PCI-1428 and Dalsa Piranha2 IK linescan camera

[John Birmingham]

When I enable a test pattern mode from the camera, instead of seeing a linear 0-255 ramp per tap output, I see this corrupted with some sections of data having either 64 added or subtracted. This also happens when real image data is used. This is independent of whether MAX or my own software is in use. Does this indicate a problem with the *.icd file, 1428 memory management or what ? One icd file parameter that changes the way this problem looks is the DataFPGA value, however I never see the correct test pattern. There is a basic raw video line dump mode with this camera and there is no +/- 64 problem apparent with the raw data from the camera seen through such direct serial read, just with same as seen through 1428 higher level
functions like Snap or Grab. Any help gratefully recieved.

[Mike H2]

Sir,

I talked to the engineer who wrote that camera file for that camera. The camera file was written relatively recently, and it does work. There are two possible problems that we can think of. One, the parellel to camera link converter chip on the camera my have one line that is not working. Or it may be the chip on the board. Can you send me some screen shots of what you are seeing? Is it consistant?

[John Birmingham]

Mike H,

Attached find two screenshots of single line grab, one with the test pattern, other with a real optical source, you can see the sharp +/- 64 stuff coming in at various places. The exact pattern depends a bit on line rate but it is generally consistent in that I cannot eliminate it !

[Mike H2]

Sir,

We cannot tell whether it is the fault of the board or the camera. What we can tell you is that we have not seen anything like this before. We know that it is not just plain noise or a problem with the cable since it is off by a consistant 64 pixels. I would recommend either RMAing the camera or the board. I would start with the camera since we have not seen this problem before.

[John Birmingham]

Mike H,

Thanks for your reply. One thing I forgot to mention is that I occasionally see an Imaq-1428 board diagnostics failure in MAX, on the lines of Critical Failure re Memory Writing, details say something like "interrupt X asserted incorrectly before interrupt Y". However this often disappears when running the diagnostics a second time and none of this really alters what I see as regards data from camera when reading normally via imgGrabArea or imgSnap. Also, the Piranha2 has a low level serial write of current raw video line mode which I access via imgSerialRead function, this always looks OK with no +/- 64 jumps at all. However I don't know whether this is a good enough check of the CameraLink hardware on the camera end or not.