FMC Projects » FMC PCIe Carrier (PFC)

CERN's PFC card and the Gennum to Wishbone core is featured on Gennum's GN4124 Design blog

A first DMA transfer to the PFC DDR3 and back to the host has been done this morning.
The DDR3 interface seems to work well, but more exhaustive test of the DDR3 interface are still to be done.

The communication chain of writing and reading Wishbone registers inside the Xilinx via the Gennum PCIe bus interface is working.

Thanks everybody for your support, and specially Tim from Gennum for his tips. The GN4124 is now listed under the PCI modules. We have no problems accessing the configuration space, everything seems to be running really smoothly. Next step is to program the GN4124 flash and gain communication the Spartan6.

The first 3 prototype arrived this afternoon from the assembly office. The boards powered up correctly from a desktop power supply. QDRII, Spartan6 and GN4124 were easily recognized by impact running on a Windows machine. The linux version of impact misteriously refused to detect the chain. To be continued....

The PCI Express front-panels will be shipped from the US on 15 December.

The component assembly of 2 Standard Negative PFC and 1 Standard PFC will start tomorrow 4th November. The estimated delivery delay is aproximately 1 week.

The PCIe front-panel with the cut-out for the FMC front (and fibre-optic connector) has been ordered. It will be a one-piece design. Another, more conventional 3-piece design will also be ordered soon.

The PCB layout of the FMC PCIe carrier has been finalised and the production of four PCBs has been requested. Three of them will be assembled and one will be kept empty for being able to measure for connectivity, if needed.

A first idea of how the board will look like can be seen from the image of the component placement (see the wiki page). The board will be implemented as a full-size (long) PCIe card that will ease routing. As the usage will be in industrial PICMG PCs, size is not a problem.