Troubleshooting the Cookbook CCD camera

by Veikko A. Kanto

This document offers help for the ten most common types of problems encountered by Cookbook builders. At the time we completed The CCD Camera Cookbook, we had overseen the construction of 15 Cookbook CCD cameras. With well over 2,500 Coookbook camera now in operation, we have learned a great deal more about the problems you may encounter when you construct your Cookbook camera.

Please don't let a list of problems deter you from building a Cookbook camera. This list is here to provide guidance if you encounter problems. Most Cookbook builders have constructed successful cameras working only from the book. Companies that design and build electronics find similar problems as the product goes into production. As a customer, you're not aware of these fixes because the production manager insures that delivered product will work.

Think of yourself as a production manager: your job is to become aware of problems and fix them, carry out the work carefully, and never, ever cheat on quality. Remember, when you build a cookbook camera, you are your own customer and you want nothing short of the best possible product.

Before going any further, we want you to eliminate a potential problem. In some Cookbook cameras, the -9.5V regulator circuit oscillates. Please add two 1000uF electrolytic capacitors per the instructions given on page 160 of the book (Appendix C.3) before performing the interface card power supply test. Adding these capacitors insures that this problem will not bother you.

1. Soldering Problems

Intermittent or poor solder connections are often the causes of camera problems. This problem is difficult to troubleshoot since it may come and go intermittently.

• Visual inspection with a magnifier is your best bet for finding poor solder joints. Good joints are shiny and the solder clearly wets both the pad and the wire. The camera that my wife Betsy built for the book stopped working while Willmann-Bell was testing it. The problem was caused by a poor solder connection.
• Poor solder connections are those in which the solder does not "wet" the metal surface. The solder may initially stick to the surface only to break away later. Metal surfaces that are only in contact with each other may fail when an oxide layer forms between the metal surfaces. Electrical conductivity is then lost. In other words, the joint may work for a while and then stop working. One common problem occurs when the solder only flows onto the pin of a socket. It may look like it is soldered to the card but it is not. The giveaway is a negative meniscus on the solder. When you solder, be sure to press the hot tip of the soldering iron against both the solder pad and the component lead. For a good solder joint, both soldering surfaces must be hot because solder flows only onto hot surfaces.
• When soldering stranded wires to the cards and connectors, it is important to wet the wire with solder first. Heat the wire with the soldering tip and flow solder onto the wire. The solder should evenly coat the wire and an outline of strands should be visible. The insulation normally melts back a little and you can clip the wire off a little to bring the uninsulated wire back to a reasonable length. If you solder a wire in that has too much conductor exposed, it can short to adjacent wires or traces when it is moved around. Check that the wire fits into the card hole before soldering it. If it does not, you may have too much solder or a wire size that is too big. If you wet the socket solder cups with solder, a little flux is left on the solder surface. To make the wire connection, just put the tinned wire in the tinned solder cup and press the iron tip on the wire to reflow the solder. No extra solder is needed. Don't use old wire. Copper wire is often tin plated and when it ages, a copper-tin compound forms that does not wet with solder.

2. Socket Problems

Sockets amd plugs are the least reliable components in most electronic cirucuits. They can cause problems ranging from erratic performance to total inoperability.

• Poor socket mating is a frequent cause of intermittent electrical connections. These connections rely on spring compressive forces to make metal-to-metal contacts. The pins generally have a wiping action that scrapes oxides and organic contaminants from the contact surfaces as they are mated.
• It is important to use the machine tooled sockets recommended in the CCD Camera Cookbook. Do not be tempted to substitute other types of sockets. Inexpensive low-profile sockets have flat blade contacts can deform permanently if a large diameter pin or probe is pushed in the socket hole. Once deformed, the contact in the socket fails to mate with the part. The machine tooled sockets have the pin conductor exposed on the top of the socket strip and you don't have to push a probe into the socket hole to get a meter reading.
• Intermittent connections to the J2/J3 socket will occur if the connectors do not fit squarely together. The preamp card should be inserted into the socket so that there is no tilt and no gap between the plug and socket. With the connectors squarely mated, the preamp card fastening screw location should be marked on the rear aluminum cover bracket. Drill and tap a 4-40 hole in this position. Cut off a 4-40 screw to make a stud. Insert this in the bracket and tighten a 4-40 nut onto the stud to clamp it to the bracket. Thread another 4-40 nut on the stud. Use this nut to adjust the tilt on the preamp card when it is installed on the camera. Lock the card to the stud with another 4-40 nut over the top of the card.

