Arduino spectroscope with TSL1401 line sensor and TFT-display - Arduino Spektroskop
Vložit
- čas přidán 19. 12. 2018
- more information & code: stoppi-homemade-physics.de/sp...
www.instructables.com/id/Ardu...
music: coward - Hans Zimmer - Věda a technologie
Spectacular project. Thanks for sharing.
This is crazily great! congrats
I need this in my life. Thanks!
Very nice!
I would like to see how well it works with some sample in cuvette.
this is a most wonderful line scan project and hard to find DIY info. 💪 it would beneficial be to have an arduino/raspberryPi project book on line-cameras like this one. more than following a silly line. something such as building a raster line by line in a scrolling mode...😎 thanks..:)
Großartiges Projekt! Ich glaub das ist die beste Umsetzung eines DIY-Spektroskops die ich bislang gesehen habe. Sieht schön robust aus.
Danke ;-) Die Umsetzung mit integrierter Anzeige habe ich auch noch nicht gesehen...
habe noch einige TCD1304 mit entsprechenden Trägerplatinen (tcd1304.wordpress.com) und 1000er beugungsgitter sowie ein Nucleo401RE hier rumfliegen, da ich auch mal ein spektrometer bauen wollte, aber etwas unsicher auf optischer Seite war. Einfach mal kleine achromaten zu verwenden scheint das Problem ja zu lösen. Spielt die Größe der Linse eigentlich eine Rolle? der TCD1304 hat ganz schön viele Pixel auf 29,1mm verteilt, da braucht man bestimmt ein f von 50-60...ob wohl 13mm Durchmesser mit dem f ok sind?
Es gibt ja eine Firmware für das STM32-board für den TCD1304, ich dachte man könnte einfach nochmal eine Display/MCU-Kombo als interface da dranbasteln. Kost ja alles nix heutzutage.
Hört sich spannend an. An ein CCD-array mit deutlich höherer Auflösung habe ich auch schon gedacht. Gibt sogar einige Ansätze, dies mit Arduino umzusetzen. Wenn dein Sensor deutlich größer ist (29mm zu 9mm beim TSL1401) brauchst du einfach wie du richtig schreibst eine größere Brennweite, z.B. 75 mm. Schau dich mal bei astromedia um, die haben einige gute Glaslinsen im Angebot... Meine verwendeten Linsen haben einen Durchmesser von 18 mm. Bei einer größeren Brennweite bekommst du auch leichter welche mit größerem Durchmesser. Bestimmt dann die Lichtempfindlichkeit deines Spektrographen. Obwohl ich mit unzureichender Belichtung keine Probleme habe, eher im Gegenteil bei den verwendeten Lichtquellen. Bei den Laserpointern bin ich bei einer Belichtungszeit = 0 (d.h. delay = 0), da sonst die "Aufnahme" total überbelichtet wird. In Wahrheit natürlich >0, da die clock-Schleife (diehe Datenblatt TSL1401) ja selbst ohne zusätzlicher Verzögerung auch ihre Zeit braucht und währenddessen ja schon belichtet wird ab dem 18ten clock...
@@stoppi danke für die ausführliche Antwort. Habe zwei f=55mm Achromaten von Zeiss, die nur 13mm im Durchmesser haben aus irgendeinem Optik-ebay-shop. Waren schön günstig. Der Astro-shop hat jede menge varianten mit f>90mm. Das mit der Lichtempfindlichkeit in Abhängigkeit von der Linsengröße leuchtet mir ein. Man kann wohl bei dieser stm32 firmware die exposure time in ziemlich weitem Bereich einstellen.
Selbst wenn man nur teile dieser 3tausendIrgendwas Pixel benutzt, reicht das dicke. Das sind jede menge Daten.
Nice job!!! Can you measure in uvc range 200nm to 300nm with this? or is another sensor needed?
