A DIY CCD camera.

[Icosahedron]

Over the years I’ve collected a number of CCD sensors with the intention to one day build a camera. That day is now imminent and over the past week I’ve examined my requirements in order to set out a specification. This is it so far:

1. Sony ICX413AQ sensor.
2. AD9826 16-bit CDS signal processor.
3. Arduino Mega 2560 controller, 1.8” TFT screen with SD card reader.
4. Images to be saved to a SD card eliminating the need to have a pc attached.
5. Pentax K-mount so that camera lenses may be attached.
6. Mechanical shutter to allow short exposures.
7. Flip-mirror for focusing.
8. 12 VDC power supply.
9. Home-made PCB (hopefully).
10. Cooling of course.

Basically this amounts to a cooled DLSR so camera parts will be the order of the day. Unfortunately, I destroyed my CCD donor camera by breaking a crucial flex circuit and now I’m stripping and examining a SLR body as a replacement.

 

[Icosahedron]

After stripping and trimming a Minolta Dynax 500si SLR body with a hacksaw, I salvaged the mirror box, shutter assembly and motor drive.

Operation of the mirror and shutter is straightforward: Apply power to the shutter solenoids and then the motor drive. The mirror flips up, both shutter curtains move and a set of contacts on the shutter assembly makes contact. The shutter is opened by removing power from one solenoid and closed by releasing the other. Again apply power to the motor drive and the mirror flips down with a set of contacts on the mirror box briefly making contact.

I also managed to remove 3mm off the front of the mirror box. This will allow me to fix the Pentax lens mount and SLR body onto a common mounting plate. It also puts the CCD plane behind what is the film plane on the Minolta, allowing for adjustment of the CCD plane. The focusing screen needs to be moved accordingly. It will stay in place during development of the camera, but I intend to eventually ditch it.

Two issues remain: The aperture control lever on a Pentax lens interferes with the front of the mirror box and a lens locking pin will have to be improvised.

[Icosahedron]

The design of the first of two PCB’s is now complete. The board is a customised Arduino Mega shield with the following on board:

1. Breakout provided for all Mega pins.
2. TFT screen and SD card reader accommodated.
3. A 40-pin interface to the second board with CCD and control circuits.
4. Buttons for navigating a menu system.
5. An interface for a remote control.
6. 32 Kbyte of serial RAM. The write latency of SD cards makes it impossible to use while the CCD is being read out.
7. Provision made for the Mega not being able to handle an input voltage of 12V.
8. Backlight control for the TFT display.
9. A TEC controller with a separate power supply input. The controller is a simplified type of SMPS. It is driven by a PWM signal and utilises a Buck converter.

 

[RuralBill]

Following with great interest

Bill

[johnfosteruk]

lovely project, I'll be following too.

[RFE3]

Nice project, This looks like a really neat DIY adventure...Thanks for sharing.

[Icosahedron]

Thanks all for the support; I hope this doesn’t turn into a disappointment.

I’ve already acquired a 60W peltier device as well as a 65W TDP cooling fan. This will help me to determine the size and layout of the camera.

Once I receive more parts, I first want to build and test the TEC controller before committing to the design. Expecting that the peltier will be most efficient with a 10V supply, I would like to measure the ripple at this voltage. I’m not bothered with high ripple at lower voltages which is to be expected from a low duty cycle PWM signal.

The PCB design has already seen a few changes. I’ve a lot to learn about the Arduino platform; discovered that the available PWM frequencies are too low for the TEC controller. Fortunately information is available to get higher frequencies by changing the timer prescaler. I had to move the output to a pin controlled by a different timer as the original is used for the Arduino timing functions. I also neglected to route the I2C lines to the 40-pin interface.

Yesterday I removed the CCD from the donor camera. Desoldering went well and the CCD easily separated from its backing plate. The reason for this is that the epoxy adhesive did not adhere to the backing plate at all but is firmly stuck to the CCD!

[Icosahedron]

The reason for going with a PWM TEC controller is that I intend running the Peltier device off a 12V battery.  A linear regulated supply would mean dumping 10W as heat which would seriously limit battery life. I was hoping for slightly better performance of the controller having measured 1.9% voltage ripple with a 30 kHz PWM signal. When I have more time available, I’ll consult the Atmel data sheet and figure out how to get higher frequencies.

Otherwise I’m very happy with the design. The NTP85N03 power MOSFET has an "on" resistance of only 6.1 milliohm. At 5A the dissipation is about 150 mW, which is a little more than what a 1000 ohm resistor dissipates when connected to a 12V supply. My finger tells me that the dissipation of the inductor is slightly higher. The PCB itself will be an adequate heat sink for these components.

It took only two minutes to remove the epoxy adhesive from the back of the CCD with methylene chloride.

[Icosahedron]

The Mega shield board has been assembled and tested successfully. Shortly before making the PCB, I thought it would be convenient to be able to mount the camera on a tripod as well. This meant moving all the interfaces to the same side of the board. The bottom of the camera is now reserved for a tripod mounting, the top for a viewfinder and SD card access and the remaining for accommodating the cooling fan.

The Microchip 23K256 SPI RAM chip runs off the Mega 3.3V supply. To simplify things, I opted for 1k/1.5k ohm resistance dividers for converting the 5V signals to 3.3V, knowing full well that I might not be able to run the device at the highest speed available. I was only able to test this after assembly and am glad to report that it performs flawlessly.

[Elliot Lucak]

Hello, have you done any more work on this?

I'm very interested to see your progress on this project as I stupidly managed to wreck my d100 (Sony ICX413AQ), and want to create a very basic ccd camera with it much like your project.

 

The only other websites I've found are far beyond my scope\skill level with electronics as they're also astronomy projects, which I suppose have very high requirements in image quality, while I only want to have a usable camera. Perhaps even a bad quality camera is out of my scope, but I would like to hear what you would suggest.

[Icosahedron]

Hi, at the moment I’m working on the design of the case and choice of raw image file format. I’m still months away from doing the circuitry and software for reading the CCD. Also, I only have experience of reading linear sensors.

There doesn’t seem to be any projects to be found on the internet making use of this sensor. This project inspired me to start with mine. They have been very generous and have published their hardware, software and PCB designs. I wanted something a little different and opted for the ICX413AQ as the data sheet is available. The ICX453AQ is a proprietary Nikon chip.

A lot more information is available here.

[Elliot Lucak]

Ah, I see. That is much simpler than the projects I was looking at before, and actually gives me a much more clear idea of what I need to do and things that I could do with my project. 

I also could have sworn I already searched for diy projects involving the ICX453AQ with no results, but worst case (If ICX413AQ doesn't work) I do actually have an ICX453AQ that I could use with some slight case modification.  

Thanks, and best of luck with your own project!