Audine doesn’t need to be adjusted with an oscilloscope : the following information apply essentially to the specialists who want to observe how their camera internally operates.

The timing diagrams which illustrate this chapter show the look of the signals at some strategic points of the electronic circuit. To produce them, it is necessary to launch special commands from the console mode of PISCO. Two commands are used : TEST and TEST2 :
 

Command TEST [NUMBER]

This command executes NUMBER transfer cycles from the image area to the horizontal register. The following sequence of instructions describes each basic cycle (for further explanations, click here):

_asm
      {
      mov dx,base
      mov al,11111011b
      out dx,al
      out dx,al
      out dx,al
      out dx,al
      out dx,al
      out dx,al
      out dx,al
      out dx,al
      mov al,11111010b
      out dx,al
      out dx,al
      out dx,al
      out dx,al
      out dx,al
      out dx,al
      out dx,al
      out dx,al
      mov al,11111001b
      out dx,al
      out dx,al
      out dx,al
      out dx,al
      out dx,al
      out dx,al
      out dx,al
      out dx,al
      mov al,11111010b
      out dx,al
      out dx,al
      out dx,al
      out dx,al
      out dx,al
      out dx,al
      out dx,al
      out dx,al
      mov al,11111011b
      out dx,al
      out dx,al
      out dx,al
      out dx,al
      out dx,al
      out dx,al
      out dx,al
      out dx,al
      }

The code sets the clocks V1 and V2 in action, and performs the  charge transfer from the image area. As the code is very short, it complies very quickly too. So to observe the clocks V1 and V2 with the oscilloscope a long time enough, you must give the argument NUMBER a big value. Try for instance :

TEST 10000
 

This command executes NUMBER cycles of quick reading of the camera. The following sequence of instructions describes each basic cycle :

for (i=0;i<521;i++)
   {
   zi_zh(base);
   for (j=0;j<790;j++)
      {
      _asm
         {
         mov dx,base
         mov al,11110111b
         out dx,al
         mov al,11111111b
         out dx,al
         out dx,al
         out dx,al
         mov al,11111011b
         out dx,al
         out dx,al
         out dx,al
         }
      }
   }

This code activates all the clocks that command the CCD (clocks V1, V2, H1, H2, R). The TEST2 command is particularly aimed to examine the look of the video signal at the pixel level. For instance, to observe 3 successive frames, you will execute :

TEST2 3

Nota : It is very dangerous to observe the video signal directly at the CCD output (on pin 2). The least shortcut between this pin and its neighbors can destroy the CCD internal output amplifier. Record the video signal in these conditions only with a special care and if it is really necessary. By measuring the video signal after the coupling capacitor C1 , as recommended before, you will protect the output amplifier of the CCD.

Video signal on the pin 3 of U6 (input of the amplifying chain). The CCD is lighted so as the pixels are saturated.  The first timing diagram (canal 1) is a synchronization signal, taken on the pin 1 of U4 (MAX333A). It is the command signal P3, which produces the CCD clocks H1 and H2.
The second timing diagram (canal 2) is the actual video signal. It is easy to distinguish the reset stage, the reference stage and the video stage (this one with the lowest level). The little arrow marked 2, on the left side of the picture, indicates the ground level. Let’s notice that the video signal average level is close to the ground level after the coupling capacitor C1, and that the pixel level potential is negative.

These timing diagrams are equivalent to the previous, but this time the CCD is plunged into darkness. The video stage level comes closer to  the reference stage level, but it is still negative in comparison with the ground.

The following picture takes a close-up on the video signal according to where it is measured in the amplifying chain. The TEST2 command is still used here.

Video signal on the pin 3 of U6, that is to say before the amplification. The illumination conditions of the detector make the signal fill half the useful dynamic. By examining the code of the TEST2 routine, you will notice that the reset stage duration is equivalent to an OUT assembler instruction, whereas the reference and video stage duration is equivalent to 3 OUT instructions (the horizontal scale is 2µs/cm).

Signal on the pin 1 of the U6, after an amplification of about 15. The limitation of  the amplifier “gain-band product” is at the origin of the clock edge smoothing. Let’s notice that the reference and video stages are established after about 2 µs. So it is necessary to wait at least this delay to digitize the video stage after the rising edge of the command signal P3.

Signal on the pin 7 of U6. The look of this signal is very close to the previous one. Indeed, the clamping circuit gain is unitary, and the bandpass is maximum. Let’s notice that the routine TEST2 does not generate the clamping signal : the clamping capacitor C25 is then shown like any other coupling capacitor. The signal still has an average level close to zero.