Maybe i try this. Looks better than mine paper one :-D
hi your video is very instructive. Actually I’m working on same by following your website. I have some question. I don’t have a TFT display but I want to draw a line chat by using Excel. What I’m trying to do is print brightness from first to 128. Do you think is right?
excellence my friend...now addin a sd card to record the info in the lcd
Hi stoppi, I'm a student from Malaysia. I followed your steps in making the arduino spectrocopy. I did insert the coding that you provided but the spectrum in my LCD did not give the peaks according to the LED colour. The peaks only shows where I move the light not following the colour of LED. Do you know what is the problem?
amazing
Thank's a lot
Hi there Stoppi, is there a more direct way to communicate with you regarding your work on this project? I'm very keen to have a go at this project myself but would very much like to chat to you for guidance with building it, more so with making a few modifications to the Arduino code.
Hey friend! Where is it used?
Great project and well presented. One question; what is the material you are holding in the peg and what is it's purpose? I can't see this mentioned in your discussion? Cheers
Hi! Well observed and good question. It's a neutral density filter to weaken the laser-beam, because without it the sensor gets overexposured even with the shortest exposure time ;-) This is just a problem with laser-beams, because they are so colliminated. With other light sources like flashlights or LED's you don't have this problem...
Thanks mate - will any neutral density filter do or is this a specific type and/or brand? Where did you get yours from? Cheers, from Down Under
Also (from above) what power output are your RGB lasers. And do you think it would be better to coat the inside of the box matt black to prevent any undesirable effects from to stray light? cheers again.
@@ozgemmo3445 Hi Ian! I used the Kodak Wratten Nr. 96 with the neutral density = 1, which means that just 10^-1 = 0.1 = 10% get through the filter. ebay: www.ebay.com/itm/Wratten-96-ND-0-70-NEW-Filter-3x3-inch-75mm-Broken-Seal-2-1-3-stops-Kodak/123515065929?hash=item1cc2125649:g:dDMAAOSwwvZZWT2o
I'd recommend a ND = 0.4 to 1.0. If the attenuation is too low you can stack the filters. So 0.4 = 10^-0.4 = 0.39 = 39% is a good value. Tow of the filters will reduce the light to just 0.39^2 = 0.15 = 15% and so on...
Greetings from austria to australia, stoppi
@@ozgemmo3445 Here in this video (czcams.com/video/9ZaDhGMfNkU/video.html) I used the same box as for the sprectroscope. The problem was, that the box isn't absolutely light-tight. For the photomultiplier, which is far more sensitive than the TSL1401-line sensor, this was a great problem and therefore I had to cover the box with black tape. For the spectroscope it isn't necessary.
The laser-pointers have different power output, between 5mW (red) to 45mW (blue, though declared as a 5mW-pointer too). I've measured the power with my homemade thermopile (czcams.com/video/wkKf5D6BxUg/video.html).
For the neutral density filter you can also take a cheaper version not from Kodak like www.ebay.com/itm/55mm-Neutral-Density-ND8-DSLR-Lens-Filter-For-Tamron-SP-AF90mm-f-2-8-Di-MACRO1-1/392113632966?hash=item5b4bcb36c6:g:nGUAAOxyIAZRuZJk
But they are made of glass and therefore you can't stack them as simple as the Kodak wratten filter. And be aware, the specification ND8 doesn't mean the same as ND8 from the Kodak wratten filter!
Hey stopper! It was really a nice video and i really got a lot of help. But i have few questions regarding the setup like what is the distance between the slit, first lens, grating, second lens and the sensor and what is the angle between the grating and the sensor.
The distance between the slit and the first lens should be the focal length of the first lens. Then the light which goes through the grating is parallel. Therefore the distance between the second lens and the Tsl1401 sensor must be the focal length of the second lens. To find out the right angle between the grating and the second lens/Tsl1401 sensor, I used laserpointer and tried to focuse them right on the sensor. It's a bit tricky... Good luck
@@stoppi thank you so much.
@@sanjayrathore4911 you're welcome
Also if i'm using halogen lamp as the light source, do i need neutral density filter?
HI! No, just for the laserpointer I needed a neutral density filter to decrease the intensity...
Great looking project!
Can be data stored in memory card or sent to pc for detailed analysis? I want to build similar one, but I need this functionality...