3. Wiring Problems

We suspect that neophytes make fewer wiring errors than electronics professionals because they make full use of the test software included with the Cookbook.

• Miswiring a connector or a card is certain to cause a test failure of the camera. Miswiring generally occurs because a jumper wire is forgotten, a wire was soldered into the wrong card solder pad, or the connector was viewed from the wrong side when probing pins for a test. After passing low-levels tests, the card may fail at a higher test level -- but when it is retested to look for the problem, the socket pins or card pad are once again probed at the wrong point. To correct this problem, make sure you view the card or the socket from the correct side. Pin references have been marked throughout the book and in the test software in an effort to identify the pin and pad locations. Miswiring is one of the most common causes of test failure right up to failure at final checkout.

4. Miscellaneous Problems

4.1 Component Problems

Component problems with the Cookbook camera have been mostly due to confusion with respect to the device markings and which package type to use.

• The 2N2907 transistor, the 2N3904 transistor, the LM7912 regulator, the LM78L12 regulator and the LM336Z-2.5 voltage reference should be purchased as plastic packaged parts. The silk screen on the cards shows the correct orientation for these parts when they are molded plastic packages. The use of metal packaged parts may result in the incorrect orientation of the part.
• Some tantalum capacitors do not have a + (plus sign) marked on the package, but the plus lead is usually marked in some fashion. Look for a longer lead; it is the plus lead. A stripe painted on one side of the capacitor usually indicates the plus lead. Color coded capacitors may have a dot in the center. With the leads pointing down and the dot facing you, the plus lead is the one on the right.
• Although the book did not recommend a specific type of resistor, you will find that carbon film resistors are a better deal than carbon composition resistors. Film resistors have a material deposited onto a mandrel and the film is cut in a spiral pattern to trim the resistor. Carbon composition resistors have a central plug of carbon material with a molded body. After molding, they are tested and binned into the values. Carbon compositions drift more with humidity and temperature than the film types.

4.2 Component Failure

Component failures have been relatively rare. If you have trouble, suspect the component last. Almost all component failures are secondary, that is they failed from an overstress.

• A short in the wiring of the CCD chip to ground, a short in the J2/J3 socket to ground or incorrect installation of the J2/J3 plug into the socket, can cause the output resistors on the DS0026 drivers and the DS0026 drivers to heat up and fail.
• The DS0026 drivers can overheat and fail if the -9.5V power supply starts to oscillate (install those 1000uF capacitors).
• Only one failure of a CCD chip is known to date. The chip was bad as received and it was replaced by the supplier. Nevertheless, we continue to recommend that you apply tight ESD procedures.
• Tantalum capacitors installed backwards are failures just waiting to happen. They always fail short.
• The 7805 voltage regulator can get very hot when testing the cards. Attach a temporary heat sink to the metal tab. If the regulator gets to hot, it normally will not fail but shut down. This can cause confusion when you are testing. Be careful if this happens because the tab on the regulator will be about 125 degrees Celsius. Don't burn your fingers!

4.3 Pump Problems

• Pump problems have been short life, running hot, or not running at all. The Autozone universal replacement windshield washer pump selected for this project turns out to be one of the few pumps which is reliable for this application. It also operates satisfactory on the 6V provided by power supply.
• Low-voltage pumps are available through chemical equipment suppliers. These pumps vary in design. Look for a self-priming type. The maximum pumping head height is not very critical since the return side of the hose counters gravity once the cooling system is filled. Watch out for too much pumping pressure. Keep the inlet pressure to the camera low as not to blow off a hose or to create a leak in the heat exchanger. A flow of 100 to 250 milliliters per minute is entirely satisfactory.
• Richard Berry has switched to an open-loop water cooling system and uses a submersible 115VAC Little Giant fountain pump costing about $40. The pump is virtual silent, made for nonstop running, and built to run immersed in water. Nevertheless, anyone using such any 115VAC system should be aware of potential hazards of using house current in an outdoor setting.

