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

[NightLase]

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

To me it looks like the light is not evenly spread across the CCD sensor..

The light source is from the Iphone LED, they have a broad curve with a blue peak, hence the appearance.

 

[NightLase]

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

One: Notice how the 585nm line is clipped. You're saturating the sensor.

Yes, this I Understand... To get the other peaks to come up, the 585 line was clipping unfortunately. As I understand too, this spectrometer is very sensitive to low light levels (as required for the Raman application it was used in).

Two: Notice how the base of the spectral line is broader on the right than on the left. This is an astigmatism in the optics alignment that I've mentioned a few times prior, and should be fixed.

Ah excellent, thank you for confirming this.. I do understand that alignment would be needed, I guess is how much from the original application settings.

compact fluorescent lamps have Hg and Ar,

Yes, I have seen several posts indicative to using CFLs and Neon Indicators as calibration sources..

My only question would be how accurate a CFL would be across the bandwidth as they are not "clear" glass allowing for the pure spectral lines to come through (due to phosphor converting to visible.

Hence my understanding why there are specific element spectral lamps used for calibration (yes at a cost)

I have pure Neon tube (neon sign) and an Argon (which has added Mercury to brighten).

As im still learning and understanding the concepts including calculation of the coefficients, I made initial calibration on wavelengths using what sources I have.

The screen captures was just of Neon as a quick method to present something in the rough.

Cheers

 

[diber]

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

My only question would be how accurate a CFL would be across the bandwidth as they are not "clear" glass allowing for the pure spectral lines to come through (due to phosphor converting to visible.

The phosphor coating causes indeed very broad peaks in the spectrum and can change from brand to brand and lamp type to lamp type. However, the discrete lines from the individual elements are still clearly present above this wide background fluorescent spectrum. The position of these narrow lines is, as far as I know (anyone: please correct me if I am wrong on this !), not affected by the fluorescence and can hence be perfectly used for the spectral calibration of the spectrometer. However, the intensity ratio between the different peaks could be affected by the background fluorescence spectrum, but you do not need the peak intensity for the determination of the calibration coefficients of the spectrometer

Thus, according to me:
- if you need nice spectra of the elements, without a background and correct peak intensity ratios -> you need the special lamps.
- if you just want to determine the calibration coefficients of the spectrometer -> a CFL lamp is perfectly fine.

When you receive your spectral lamps, maybe you can do the comparison and let us know the results !

 

[Mosc007]

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

I have been running some tests on both the B&W units I have.


Has anyone noticed the CCD in these units does not detect anything below about 473nm ?

I have been playing with Calibration using a CFL tube and both units have no Blue response. I have opened one up and directed a White LED torch into it. I can easily see were on the CCD the Spectrum falls. This one was right up the Red end. It is only using about 50% of the CCD. I adjusted the final mirror and placed the spectrum in the middle of the CCD. I now have a much larger Red end. The Spectrum from the CFL is still correct. Just shifted left. But still no Blue Response.

I can see the Blue is missing when I shine the LED torch into it. The Red end looks like it goes all the way into IR until I can't see. But the Blue end stops around Cyan. No darker Blue color is visible.

This is very strange. I am able to get a VERY good Calibration from 473 up to 650. It seems very accurate. My Sharp 490 is reading 490. I used quite a few calibration points from the CFL. But the Blue end is trash as it doesn't respond to Blue light.

Is it possible one of the Mirrors has a Blue Cut Filter on it ?


PS, There is a Bug in my Spectro software which occured when I added the Save and Load Spectrums. It's not saving the new Calibration Data. You can of course manually enter the data into the DAT file for the Serial number you have. I have fixed the bug but wont upload the fixed version yet.


Here is a capture of the CFL showing the Lines I used to Calibrate. I also added 473nm into the calibration but its not show on this graph. I also added an inaccurate 650nm from a pointer just to get a better top end.


EDIT, Not easy to capture but I managed to capture the light that falls on the CCD from the White LED Torch.. It's pretty obvious Blue light is getting filtered out before it makes the CCD. See the Pic.

 

[jnrpop]

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

Mosc, yes there's an optical filter near the grating. Only after I removed mine was I able to get any spectral readings below ~475nm. See back in this thread 5 or so pages. I'll try find the exact post

Edit: post 324 & 331

 

[Agastar]

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

Here is a graph of the CFL bulbs in my ceiling fan. Even with two different color temp bulbs in the fan, I'm still able to get a graph that looks similar to ones I found on this site Compact Fluorescent Light Spectra.

I'm planning making a black box that I can put the spectrometer in and test different light sources that I can find around the house. The box will be able to extend or collapse its length so I can adjust the distance of the light source from the spectrometer.

I also powered my spectrometer with my Rigol DP711 power supply and at 5V, the most current drawn was 360mA. The power supply that comes with it is rated for a little over 4 amps which is overkill but I'm going to include it in my case for this spectrometer with a PC style plug and switch on the back.

 

[Mosc007]

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

Mosc, yes there's an optical filter near the grating. Only after I removed mine was I able to get any spectral readings below ~475nm. See back in this thread 5 or so pages. I'll try find the exact post

Edit: post 324 & 331

If you are referring to the Baffles, I removed them when I first got the Spectrometers.

Or is there something I cant see ?

 

[Mosc007]

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

Here is a graph of the CFL bulbs in my ceiling fan. Even with two different color temp bulbs in the fan, I'm still able to get a graph that looks similar to ones I found on this site Compact Fluorescent Light Spectra.

I'm planning making a black box that I can put the spectrometer in and test different light sources that I can find around the house. The box will be able to extend or collapse its length so I can adjust the distance of the light source from the spectrometer.