Hello! Thank's. Of course you can add a memory card to the existing system and save the intensities after pushing a button f.e.
@@stoppi Will it works with TSL1402 for better spectral resolution?
@@MichaelWeizenfeld the Tsl1402 is just twice the Tsl1401 so I couldn't find any arguments that it will not work. The only problem could be the small separation between the two sensor modules of the Tsl1402. If this separation is larger than the separation between the other pixels, then it may cause problems (step on the x-axis). When you use the Tsl1402, you will need a lens with double focal length compared with my setup...
@@stoppi thanks ! One more question. Is Arduino capable to work with other linear sensors like TCD1304 (they are cheaper and has more resolution) or it is not possible?
@@MichaelWeizenfeld I have looked for those combinations too but I didn't find a successful way based on arduino...
As far as I remember only with reduced resolution but this doesn't really make sense then...
Have you used a the 3.5 TFT LCD shield with this setup? would it require modification on the software etc?
I haven't used a shield. The TFT-display is directly connected with the header of the arduino mega.
@@stoppi thanks i did a little investigating and i noticed i will have to rewrite some of the sketch to use other pins since the display plugs in to the pins being used by the TFT shield display etc... Also will probably need to up load drivers to library since it is a different drive IC of ILI9486 than what you used etc...
@@georgeyork8975 okay, good luck
Don't ever stoppi making vidoes
Haha, I do my best. At the moment there are around 40 videos waiting to be uploaded 😉
@@stoppi awesome! Can't wait!
Nice project but wouldn't you be better using a webcam with more resolution ? I think these are 128 array?
Hi! Yes, the resolution is poor but my intention was to build a compact device with no need of an additional computer or laptop, like all-in-one...
Вы должны использовать МОНОХРОМНУЮ камеру, обычные имеют цветные фильтры перед матрицей, они не позволят корректно измерить интенсивность спектра.
were did you buy the ccd?
I bought it on ebay...
can u helpme please
*Interstellar*
Does it see UV and IR spectrum?
At the moment I have adjusted a spectral range from 390nm to 690nm. To increase this to detect also ir you'll have to use a second lens with a shorter focal length...
@@stoppi Sounds great! Thanks!
@@user-ug4ow1qq2h but you'll have to proof, whether the glass lense doesn't absorb uv- or ir-light. Therefore maybe you will have to use quartz lenses...
@@stoppi This is an absolutely awesome project. Thanks you for sharing. I have a couple of questions. If you wanted to see between 390nm and say 740nm, I see you recommend a shorter focal length lens. What focal length would you recommend? Also, for the diffraction grating, would 600 lines per mm be sufficient or is 1000 lines per mm recommended? Thanks again.
@@smabonham Hello! Thank's for the compliment. You can calculate, which focal length you would need. The formular for the diffraction of the grating is sin(alpha) = lambda/d with alpha = angle, lambda = wavelength and d = grating-constant (f.e. 10^-6m). For example you want to cover 390 to 740nm then you calculate for both wavelengths the angle alpha and calculate alpha_740 - alpha_390. Then you need the length of the TSL1401 = L. To calculate the necessary focal length f you need the formular tan(alpha_740 - alpha_390) = L/f. With the known alpha's and L you can calculate the focal length f you will need ;-)
>>> ENGLISH LANGUAGE
Hi Livio! Wouldn't it be possible to calibrate the intensitiy of the spectrometer by for example using 1...10 LEDs in steps and comparing the 10 generated values for each frequency? Even if the sensors sensitivity is a nonlinear function of frequency and light intensity i'd think you should be able to get some kind of calibration curve (sensitivity over light intensity) for each freqency. What am i getting wrong?
Also i am wondering why a whole lot of people made videos of their specs with plans to (sometimes quantitatively) compare the absorption of different solutions but no one seems to actually show a working result. Shouldn't a rather small difference in the intensity spectrum already generate a sufficient spike in the absorbtion graph because of the log10 in the absorption formula?
Sorry for the long questions, just a mechanical engineer here not really getting it :D