4.4 Variations in Design

Not following directions or "doing it someone else's way" can be a problem, especially if you have not built a Cookbook camera before.

• We encourage experimenting and improving the camera design, but for your own peace of mind, we ask that you not change the design before you have succeeded in making the camera work. A lot of effort went into providing assembly, processes, and test information in the CCD Camera Cookbook. The methods are almost foolproof if followed. Perform all of the electrical tests provided in the CARD and PREAMP test programs.
• If you change the design of your camera and you have problems, return the camera to the original book design unless you understand what is causing the failure. If someone suggests a modification or improvement, find out whether they know why the improvement is better and whether it affects other aspects of the camera such as mechanical fit. They should also have an in-depth understanding of the camera and be able to provide help beyond the suggestion.
• Several independently published versions of software drivers for the Cookbook cameras available on the Internet. We offer no assistance with respect to the compatibility of these software items with the Cookbook cameras. The AP211 and 211Plus software has also been reported to operate with at least one commercially available 211 type of camera. Again we offer no assistance to this claim. Verification of independent product compatibility is an extensive task which left to the proprietors of these products.

5. Power Supply Problems

The power supplies on the interface card are a weak point in the Cookbook Camera design because the regulators can oscillate. The installation of two 1000uF electrolytic capacitors (refer to section C.3 on page 160) elimimates this problem.

• Power supply oscillation is often load dependent and may occur only when the regulator must source more than some level of current. This behavior causes great confusion since the regulators will test out normal during initial testing, but the voltage measures incorrectly during later test sequences.
• The regulators can be stabilized on the Interface Card by adding two 1000uF electrolytic capacitors as described on page 160 of The CCD Camera Cookbook. Add these capacitors to the card before performing the regulator voltage test.
• It has been a common problem that one of the power supplies voltages on the interface card or the preamp card does not have the correct voltage level. This is usually noticed at higher level testing when chips have been added to the cards. The +2.4V and -9.5V supplies are the culprits. Usually the cause is that these supplies were not trimmed at lower level testing. Both of these supplies require the addition of a resistor to trim the voltage to the proper value. Press ? or H to check what the help file has to say about trimming these supplies.
• The Peltier power supply has been occasionally reported to exhibit abnormal voltage outputs and oscillations. The Peltier supply should not normally oscillate and care was taken in the design to circumvent oscillatory behavior. The following layout rules are suggested: 1) Keep the wire from C2+ to the LM317 Input as short as possible. Make sure the connection is to the C2 capacitor and not to the collector connected end of the R1 and R2 dropping resistors. 2) The 0.1uF bypass capacitors on the LM317 input and output should be connected as close to these pins as possible. The grounded ends of these capacitors should form a tie point for the 1500 ohm resistor and the 1K potentiometer. In general, keep all component lead lengths to the LM317 short. 3) Route all ground wires for the Peltier supply to a common point. The nylon binding post is a good point. Try to keep them as short as possible. 4) Keep the wires from the 1K potentiometer short. If long wires are unavoidable, you can twist them together to make a twisted pair which has more shielding than straight wires. 5) If the supply oscillates, try adding a few ferrite beads to the wire coming from the emitter of the 2N3055 transistor. If you can't find ferrite beads, Radioshack sells snap-together ferrite cores for radio frequency interference (RFI) elimination. Try a few turns of the wire leading from the emitter around one of these cores. An alternative oscillation damping method is to add a series resistor (carbon film) of 5 to 10 ohms between the regulator output and the base of the 2N3055 transistor.
• The plus and minus 15V supplies are unregulated DC power supplies. This means that the output voltage will vary with the voltage at the 115VAC outlet. If the outlet voltage drops too low, the +15V output will go below the drop-out-point of the 78L12 voltage regulator. The camera will still work but the image will show banding of around a hundred counts caused by the power-line frequency modulating the 12V on the card. Paralleling two transformers as suggested in the book will raise the voltage level to operate most cameras. Also using a 25.2V center tapped transformer rated at 2A (Radioshack #273-1512) instead of two 450mA transformers has been observed to get a slightly higher output voltage. If you live in a remote area, you may have low line voltage and the paralleling of another transformer may not increase the voltage enough. A constant voltage transformer made for computer line conditioning is suggested for the low-line-voltage problem areas.
• Fuse blowing has been reported on the Peltier supply. The current draw at the 115VAC side is close to the 3/4 amperes that the fuse is rated for. This only occurs when the Peltier voltage is at the maximum level. If the 3/4 amp fuse blows occasionally when the Peltier supply is set to maximum, it is probably not a problem with the supply and you can substitute a 1 amp fuse. If the 3/4 amp fuse blows when the Peltier supply is turned on or at lower voltage settings, there is something in the supply drawing too much current. (In that case, fix the problem and do not install a 1 amp fuse).

