linear CCD TCD1201d readout and timing issue

yes ofg course, only the array of data use 1k, so you would not have ram for anything else.

you may solve reducing the size to 512 points and increse the "sdacCounter" only when "i" is even, for example.

Hello, help to connect tcd2252d to arduino mega. Want to make a scanner.

btw i got this sensor (TCD1201D) time ago and wrote an article about it (it works!)

http://mauromombelli.com/2015/02/20/TCD1201D.html

picture details:
WP_20150922_002.jpg show signals SH, PHI1, PHI2
WP_20150922_003.jpg show signals PHI1, PHI2, RS, BT
WP_20150922_004.jpg show signals OS, PHI1, RS, BT

I must mask the CCD Window…this is the right time, i hope!

…the problem was the very sensitivity of the CCD, i mask with tape (2 layer) on the CCD window and it run well (finnally). How can utilise it, for example to measure an object ( see https://measure.feld.cvut.cz/groups/videometry/Publications/Papers/MEAS03_Radil_slides.pdf) ?

yay! good job!

Hi there !

I'm trying to implement a TCD1201 with an Arduino UNO with the hope to use it to complete a spectrometer. I've used the code lesto gave which is working perfectly well.

As I would like to work with long integration times (typically above a second) I would like to know if I could still use the full 2048px of the array and spare RAM by reading & sending pixels by packets of say 512 ? Do you think it would significantly impact the total integration time ?

Thanks btw for such a well documented work :) !

Alex

the sensor output pixel by pixel, as you send out the clock signal. Right now the bottle neck is the analog read, way too slow, you could remove it and get fast reading and smaller integration time..

so you can disable the analog read for some measurement (if you look at the code history i think you will find in the beginning i was reading just odd pixel)

if you want a suggestion, use a more powerful MCU if you need something serious.

Andrea1970: ...the problem was the very sensitivity of the CCD, i mask with tape (2 layer) on the CCD window and it run well (finnally). How can utilise it, for example to measure an object ( see https://measure.feld.cvut.cz/groups/videometry/Publications/Papers/MEAS03_Radil_slides.pdf) ?

very interesting, seems to be possible in theory

lesto: if you want a suggestion, use a more powerful MCU if you need something serious.

Thank your for such a quick answer :) ! I'm actually thinking about reading all pixels, this is the huge problem..

I'll probably be adding a 23K256 32k SRAM chip to store pixel values and an external 14-bit ADC in order to get a good read, I chose Arduino because of its simplicity + built-in USB interface, I can't find a similar product for such a cheap price..

You could use an Arduino Mega as it has enough SRAM to read the entire CCD and you can buy clones on eBay for about £11. I did a similar thing using Mega and an ILX511 here.

Feynmaniac: I'll probably be adding a 23K256 32k SRAM

you cannot add sram to a 328 and work on it like internal sram. I mean you cannot extend internal sram

Maybe an STM32 like the Maple Mini or the cheap board callde Blue Pill.

Hey guys, i'm in a project with this sensor. So, I saw the lesto's project, but i don't know how it doesn't work with me. I've try everything already. But it just doesn't work.

Some questions: In this prototype, Do I need some resistor, capacitor or something else? or Can I put all directly on arduino?.

In this project: http://imgur.com/a/ihUWY, What is the little yellow piece on left up side?

I try this too, and doesn't work.

Some one can give me a light about this?

Sorry for the bad english.

Thanks!

The photo you linked shows direct connections and there is only one capacitor for filtering noise in power line, it is the little yellow piece on left up side.

im extremly sceptical, i had to put 3 layer of eletricaltape and the sensor was still sensible to ambient light and going in saturation when using the 5w classic red laser...

I guess you can decrease the integration time as it seems the integration is the time of the previus reading, so if you do a reading without ADC the next one may be less sensible

so, cover that sensor as much as possible and then shine a laser/light on it

I was following same code and wiring as Lesto, but it act's very strangely. |500x400

First is analog input, second clock 1 (0-2.5V), third clock 2 (strange 0-(-0,4V)), fourth seems like shift gate.

Also when I partially cover the sensor reading seems to stop in like 70% of sensor then goes black. |500x281

The sensor isn't covered in two places but It shows like it is. Also. Have anybody tried reading using Mega? I tried but failed. I even changed ports in code but it didin't helped either. Can anyone please know what's going on, if it's sensor fault?

the sensor must be completely covered, otherwhile the light that enter will diffuse in all the sensor. use a laser pointer to shine light on a specific point.

Also your clk 2 seems estrmely strange, i guess you have some issue with the ground of your probe, check connection again

the image is not visible, attach directly it here

Dragonsired1:
Have anybody tried reading using Mega? I tried but failed. I even changed ports in code but it didin’t helped either.

I have read a similar CCD array (ILX511) using Mega but as the MCU is barely able to keep up the data rate needed the lens needs heavy shielding from light (as lesto states) else the CCD reaches saturation between reads.