All, For what it's worth, our batch had the correct crystals (Epson EEG-2121-100.000L).
The boards we have are: D01020811 D01020829 D01020832 Regards, Richard Black On Thu, Feb 27, 2014 at 6:04 AM, Dan Werthimer <[email protected]>wrote: > > casperites, > > regarding alec's email appended below: > > the serial numbers of roach2 boards that might have the wrong crystal > oscillator, > range from D01020800 to D01020900 (D010208xx). > > best wishes, > > dan > > On Thu, Feb 27, 2014 at 3:35 AM, Alec Rust <[email protected]> wrote: > >> Hi all, we have found that an incorrect batch of crystal oscillators were >> delivered to Digicom and these made their way onto some ROACH2 boards. The >> batch affected is D01020800. >> >> The oscillator in question is U72, which is the 100 MHz FPGA clock. It is >> supposed to be a LVDS output clock, but the incorrect parts are LVPECL. The >> FPGA does work but intermittent problems occur. This also causes QDR >> read/write errors. >> >> U72 is between the FPGA and the DDR3 DIMM slot. >> >> The incorrect part fitted is marked CCPD33. See attached photo of an >> example of an incorrect and correct part. There are 2 parts that work: >> Crystek CCLD-023-25-100.000 and the part currently shipping on ROACH2 >> boards which is the Epson EEG-2121-100.000L. >> >> Please check your boards and let Mo at Digicom know if you have the >> incorrect part fitted. >> >> Regards >> Alec >> >> >
