From: ucfchuck on
i have a cypress fx2lp usb2 that emulates an ACE hdl player and outputs the
Jtag directly to an XCF32P and an SX55.

there are a couple of issues i am still not sure about though,
1- on a TDO check, the ace file gives me the output data for TDI, a mask,
and an expected return value.
The SX55 and XCF32P have a combined TDO buffer of 26 bits, so from what
i gather i dont get the data back out for 26 clock cycles, does this mean
that i should just continue clocking with no new instructions or data until
i get all the data back or do i continue feeding in new instructions and
just wait the 26 clocks like that?

2- what is the maximum TDO length for an ACE file? i have the FX2LP packed
full and can barely fit in another byte of storage, if i have to hold 100+
bytes in ram to check later im kind of screwed...

3- the XAPP424 talks about having to clock on the runtest instruction if it
is configuring an FPGA, the FX2LP link is not exactly fast enough to
continuously feed data through so the clock is sometimes just held high or
low for slightly longer periods on different instructions as im just
bit-banging it. should i slow it all down so i have a continuous clock
period or does it not matter as only the rising edge and minimum clock high
periods are the only important parts?





From: ucfchuck on
I realized i was reading the TDO instructions incorrectly from the XAPP424
and so i only need to store 26 bits or mask and 26 of expected TDO and 26
bits of received TDO, i thought it was <opcode><all TDI data><all TDO
expected><all TDO mask>, but it looks like it is broken into bytes of
each...
<0x04><numbits> <TDI byte> <TDO exp byte> <TDO mask byte> <TDI byte> <TDO
exp> ...

unfortunately it takes about an hour to run this programming procedure as
the ace file is about 40MB which has to go down endpoint0 and i have an
error check so its 2 transfers for every 32bytes. i will update if it
worked...

---------------------------------------
This message was sent using the comp.arch.fpga web interface on
http://www.FPGARelated.com
From: ucfchuck on
still not programming....

---------------------------------------
This message was sent using the comp.arch.fpga web interface on
http://www.FPGARelated.com