6. Measurement Related Problems

Reading digital voltmeters has been a cause of much confusion when trying to interpret whether the card or a power supply actually failed the test.

• Meters vary considerably in design, and if you have little experience in electronics the meter itself can be a source of error. We recommend testing with modern digital volt meters (DVM). They are inexpensive (on sale, they cost about $20) and when you are finished with the camera, the meter is handy for checking batteries and light bulbs.
• Not all meters measure AC or alternating current voltage the same way. A true AC voltage measurement decouples the DC or direct current from the measurement, that is, when set on AC volts, the meter will read zero volts when it is attached to a battery. This is what is needed to measure the ripple voltage on the plus and minus 15VDC supplies. Test your meter on a battery, say a 9V, with the meter set to AC volts. Check it for both polarities on the probes, that is swap the terminals that the test leads are probing. If you don't get zero volts, the meter can't measure the ripple on your power supply correctly. If this is so, place a 0.1uF capacitor in series with one of the test leads. The capacitor will block the DC and you can measure the ripple. Use a capacitor that has a rated voltage greater than the voltage you are measuring. Do not use the capacitor for measuring the power line voltage. This voltage has no DC on it and your meter will read correctly on the AC setting. Line voltage is potentially lethal and it is best to keep your hands on the insulated part of the probes when measuring high voltage.
• Electronic components and signal levels in electronic systems can vary in magnitude by a million times or more. The units used to express the magnitude can be confusing. Many of the test sequences ask for a specific value. A resistance of 200 ohms might be called out but your meter reads 0.200K ohms. They are actually the same value. The K (Kilo) means to multiply by 1000, so 0.200 times 1000 is 200 ohm. Other multipliers are used in electronics. Here are some common ones: m (milli) is times 0.001, u (micro) is times 0.000001, n (nano) is times 0.000000001, p (pico) is times 0.000000000001, K (Kilo) is times 1000, and M (Mega) is times 1,000,000. Another example is 450mV which is 0.450 volts.
• Sometimes a test reading is just a little out of what the limits say it should be. What do you do? Well, it depends on what you're measuring. The test measurement limits were chosen with margins as wide as possible to keep good components from causing test failures. If you are measuring a power supply, it may need adjustment to trim it to the correct value. Always press ? or H to check what the help file has to say. If it is a resistance value that is just out, it likely to be OK to leave the resistor alone. If you are using carbon composition resistors (not carbon film types) they vary in value with the relative humidity. It is not uncommon for a 5% tolerance carbon composition resistor to be plus or minus 8% from the nominal value. If you are using this type of resistor, measure the value before you install it, and if it is too far off, use one that is within the specifications. To find the normal range of resistance, multiply the nominal value by 0.95 and by 1.05. The resistance should read in between these two values.

7. Software Related Problems

Yes, the original Cookbook camera has software bugs and limitations. However, these problems are minor and they do not prevent you from obtaining good images with your camera. Many of the problems have been fixed with the new 245Plus and 211Plus software, and the new software is generally friendlier and easier to use. Order the upgrade package from Willmann-Bell.

