Here's some stuff I wrote up while I was trying to figure out Aaron's FFT
magic which might be useful to others. Read at your own risk of course.
http://www.eecs.berkeley.edu/~sgowda/parallel_fft_algorithm.pdf
-Suraj
On Mon, Jun 23, 2014 at 11:43 AM, Aaron Parsons
apars...@astron.berkeley.edu
Hi all,
The memory footprint of the fft_direct block could be reduced if it is
split into two blocks, one for the operator to manipulate the phase of the
FFT of each input and another block to calculate the true direct-form FFT,
i.e. not mapping together a larger FFT (see p. 615 of
Hi all,
Does anyone know what causes the error 'cannot execute binary file' when
trying to execute a bof file from the powerpc shell? This error occurs
with all bof files, including the test designs from github
Thanks
Suraj
Hi everyone,
Does anyone have a ROACH2-compatible pcore for the adc083000? The
auto-update of the clock manager from DCM to MMCM doesn't seem to work
properly. I remember discussing this issue before but I can't remember if
it was ever resolved.
Thanks,
Suraj
If I remember correctly, the SMA inputs are not normally run through a DCM,
they're just IO pins (I could be wrong about this..). Specifics on how to
make a yellow block are on the wiki:
https://casper.berkeley.edu/wiki/How_to_make_a_yellow_
block. Assuming your block is named 'newblock':
0)
Using DSP48s with for addition with a latency of 2 yields the best timing
performance, from experience and Xilinx's recommendations, which is why it's
hard-coded.
-Suraj
On Thu, Jul 28, 2011 at 11:53 AM, Samuel Tun samuel@gmail.com wrote:
Yes, unchecking the DSP48 adders allows me to change
Hi Kim,
The MPD, BBD, and PAO files should go in
mlib_devel/xps_lib/XPS_ROACH_base/pcores/adc_iserdes_1_00_a/ data/
If you're re-compiling from MATLAB, make sure to re-copy the base
system.
If this can be clarified on the Wiki page instructions, let me know
-Suraj
On Apr 13, 2011, at 4:46
Sorry for the response delay, I've been on a hacker-induced hiatus
from Simulink.
On Mar 25, 2011, at 6:23 AM, Andrew Martens wrote:
Hi all
I have read the bit of documentation available on the xBlock
interface, looked
at some of the scripts, and tried to debug an xBlock problem.
It
Hi all,
I've developed a PFB FIR real optimized for V5, making the (now
usual) modification of directly instantiating dsp48s to use the same
multipliers and adders. The problem is that this particular
optimization requires eliminating the tap block in the PFB and doing
all the
Hi Sam,
This might be a case where Simulink doesn't find the library. Are the
libraries visible in the Simulink LIbrary browser? if they are, try
pulling any block from the CASPER DSP blockset into your design,
minimize the .mdl and then re-examine it. Are the links still broken
after
Hi all,
Does the ROACH come with a convenient way of measuring power
dissipation?
-Suraj
Hi all,
In push eb52771 (I think, I'm still pretty bad with git...), I rewrote
the fft_wideband_real_init.m script to draw pipeline delays inside the
FFT, for the inputs to the biplex cores and between the fft_direct
block.
I also added a script init_planahead.tcl to
Hi John,
Just to clarify, do you mean 1 board with 16 outputs / cycle at half
the clock rate? I implemented this feature but I wouldn't recommend
using it for serious computation. For operations like FFT, speed is
not as much a problem as the number of simultaneous inputs.
-Suraj
On
At 3 Gs/s, the FPGA clock rate is 187.5 MHz. Is this in range of the
DRAM clock?
Maybe you could describe what the output is and/or how it's incorrect?
It's been a while since I attempted to use that interface.
_Suraj
On Sep 29, 2010, at 6:08 PM, John Ford wrote:
Hi John,
Just to
Hi Mandana,
Your error is not a Simulink error. It seems that Xilinx's synthesizer
can't find the distributed RAM block used in the ROM for the FFT
unscrambler. Check in the XPS_ROACH_base/implementation for your design
and see if you can find the ngc netlists that it mentions. If not, system
...@astro.berkeley.edu wrote:
On Jul 27, 2010, at 10:33 , Suraj Gowda wrote:
When I try to run the .bof from the ROACH command line, it returns
with
Input/Output Error. The ROACH is NFS mounted and I put the bof
in the
filesystem via scp.
Sounds like a corrupted/incomplete bof file. What are its
Hi casperites,
When trying to draw a 16-input real FFT, I get the following error
message in 10.1 about 109830498 times:
Warning: In instantiating linked block
'spec_256ch_250MHz_clk/fft_wideband_real/fft_direct/butterfly1_5/
twiddle/convert_of2/never' : Xilinx Constant Block Block block
Agreed. At the very least, you shouldn't be able to override function
names with variable names. 1 point to strongly typed languages...
On Jun 24, 2010, at 12:40 PM, Aaron Parsons wrote:
FWIW, I've noticed another related problem in the pfb_fir and
pfb_fir_real
blocks. The mask callbacks
Hello casperites,
Does anyone know the format for lines in XPS_ROACH_base/
core_info.tab? All I see is 3 numbers:
iadc_controller 321
Thanks,
-Suraj
Hello casperites,
I am looking at the code in opb_adccontroller.vhd and user_logic.vhd.
I understand the state machine operation, but the operation seems to
depend on the input data Bus2IP_Data, and I'm not seeing where that
bus actually changes values. Is it something hidden in
Hi Steve,
The GUI interface to the block will not change from what you've done
unless the mask script dynamically redraws the ports. This may be the
case, depending on the block. If not, then deleting readding the
block won't do anything to update the ports. You'll have to unlock
the
Hi John,
I don't think anyone has been able to get a spectrometer working at
the full 3GS/s interleaved yet.
We are able to digitize and plot raw ADC data at full speed. We run
into timing problems when the chip becomes more full, i.e. with an
FFT. One possible solution we are (I am)
Xilinx style fix: Use Glenn's vector accumulator from the GAVRT library.
On Apr 7, 2010, at 7:32 AM, Aaron Parsons wrote:
It looks like this block is not correctly configured in the library.
Volunteers to fix/test?
--
Aaron Parsons
510-406-4322 (cell)
Campbell Hall 523, UCB
The yellowblocks for the TI DAC part is in xps_library/DACs. I added
(but didn't write) the blocks to the main CASPER SVN maybe 2 weeks
ago, but did not do any testing as I don't have access to either
board. Both blocks are bare bones and contain no simulation
interfaces, but they are
Multiple times, after starting a simulation MATLAB will crash and then
suddenly nothing from that design will work. So copying the design to
a new file and trying to simulate does hangs in the initialization
part of the simulation. Anybody else have this problem and/or a
solution?
I'm
My issue was solved under a different alias. If anyone else has this
problem, the thread 'Reference to non-existent field clock_loc'
contains a hack.
On Feb 16, 2010, at 12:46 AM, Suraj Gowda wrote:
Multiple times, after starting a simulation MATLAB will crash and
then suddenly nothing
Hello,
On Nov 3, 2009, at 6:39 PM, C-H Cheng wrote:
Hello All
If I want to simulate a design in ISE and generate a bit file to
download to ROACH over JTAG.
You can do this using the .bit generated by the CASPER toolflow,
available in the same location as the .bof. bof files are generated
,
-Suraj Gowda
-
opened. Is this a known problem with a solution?
Thanks,
-Suraj Gowda
29 matches
Mail list logo