1
\$\begingroup\$

I am trying to connect FT2232H breakout board (http://www.digikey.com/catalog/en/partgroup/ft2232h-evaluation-board-ft2232h-mini-module/15377) to DE0-Nano FPGA board (http://www.altera.com/education/univ/materials/boards/de0-nano/unv-de0-nano-board.html) and use it in FT245-style synchronous FIFO. In general, everything works, but system is unusable, because I receive some bytes twice (usually, receive 10-15% more data than send).

FPGA is clocked only by FT2232H (60 MHz) and entire system is synchronous so I think this is happening because I connect these two boards with 10 cm prototyping wires which are not exactly suitable for frequencies this high and don't use any pull-up/pull-down/series resistors because can't figure out which ones should be used. I've tried to connect 100 ohm (50 — 300, as suggested on some forum I've googled) resistor in series of clock signal — it decreased byte repeat ratio but not entirely. Is it this or something else causing problem and how should I solve it?

\$\endgroup\$
4
  • \$\begingroup\$ So your theory is the clock is 'noisy' and so the FPGA is clocking in the same data twice? Is the FPGA 'hand-shaking' with the FT2232H? AFAICT, RD# and OE# can be used to handshake with the FT2232H, and read the data at a rate convenient to the receiver. Do you do that? It might be worth posting a schematic of the wiring (even hand drawn and a photo might help). \$\endgroup\$
    – gbulmer
    Commented Sep 20, 2014 at 13:03
  • \$\begingroup\$ @gbulmer, yes, FPGA is handshaking. RD# and OE# are used at the beginning and in the end of transmission, during the transmission data is latched on rising edge of the clock and sometimes it happend twice. Schematic of wiring is just TXE, WR, OE, RXF, RD, SIWU, CLK connected to the appropriate DE0-Nano pins with 10 cm prototyping wires. \$\endgroup\$
    – themylogin
    Commented Sep 20, 2014 at 17:24
  • \$\begingroup\$ And you connected GND between the two boards, right? Without a ground return wire, it will "kind of" work sometimes, but not reliably. \$\endgroup\$
    – MarkU
    Commented Sep 20, 2014 at 19:57
  • \$\begingroup\$ @MarkU yes, of course \$\endgroup\$
    – themylogin
    Commented Sep 21, 2014 at 3:34

1 Answer 1

Reset to default
2
\$\begingroup\$

Lack of pull-down resistors for CLK, RXF, RD, OE (even with breadboard connection) was causing the issue.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.