Problem with FT245BM

• Previous message: uguess_at_nowhere.net: "Re: Smart Media or Compact Flash"
• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

Date: Tue, 13 Jul 2004 07:04:44 +0200

I have a strange problem with a FTDI FT245BM USB interface IC.

As soon as i enable the "pull down IO pins in USB suspend" option in the
config EEPROM, I read a bunch of garbage data form the FT245BM after
start-up.
Without the pull-down option the FIFO is (as expected) empty after
start-up.

I have the following setup:
- FT245BM connected to a ATmega64
- Configured as bus powered device (500mA)
- VCC and VCCIO are (for now) 5V
- VCC for the ATmega64 is switched using a IRLML6402 P-MOSFET

I think a have found what causes the FIFO state/content to be corrupted.
FTDI write in one of the appnotes that reading from the FIFO while it is
empty or writing to it while the TX buffer is full may lead to data
corruption.

Using a scope i see that during init of the FT245BM (while PWREN is still
high) the pull-ups of all interface pins (D0-7, WR, RD, TXE, RXF, SI) get
turned on and off 3 times!?

As this only happens with the pull-down option enabled, my best guess is
that the edges on RD corrupt the internal state of the FIFO.

The only help I got from FTDI is that the pull-ups are initially turned on,
then the EEPROM is read and if the pull-down option is enabled they should
get turned off until PWREN goes low.
(I guess after they noticed that i'm "only" a hobbyist they stopped
responding ...)

Has anyone here used the pull-down option with success?

Thanks in advance,
Bertolt

-- 

• Previous message: uguess_at_nowhere.net: "Re: Smart Media or Compact Flash"
• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]