber coupled CCD spectrometers. These spectrometers are crossed Czerny-Turner design, built by B&W Tek (Model BTC-110S). These are OEM instruments formerly used in a medical device."

If you want to get these working to a decent degree, there's hardware work to be done. "We install a new grating, re-align the optics, and place the OEM unit in a new custom enclosure as shown in the photo. We also install a basic order sorting filter to reduce stray light from higher diffraction orders."

Also, the (raw OEM) B&W Tek, the science-surplus, and the ocean optics usb2000+ ALL have a 2048 pixel detector. All other specs are down to the choice of diffraction grating, which is always a trade-off between range and resolution.

 

[Atomicrox]

Re: B&Wtech 473nm unit / Spectrometer Mods & Info

Good catch!

Here are the alignment procedures: http://www.science-surplus.com/prod.../AlignmentInstructions.pdf?attredirects=0&d=1
Shouldn't be too hard.

I'm pretty sure I can modify the provided software to allow insertion of the calibration parameters. In fact, if I'm not mistaken, they're hardcoded in the crappy code:

public static float CX2 = -3E-05f;
public static float CX1 = 0.1866f;
public static float CX0 = 390.47f;
public static float IX3 = -3.37E-09f;
public static float IX2 = 1.59E-05f;
public static float IX1 = -0.0226f;
public static float IX0 = 11.1f;

while (true)
{
int arg_183_0 = num2;
int num3 = num;
if (arg_183_0 > num3)
{
break;
}
int num4 = num2 - 2;
unchecked
{
float num5 = (float)((double)GLOBALS.CX2 * Math.Pow((double)num4, 2.0) + (double)(GLOBALS.CX1 * (float)num4) + (double)GLOBALS.CX0);
float num6 = (float)Conversion.Val(GLOBALS.VALARRAY[num2]);
float num7 = num6 - (float)GLOBALS.BACKGROUND[num4];
float num8 = (float)((double)num7 * ((double)GLOBALS.IX3 * Math.Pow((double)num4, 3.0) + (double)GLOBALS.IX2 * Math.Pow((double)num4, 2.0) + (double)(GLOBALS.IX1 * (float)num4) + (double)GLOBALS.IX0));
flag = (GLOBALS.BACKGROUND_COUNT > 0 & this.LevelCorrectCheckBox.Checked);
if (flag)
{
GLOBALS.SPECTRUM_PLOTLIST.Add((double)num5, (double)num8);
GLOBALS.TOTALCOUNTS += num8;
}
else
{
GLOBALS.SPECTRUM_PLOTLIST.Add((double)num5, (double)num6);
GLOBALS.TOTALCOUNTS += num6;
}
}
num2++;
}

Now I only need one of these units to test

 

[paul1598419]

Re: B&Wtech 473nm unit / Spectrometer Mods & Info

Yes, in my hunt for the power supply for these I did skim the Science Surplus version. I didn't look into it enough to see they were using the same optical bench. My mistake. Still, I think their software isn't going to be as good as BWSpec, which is the OEM for these spectrometers.

 

[Mosc007]

Re: B&Wtech 473nm unit / Spectrometer Mods & Info

Interesting, Mosc007. I found the integration time on my software is so slow it can't lock and hold a spectrum long enough to even read it. I see your base line is high on the intensity axis when nothing is being read. Your software seems to have some advantages mine does not. I would like to get together and see if there is a way to make this software completely usable. Unfortunately, I have my day and tomorrow locked down for most of the day.

The software I was using is the one that was posted in this thread.

 

[Mosc007]

Re: B&Wtech 473nm unit / Spectrometer Mods & Info

As I thought, this has the baffles in it. Once I removed them it started reading more like it should. The calibration is off, but since it is not a linear calibration, it is off more in some areas than others. My 477nm actually measured quite close to that wavelength, but a 532nm measured ~545nm. This software cuts off on the ends of the full spectrum it is said to be aligned for, so I couldn't measure some WL with it. Like my 660nm was off scale on this software. It seems highly unlikely that any calibration will be off by the same amount and in the same direction throughout the whole spectrum.

Brad, are you certain the baffles aren't inside yours? The are small and easy to miss. It sure sounds like they are in yours from what you have said.

I thought 400-650 was rather narrow. Maybe I better look at writing my own software and find out what the full range is. It sounds like we will need it for proper calibration.

Before I removed the baffles I was getting strange readings. Pumping enough 532 down the fibre I could get some reading. That's when I decided to open it up. Once I removed the baffles it came to life as shown in my Pics.

The seller did say it had baffles. But I have seen some post here it didnt have baffles. The strange readings was enough to motivate me to remove it and open it up. I didn't like doing that but I had no choice. Glad that I did.

 

[paul1598419]

Re: B&Wtech 473nm unit / Spectrometer Mods & Info

As long as you're careful and don't introduce any foreign particulate matter or damage the optics, you can open the bench. If you ever need to align it, you will have to again. :yh:

 

[diachi]

Re: B&Wtech 473nm unit / Spectrometer Mods & Info

Mine showed up today, won't be able to do anything with it until I get home though.

Also, the 12V OEM supply for my HeNe supply showed up too - hopefully get that all mounted and looking nice soon! Ordered 3, one can go into my UV exposure box once my beefy 10W UV LEDs show up.

