May 10, 2001
E. Pedretti

Present State of the PICNIC Camera FPGA Readout
===============================================

1. Quadrant readout:

It seems to work well now. The image is free of stripes or other
visible artificial feature present before.

The circuit now uses three separate state machines working in
parallel, to avoid the race conditions happening in the single, big
state machine used before. Most of the problems in this circuit seemed
to be caused by internal delays: since the pin assignment is fixed on
the Tecnobox card the compiler is not free to route the signals
chosing the shortest path.  this might have been the cause of the many
problems not seen in simulations.

I haven't tested the part which takes the difference of the reset
frame and the integrated frame yet since I'm working to the single
pixels' acquisition.


2. Single pixels readout:

I've got the clocking pattern right in the simulation. I need to
implement the sums for the n-reads and the n-loops part. I also need
to modify the state machine which controls the PCI bus to avoid
hunging up the CPU.  testing on the Picnic camera will follow.