RE: [oc] Modular FPGA board

[John Dalton <john.dalton@bigfoot.com>]

> It depends on the frequency (33 or 66MHz) and the bus width (32 or 64
> bits).

I would say the FPGA size should be decided on 'worst case'.  Presumably
64 bits at 66MHz.  I note the OCRP-2 schematics indicate a 150k gate part.

> It depends on the bus width. For safety, 100 pins.

I presume 100 pins is for a 64 bit bus?

> There is not too many different FPGAs in TQFP packet, then the common
> footprint should be based on a BGA packet. But if your plan is to build
> the PCB at home, BGA is not a good alternative. Or do you have equipment
> do solder a BGA chip at home?

Most of the XIlinx families (to 800k gates) appear to be
available in some form of flat pack. My intention would be
to use a flat pack for version 1, as they are easier to route
and to install.  The biggest limitation is the restricted number of
pins.  Later versions can use more sophisticated packages.
After all, one of the ideas of modularity is upgradability.

> We can design the board with a PCI edge connector and with all PCI tracks
> going from the PCI connector just to the FPGA that will be the PCI core.
> Then, the remaining pins of this FPGA would be connected by the board to
> the SIMMs sockets. Our specific hardware would be another FPGA soldered in
> an extended SIMM board, and maybe, for some applications, one of the
> sockets would accept a real SIMM memory module. Is this flexibility good
> enough?

Similar to what I am thinking.  I'm not sure the connection between the main board
and the PCI board can be via a SIMM.  Possibly more pins are required than a SIMM
can provide.  Using a SIMM may also rule out fitting the whole assembly in a PCI slot.
Maybe the connection between the PCI board and main board can even be via a PCI
connector (depending on price and availability, early days in the design yet).

> Could you please explain what is this Riston?

It is a system of blank PCBs with a photoresist applied.  A
transparency of the PCB is layed over the board and exposed to
UV light.  The board is then developed and etched.  I haven't
actually used it that much so I don't have a feeling for the maximum
resolution it can do.

> PGA and BGA packets requires at least a 3-layer board,

I'm thinking of a flat pack, with single layer routing.  The idea is to radiate tracks
out from the edges of the FPGA with no crossovers on the PCB.  All crossing
over of signals is to be done inside the FPGA.

Regards
John