• The CARD and PREAMP test software shipped with the book, does not automatically select the printer port address. The default printer address set by the software will work for most computers. If you have a problem, you will see it when running the Computer to Interface Card Logic Level Test, test-2. You must then select Show Printer Port Address from the main menu then choose the port by pressing the number corresponding to your port (if you only have one printer port the correct address will be next to 1). Exit using this port. If you had to change the printer port address, you will need to change it every time you start the CARD and PREAMP program.
• The 245 camera acquisition software tries to do a lot within the limits of conventional memory. There is not enough room in conventional memory to handle a 378-wide format image and a dark frame plus the quarter frame image and its dark frame. The program reuses the quarter frame space for the bottom third of the 378 wide arrays. Using find/focus mode will write over the bottom third of your 378 wide image. Integrating a full frame 378 wide image will write over the find/focus quarter frame image. Always save your 378-wide image and dark frame before you search out a new target in the find/focus mode.
• AP245 does not clear the image frame when you switch between the 378 and 252 wide modes. You may observe a distorted image when displaying after a mode change. The new 245Plus software clears the frames to avoid confusion.
• AP245 will not save a 378-wide image correctly in the multiple acquisition mode. With AP245, use only the 252-wide modes for multiple acquisitions. 245Plus corrects this problem and does multiple imaging in 378-wide mode.
• Slow motion controls for guiding can be operated through the use of a serial to parallel converter (UART chip). The book circuit is based on an article in Telescope Making #46, Winter 1990/91, "Build a PC Controlled Data Acquisition and Error Correction System for Synchronous Drives" written by John Munger and myself. This is recommended reading if you can find a copy. Although simplified, the book circuit provides the same function, that is, open-loop error correction and control of the synchronous motor drive by keeping track of the 60Hz clock cycles to the motor. The older circuit, which I use a modified version of, generates a clock signal using an inverted control bit. It was natural to make the default state for the software invert this bit to match the circuit operation. An inverter gate is not needed for operation of the book circuit. From the Options Menu, select seRial then press F to toggle the not-bit0 to bit0, and then exit and save the setup.

8. Computer Related Problems

The need for lots of conventional memory can be a problem when running the 245 acquisition software.

• The AP245 and 245Plus need about 555K of available conventional memory and the 211 programs need about 325K. Where does the need for all of this memory come from? It is the image frame and the dark frame that use up most of the memory. Each image pixel is stored as an integer which uses two bytes of memory. The largest image size is 378 by 242 pixels. There are two image frames and the memory adds up to 358K (1024 bytes per K). The remaining 197K is the program, variables and stack.
• The amount of memory available for running a program will depend on the version of DOS you are running, drivers that are loaded into memory, and any terminate and stay resident software (TSRs) that have been loaded. Versions of DOS 3.x are small enough to fit into conventional memory along with the 245 program. Higher versions like 5.x and 6.x are larger and they must be loaded into high memory. This requires that you have extended memory (enough is present with a 286 machine and 1 Meg of memory).
• I run DOS 6.22 on 286 machine and I have 561K available program space (and 570K if I don't install the mouse driver). My CONFIG.SYS file consists of the code below. Use it for reference but check with your DOS User's Guide to determine what is best for your version of DOS. You may also need to install drivers for some of your hardware. I prefer to make a bootable floppy disk with the CONFIG.SYS file that I would like to test. This way the computer will still boot from the hard drive if I goof up the file.

DEVICE=C:\DOS\SETVER.EXE
FILES=20
SHELL=C:\DOS\COMMAND.COM C:\DOS\ /p
DEVICE=C:\DOS\HIMEM.SYS
DOS=HIGH
BUFFERS=10,0

• Check available memory with the MEM command in DOS Ver 5.x and 6.x. Use CHKDISK in DOS 3.x to find out how much memory is available. Take the "bytes free" value and divide by 1024 to get the number of K bytes available.
• The delays for the operational amplifier settling time and the Analog to Digital Converter (ADC) conversion time are generated using delay loops in the software. It would have been possible to make the delay calculation automatic in the software but this would not allow optimization of the acquisition speed. Even between the same types ADCs and op amps, the times can vary. The use of a faster settling op amp has already been suggested and ADCs are getting faster. For example, Sipex has a part that rated to convert between 4uS and 8uS (variation depends on the internal clock frequency) as opposed to the 10uS for the Analog devices part. The delay count has a maximum of 99 on the AP245 software. This delay time may not be long enough for some processors (I use a delay of 70 for a 33mhz 486 and a delay of 10 for a 12MHZ 286). You can set the count to zero which loops the delay for 256 counts. The new 245Plus and 211Plus software delay counts can be set as high as 255 which should cover most computers. However, I recommend buying an old 286, with a VGA display, for field use.
• Hardware keys are used to protect some software. Although they don't interfere with normal printer operations, they can cause havoc with the Cookbook camera' s operation. Remove the key from the parallel port before connecting the camera to your computer.