I also powered my spectrometer with my Rigol DP711 power supply and at 5V, the most current drawn was 360mA. The power supply that comes with it is rated for a little over 4 amps which is overkill but I'm going to include it in my case for this spectrometer with a PC style plug and switch on the back.

You are also missing the Blue. There should be a peak at 487.6nm and another at 435.8nm and also 404.6nm if you can see that far into UV. Your graph only shows the Green and Red Lines. Same as mine.

I have several CFL spectro Graphs and they all show the Blue. Which is missing in this Spectro. After calibrating accurately I should be able to see down to about 400nm. But even a 462nm Laser shows nothing.

If you look at the pic I took of the CCD with White light you can see the Blue cuts off around 473nm. There is no real blue hitting the CCD.

Another test you can do is point the Fibre at your LCD display on a white screen. There is only Green and Red. The Blue is not showing in the graph.

 

[jnrpop]

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

If you are referring to the Baffles, I removed them when I first got the Spectrometers.

Or is there something I cant see ?

Hi Mosc, in my mind, the baffle is near the CCD, the filter is in the black tube near the slit. or is this the baffle for you? have a look at the posts i referred to in my last post and you'll no what i mean by filter, little round optic

 

[diachi]

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

If you are referring to the Baffles, I removed them when I first got the Spectrometers.

Or is there something I cant see ?

There's another filter fitted to the fiber input, if you search back a few pages someone has posted pictures of the removal process. Looks like a block blue pass red/green filter, which makes sense given the application.

Also, watch for the double posting.

 

[Mosc007]

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

Hi Mosc, in my mind, the baffle is near the CCD, the filter is in the black tube near the slit. or is this the baffle for you? have a look at the posts i referred to in my last post and you'll no what i mean by filter, little round optic

Thanks for explaining this. I was not aware of the blue filter at the fibre input. I will check it out. This sounds exactly what is happening.

By adjusting the grating and aligning the yellow to about the center of the CCD it looks like it will cover from below 400nm to above 700nm. The beam was set so 650nm was at the right edge of the CCD.

I will do it tom. Its about 3:30am here.

 

[paul1598419]

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

I have found the best way to remove the filter is to remove the two screws on the back side of the slit; inside the optical bench. Once you have those out, there may be some silicone adhesive which you must pull off without touching anything else inside the bench. Just discard the little black cylinder as this is where to filter is located.

 

[Mosc007]

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

Thanks for all the help. I couldn't wait. It was so simple to remove the Filter. To remove the filter I just losened the grub screw and carefully removed the black cylinder from inside.

I now have Blue. I can now see the line at 487.6nm from the CFL. It now works on my 462nm Pointer. It is hard to be certain of the max and min range without a laser to test it with. The Calibration curve shows 400nm but my 405nm pointer does not show. It's possible moving the grating will show it. The Top end shows 681nm but maybe I will lose that if I move the grating. It looks like its easily 450nm to 650nm. And a bit more.

I re-calibrated with the CFL using 487.6 , 543.6 , 546.0 , 610.8 , 620.6 then added 473 from the laser on this Device. I also added loose lines at 462 from a pointer and 650 from a pointer.

Because of the way 3rd Order polynomial works the calibration goes wobbly above and below the max and min calibration points. But it looks very accurate between them.

Testing a few diodes. I have always thought my 505nm diode was high. It is over driven to 100mw. It reads 511.0nm after calibration. And my 490nm diode reads 491.5nm. My Cheap Chinese 532nm reads 533.0nm

Very happy with the calibration now.

I have also posted the fixed exe file for my software. Editing the Calibration data now saves correctly. This is just the exe file. If you have not used it before you will need the full install file posted some time ago.


Spectro v1.4


PS, The version number is probably the same as the old one. But it is updated.

 

[paul1598419]

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

I would redo the calibration leaving out the direct diodes as they are likely adding error to your calibration. Only use He-Ne lasers, DPSS lasers and other mercury/argon vapor sources. Then your calibration will be as accurate as you can get it without doing a realignment. Only measure your direct diodes after calibration and see how close they are then.

 

[chloderic]

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

Thanks for explaining this. I was not aware of the blue filter at the fibre input. I will check it out. This sounds exactly what is happening.

By adjusting the grating and aligning the yellow to about the center of the CCD it looks like it will cover from below 400nm to above 700nm. The beam was set so 650nm was at the right edge of the CCD.

I will do it tom. Its about 3:30am here.

Hello, before you turn ONLY the grid, you should inform yourself about the other resulting effects, such a grid always has an individually designed diffraction efficiency by the blaze angle.
If you turn the grid , you aditional to the direction of the scattered wavelengths also change the efficiency and possibly remove you from the maximum, this is basically a change in the direction of wavelength illumination ... but also has side effects.
Maybe you also consider turning only the last focus mirror to change the illumination on the ccd, but that will have negative effects too like astigmatism ..... any change of just ONE optical component actually forces you to to readjust the others ... basically you have to begin with the first part, that meens the fiber (use ONE fiber and fix it).
And one more, special CAUTION on the first focusing mirror after the slot ..... here you have quickly adjusted the parallel beam path in the direction of grid because it has a keyhole and you can rotate it AND move it .... (that's IMPORTANT).

Maybe you first read something like this page here, or something synonymous of other dealers or manufacturers in maybe other versions.

https://www.thorlabs.com/tutorials.cfm?tabID=0ca9a8bd-2332-48f8-b01a-7f8bf0c03d4e

now i have seen, that paul posted something and YES ... before changing optics you should do what he said, changing optics is difficulty and has to be the last what you may do if realy needed

 

[Agastar]

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

You are also missing the Blue. ...

That is correct, I have not removed the filter on mine yet

Looks like you have it all figured out now and yours is working better. My other 5 machines came in today and now I have 7 so I guess there is no reason not to remove it on this one