Video signal on the pin 8 of U6. It settles at the output of the last amplifier stage, which is a unity gain inverter. The video stage has now a positive level compared to the ground.

The following picture shows the video signal during the CCD reading in binning 1x1. To observe it, you just have to launch a 0 second exposure with the PISCO graphic interface, or to execute the command GLP 0 from the console and wait for the READING status to appear.

Video signal on the pin 7 of U6 when the clamping is active. The synchronizing signal is the P3. Let’s check that the clamping had brought the reference stage at the ground level (the arrow marked 2 on the left of the frame). You can measure on this figure the point frequency of the video signal when Audine acquires an image. It is about 26 kHz. The lengthening of the video stage compared to the other stages is partly due to the conversion time of the CAN and to the transfer towards the computer.

Video signal observed in the same conditions as before but this time in  2x2 binning (Notice the look of the clock P3). The charge storing effect is well visible in the video signal (the video stage disconnection of the video stage which amplitude is twice at the end).

Same thing as for the previous picture but with a 4x4 binning.

The following timing diagrams show the some clock synchronization.

Above, the P3 clock command (equivalent, amplitude more or less, to the clock H1 applied to the pin 7 of the CCD). Below, the clock R measured on the pin 5 of the CCD. The command TEST2 has been used to produce this timing diagram.

The signal R/C (beginning of the CAN conversion) taken on the pin 24 of U5 (above) compared with the video signal (below). Notice that it is necessary to wait about 6 ms before digitizing the video signal, to let it enough time to be well established.

Position of the clamping signal compared with the reference stage. The top is taken on the pin 11 of U4 (signal P5), arrives after a bit more than 7 ms to be sure that the reference stage is correctly established. The video signal is taken on the pin 7 of U6.

By using the PISCO TEST2 command and the signal P1 as a synchronization, you can observe the video signal along a whole image line. The video signal (below) is measured on the pin 7of U6 and is synchronized by the signal P1 (above), taken on the pin 1 of U5.The rounded aspect of the signal betrays a non uniform illumination of the CCD during the integration period (a piece of black adhesive tape was stuck on the CCD window, which caused a more important illumination on the CCD sides than on its center).

Detail of the previous figure, around  the end of the reading of a video line and the beginning of the reading of the next one. Let’s notice the double impulsion of the signal P1 which is equivalent to the clock V1 (amplitude more or less). The hidden pixels, and non electrically connected,  are very well visible at the beginning or at the end of the line.

This timing diagram shows the video signal (pin 7 of U6) at the beginning of the image line, but on a larger area than on the previous view. In this case, the binning is 1x1. To do that, start an exposure in a pronounced half-light and examine the video signal  when the READING status appears on the screen (you have at this moment about fifteen seconds to analyze it). The end of the previous line is characterized by a high cadence reading of the horizontal register, without digitizing. If you glance at the CCD low level reading functions, you will notice that this phase corresponds to a cleaning cycle of the horizontal register, before this one receives the content of the next image line. At the beginning of the read line, the first pixels are quickly eliminated because we are not interested in them. The actual digitization of the image points is realized just afterwards.

Here is the video signal at the beginning of the line whereas the cleaning cycles are in progress (it happens when the RAZ status appears on the screen). Notice that 4 image lines are accumulated in the horizontal register before this one is read, in order to accelerate the proceedings.

The following figures show the look of the clocks edges applied to the CCD.

Rising time of the clock H1 (below) measured on the pin 7 of the CCD (of course this one is at its place). The rising time is about 90 ns. The synchronization signal is the command P3 (pin 1 of U4). The TEST2 command was used to produce this timing diagram.

Falling edge of the clock H2, observed on the pin 8 of the CCD (the falling time value is 80 ns). The synchronization signal is the same as on the previous figure : the crossing position of clocks H1 and H2 can be analyzed. It appears approximately at 50% of the clock amplitude (this is the optimal crossing point). Let’s notice the delay of H2 in comparison with P3, due to the propagation of the signals in the MAX333A.

The rising time of the clock V1 (pins 22, 21, 16 and 15 of the CCD) is about 1 µs. The synchronization signal is the signal P1, taken on the pin 1 of U5. The TEST command is used to produce this timing diagram.

Falling time of the clock V2 ( pins 20, 19, 18, 17 of the CCD). The synchronization signal is P1 (the same as for the previous figure). Notice how the clocks V1 and V2 cross compared with the previous figure.

Above, the signal C1:2 (TTL type) from the PC whereas the link cable PC/camera is 5 m long. Below, the same signal, but after its transit through the trigger 74HCT14 (taken on the pin 4 of U1). The trigger returned stiff edges for the command signals coming from the PC printer port.