9. Troubleshooting Final Checkout

The Reference and Reset levels, that are displayed on the Main Menu page of the acquisition programs, are key values in determining if your camera is working. It is the first functional test of the system. The software does not need the timing delays to be set for these values to read correctly.

The REF and RESET levels normally change in value rapidly. Jumps of several counts are normal on the reference level and the reset level is more noisy with jumps often over ten counts. Jumps in the hundreds is abnormal. In the 211, the reset level is higher in value than the reference level at about 400 counts, and in the 245, by 800 counts. This is the most suspenseful test of the project and it is very likely that you will look up and see they do not meet your expectations. Relax, all is probably well except for a minor adjustment. Here are a few tips on troubleshooting.

• Reset and Ref. both read 4095 or Reset reads 4095
  Cover the CCD chip so that no light can reach it. Adjust the P1 potentiometer on the preamp card. It is a multi-turn device and it may need several turns in one direction or the other.
• Reset and Ref. both read either 3276 or 1092 or 1638 or 3622 or 4095 or a fixed non-changing number other than zero.
  Check that the ribbon cable is plugged into the correct computer port on one end and the interface box on the other. Check that the plus and minus 15V is present on the power supply and that it is plugged into the interface box.
• Reset and Ref. both read 0000
  Adjust the P1 potentiometer on the preamp card. It is a multi-turn device and it may need several turns in one direction or the other. Check that the JS cable is plugged into the socket on the interface card and with the correct polarity (ground pin to ground pin and signal pin to signal pin). A short between the cable shield and center wire can also cause this problem. It may be due to frayed wires or a pinch in the cable where it enters the box.
• Reset reads 0000 but Ref. is not zero
  Adjust potentiometer P1 to bring the values higher. It is normal for the readings to drift towards zero as the camera warms up and it may need some additional adjustment.
• Reset level is not much different than Ref. level
  If the two numbers do not have the same exact value, this condition can indicate a problem with the serial gate clocking on the chip. Look for incorrect values of R31, R32, R38, R49 and C38. Retest the serial clock driver high and low levels. Look for miswired serial clock lines to the preamp card and to the CCD chip.
• Both Reset and Ref. change in big jumps of hundreds or thousands as P1 is slowly adjusted.
  It is normal for the levels to increase rapidly as you turn P1, the levels should vary linearly with the turning of the potentiometer. A jump to a high value then back down or visa versa without changing direction of the potentiometer indicates the data the computer is receiving has bits missing or transposed. Check that your computer cable is fully plugged into computer and the interface box. A connection may have gone open on one of the 74LS154 multiplexers. You must check the path from the U5 ADC pins 15-20 to U3 and U4. The paths from U4 to U3 and the path from U3 out to pins 10-13 on the DB25 connector. Path means to look at the solder joints, look at the traces, check component seating in the socket and check for miswiring to the connector.
• Reset and Ref. jump suddenly as P1 is adjusted AND you have a Sipex ADC installed.
  If the preamp card offset is set more negative than -3V, the reset and ref values will jump between 1024 and 2048 (or values close to this) when P1 is adjusted. Check Pin 13 on the U5 ADC chip to find the preamp output voltage. Click here for more on the Sipex ADC.

