Parts lists

Bill of materials with ordering information: html, excel.

Parts listings by board, ordered by part number;

• Controller board: html, excel.
• Main camera board (KAF401/1602): html, excel
• Guiding camera board (TC211): html, excel.
• Power supply board: html, excel.

Schematics

• Controller board: Part 1, Part2
• Main camera board (KAF401/1602).
• Guiding camera board (TC211).
• Power supply board.
• All files in a single zip file including the original Ivex source files.

Board fabrication files

NOTE: Errors found in Version 1.3 (no longer available for download)

Shorted -12V supply to GND on controller board near C40. Short between 15V and GND on main CCD board near C34 and R35. Thank you Piotr Kucharski for troubleshooting and reporting these errors.

• Board layout maps for all boards.
• Ivex board design source files.
• Gerber files for camera boards and controller board (Version 1.4).
• Gerber files for power supply board (Version 1.1).

Troubleshooting

No comprehensive manual has been compiled as of yet, although there are some hints and suggestions in the help files distributed with the Pyxis software. This help file is also available online by following this link: Pyxis capture application online help.

Low level patterns such as stripes or bars on the CCD frame that are not always in the same location after each frame capture are most certainly due to noise pickup. Check for ground loops or bad ground connections, external sources of noise such as DC-AC converters and interference from high-inductance loads (e.g. electric motors) placed near the camera. Often the frequency of the noise source can be determined from the patterns on the frame. Use the readout time divided by the number of pixels to determine the readout time per pixel; from this you may be able to determine the frequency of the noise source. This often provides a good hint as to the provenance of the noise.

The on board regulators in the earlier versions of the Pyxis camera were incorrectly wired and oscillated in consequence. This problem manifested itself as strong pattern noise in the CCD images. This problem has been fixed, however it was noted that if the small on board regulators used on the camera clock boards were overloaded, then oscillation similar to that observed with the incorrect circuit might ensue. If a strong pattern noise is observed in the image, you should suspect a problem with the supplies on the camera boards. In particular, look for shorts of the supply to ground or incorrectly mounted tantalum decoupling capacitors. Note that a short due to an inverted tantalum capacitor may not appear when the circuit is not powered up (i.e. your continuity tester may indicate a reasonably high impedance between ground and the supply line); only comparison of the board with the component layout data sheet will turn up the culprit. This problem is sufficiently common that it should always be suspected first; it manifests itself in either a sudden or gradual fashion, often only appearing after several hours of trouble free operation.

If the pattern is fixed on the frame then this indicates a problem arising in the readout chain; either the timing of the readout clocks or the clock levels.

If the pixel values jump by large, discontinuous values then one should suspect a problem with the data lines between the ADC and the bus. Incorrect timing of the readout of the ADC word can also lead to such problems. A bad bus line may cause erratic behaviour of the digital interface, or complete loss of communications. If an image is successfully acquired under such conditions, the identity of the bad bus line will be evident; pixel values will always have one bit tied high or low.

Intermittent operation is almost always due to a dry solder joint. Try to narrow down the source of the problem, checking for good continuity between nodes on the suspected lines. Again, this is a problem that might only occur after several hours or months of operation. After having run into this problem a few times, I find that the extra time spent in ensuring good soldered connections during assembly always pays itself back a ten-fold.

If Windows does not recognize the USB device on power-on of the controller, then there is a problem with the connections to the FTDI USB chip. Having never run into this problem myself, I have few suggestions to give except to check the assembly near the USB chip. Basically the FT245 chip works, and there isn't a whole lot that can go wrong. It is important to follow the instructions given in the help file concerning the installation of the USB drivers or the interface will not work!

The following FITS image was taken using the prototype KAF-401e camera in darkness, with the CCD cooled to -20°C and the cooling supply turned off during readout. This is the expected appearance of a Pyxis offset image, although the exact offset level may vary between CCD chips. Offset_frame.fit