 

[paul1598419]

Re: B&Wtech 473nm unit / Spectrometer Mods & Info

Glad yours finally showed up, diachi. I hope you can avoid some of the mistakes posted here at least. Seems like none of these were calibrated to read a full spectrum as they only needed them to read 473nm. If you pull the spectro out of the laser assembly as I have done, you can use +5 VDC at 335 mA to power it through the mini plug. I'm pretty sure the vast majority of these have baffles in them. Good luck, buddy.

 

[diachi]

Re: B&Wtech 473nm unit / Spectrometer Mods & Info

Glad yours finally showed up, diachi. I hope you can avoid some of the mistakes posted here at least. Seems like none of these were calibrated to read a full spectrum as they only needed them to read 473nm. If you pull the spectro out of the laser assembly as I have done, you can use +5 VDC at 335 mA to power it through the mini plug. I'm pretty sure the vast majority of these have baffles in them. Good luck, buddy.

Got it up and running.

You can run it while everything is still assembled, there's a separate PSU in there for the spectro, it'd be worth pulling that and using it, at least if you plan to integrate it all into one mains powered unit. I've tested it both ways, with the barrel connector and the larger connector mounted next to it. Both ways work just fine.

I had trouble getting it running in Windows 10 but never played around with it a whole lot, just switched to Windows 7. MAKE SURE YOU HAVE THE CORRECT SERIAL CABLE. Not all serial cables are the same, just a heads up. If you're plugging your USB->RS232 adapter straight into the board that won't be an issue.

Mine did indeed have the baffles, which I've since removed (man was that a PITA to open up).

Same results as others, mine works but needs calibrated - as you can see by the graph. 473nm shows at about 455nm IIRC. 532 shows up at around 518. 650 shows up at around 620-630.

Left to right:

Laserlands "520nm" (just left of 500nm), my DIY "515nm" pen, JD850 532nm. Excuse the half assed screenshot, cut out the scale by mistake...

Either the Laserlands 520nm pointer is way lower than 520nm or my 515nm pointer is way higher than 515nm. :thinking:

Also, the software is crap.

Time for dinner, a beer and a smoke. Been working on this since I got home from work.

Will add more pictures later.

 

[Cyparagon]

Re: B&Wtech 473nm unit / Spectrometer Mods & Info

I thought 400-650 was rather narrow. Maybe I better look at writing my own software and find out what the full range is.

The full range is 2048 pixels. You can shift the wavelength RANGE by adjusting alignment. You can change the wavelength SPAN by swapping out the grating. No amount of software wizardry will change the hardware.

I think their software isn't going to be as good as BWSpec, which is the OEM for these spectrometers.

Have you even used the science-surplus software? Unless I'm reading something wrong, people here don't seem to care for the OEM software.

I'm not saying you're wrong - I've ONLY used the science-surplus software, so I have nothing to compare against. But I can tell you it is very usable in practice. Update rates of ~3Hz are plenty fast enough for everything I've used it for thus far. I think lazeerer and DTR will tell you the same.

 

[diachi]

Re: B&Wtech 473nm unit / Spectrometer Mods & Info

Have you even used the science-surplus software? Unless I'm reading something wrong, people here don't seem to care for the OEM software.

Got a copy of it you can share? The "Simple Spectro" software or whatever that we've been using with these works, but it really is simple. Actually, it couldn't possible be any more simple.

 

[Mosc007]

Re: B&Wtech 473nm unit / Spectrometer Mods & Info

Speaking of simple software.

I have started writing my own software for these Spectrometers. The Comms is very simple. Most of the work will be calibrating and having accurate grids on the graph.

Looks like these are fitted with 1800 grating which does give it a narrow range. 400-650 is probably correct. If you want a wider range the grating will need changing.

Anyway, I have my software running. Connecting and receiving data. Looping at 1 second intervals. It took a few hours to get this far and it needs many more hours work on it. Maybe more on it over the weekend.

A sample screen shot with a 505 Sharp diode.

 

[diachi]

Re: B&Wtech 473nm unit / Spectrometer Mods & Info

Speaking of simple software.

I have started writing my own software for these Spectrometers. The Comms is very simple. Most of the work will be calibrating and having accurate grids on the graph.

Looks like these are fitted with 1800 grating which does give it a narrow range. 400-650 is probably correct. If you want a wider range the grating will need changing.

Anyway, I have my software running. Connecting and receiving data. Looping at 1 second intervals. It took a few hours to get this far and it needs many more hours work on it. Maybe more on it over the weekend.

A sample screen shot with a 505 Sharp diode.

Cool! Hopefully you can come up with something better (than what we have). Looking good thus far.

 

[Mosc007]

Re: B&Wtech 473nm unit / Spectrometer Mods & Info

It's very much testing at the moment. All manual controls to set baud, change to ascii mode. It reads the integration time as well. The integration time can be set but I havnt added that command yet. You can also tell the spectrometer to average over a certain number of frames. It's done internal. Many things to do to the software. Should end up good one day.

Manual port select but it does scan and tell you which ports are available. Probably add x and y scaling next. Calibration will be some learning.

 

[diachi]

Re: B&Wtech 473nm unit / Spectrometer Mods & Info

Can these spectrometers be adjusted to work further into the NIR region, lets say above 808 nm? If so, how far might be the longest wavelength?

You'd have to adjust the alignment, unless you have one that's already aligned for that wavelength range.

As for how far into the IR it'll go, depends on the the optics and the CCD. There'll come a point where either the optics won't pass/reflect the wavelength you want and where the CCD won't detect it.