10. First Light Imaging Problems

The first step to getting an image from the camera is to select the delay values for the Analog to Digital Conversion time and the Op Amp settling time. Start with both the op amp and ADC delays equal in value. A good way to check if you have selected the delays properly, is to integrate a full frame. Set the integration time to 0.1 seconds then start the integration and your stop watch at the same time. At the last beep, stop your watch. A Internal binned 252 wide image should take about 2.5 to 3 seconds to read out. A 211 camera Full Frame image should take about 1.3 to 1.5 seconds to read out. Saturation/blooming of the image occurs when the light striking the CCD chip generates charge which exceeds the well capacity. The image will be all white. Obvious solutions are to shorten the integration time or to decrease the light level. The chip is very sensitive to light and you may need a nearly dark room when imaging with a simple lens. If you happen to wire your Peltier incorrectly, the cooler will heat the CCD chip instead of cooling it. This can also cause saturation of the image by generating a much higher thermal bias.

• Hot spots in the center of the image can be reduced by placing a baffle right on top of the camera housing optical window. I use a lint free black paper that is cut into a circular disk as large as the threaded adapter (1.25" for my camera). A rectangular hole is cut into the disk that is slightly larger than the image area of the CCD chip. When imaging with a simple lens or even a telescope, the baffle removes the central hot spot.
• Focusing an object can be difficult if your simple lens system does not have a long enough focal length for focusing at infinity. A short focal length lens will act as a macro type of system and the focus may only be a few inches away from the camera. A card with very fine news print makes a good target to focus on. Get into the find/focus menu and turn the auto scaling mode off. Next set the high stretch to 4095 and low stretch to 0000 from SET OPTIONS. Also set the time to 0.5 seconds for the 211 and about .05 seconds for the 245. Exit back to the find/focus menu and adjust the room lighting to get a mid level gray on the screen (the histogram in SET OPTIONS will also help to set the exposure up). Move your card around until a blurred image is obtained (depending on the tilt of the lens, it could also be way off of axis). Now make the card stationary and slowly adjust its position with respect to the camera. Even at a perfect focus there will be some aliasing or edge artifacts in the image of the print.
• Banding in an image, if present, should only consist of a few ADC levels. Banding of tens or hundreds of levels indicates a problem. Banding is generally caused by one of three problems. 1) The delays have not been properly set. Increase the values. 2) The plus 15V voltage is low and the plus 12V regulator is dropping out. Look in the power supply problems section above. 3) The -9.5V regulator is oscillating. Make sure you added those 1000uF capacitors described on page 160 of the book.
• A single line or compressed group of lines that are bright near the top of the image without any image formation in the rest of the frame indicates your image area gate, storage area gate or transfer register gate is not clocking properly. Go back to the preamp clock driver test and check that these clocks are working properly. A miswiring of the chip can also cause this problem.
• If you can image with the camera but the images have a limited gray scale range or the images are not smooth gray tones, the signals from the ADC might be corrupted. Go to the Z-Focus mode. Set the time to integrate a dark frame that builds up to greater than half of the range (2500 or so). The display should show a random variation of levels. Large discrete jumps between levels confirms the data from the ADC is not being read out correctly. A connection may have gone open on one of the 74LS154 multiplexers. You must check the path from the U5 ADC pins 15-20 to U3 and U4. The paths from U4 to U3 and the path from U3 out to pins 10-13 on the DB25 connector. Path means to look at the solder joints, look at the traces, check component seating in the socket and check for miswiring to the connector.
• If your camera was working then stopped working once it was buttoned up, look for a mechanical type of problem. Sockets pulling loose, wires shorting, wires breaking, covers shorting to component leads, etc.

We trust that the information provided here will aid you in troubleshooting most of the problems you may encounter in building your Cookbook camera. We do want to emphasize that most cameras work right the first time, so don't let this document "throw" you if you encounter it here on our Web Page before you have built your own Cookbook camera.

We also want to emphasize several other important things. First, if your camera stops working, the CCD is okay. We have yet to see a blown CCD. Second, the camera is not a living thing, and if something happens, it is not dead. You can find the problem, fix it, and your camera will be fine again. Third, most problems result from simple things like faulty soldering or poor connections. Look for the simple cause first.

In the event that careful reading and analysis do not reveal the cause of a problem and allow you to correct it, then browse a little deeper for directions on asking for help. It takes a lot of background information to troubleshoot a camera, and that adds up to a lot of phone hours, which is why we ask you to download our form and fill it out. That information helps us to help you more efficiently.

--Veikko A. Kanto

Return to the Text DataBase Page

Return to the Cookbook Home Page