srenblad/artiq
1
0
Fork 0
forked from M-Labs/artiq
Code Pull Requests Activity

dsp.fir: cleanup

Browse Source
This commit is contained in:
Robert Jördens 2017-06-29 12:08:43 +02:00
parent dca662a743
commit 600a48ac61
1 changed files with 5 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
artiq/gateware/dsp/fir.py
View File

@ -78,14 +78,14 @@ class ParallelFIR(Module):
self.coefficients = cs = [int(round(c*(1 << c_shift))) self.coefficients = cs = [int(round(c*(1 << c_shift)))
for c in coefficients] for c in coefficients]
assert max(bits_for(c) for c in cs) <= w.B assert max(bits_for(c) for c in cs) <= w.B
max_out = sum(abs(c)*(1 << w.A - 1) for c in cs)
assert max_out <= (1 << w.P - 1) - 1, (bits_for(max_out), w)
### ###
# Delay line: increasing delay # Delay line: increasing delay
x = [Signal((w.A, True), reset_less=True) for _ in range(n + p - 1)] x = [Signal((w.A, True), reset_less=True) for _ in range(n + p - 1)]
assert sum(abs(c)*(1 << w.A - 1) for c in cs) <= (1 << w.P - 1) - 1
for xi, xj in zip(x, self.i[::-1]): for xi, xj in zip(x, self.i[::-1]):
self.comb += xi.eq(xj) self.comb += xi.eq(xj)
for xi, xj in zip(x[len(self.i):], x): for xi, xj in zip(x[len(self.i):], x):
@ -110,7 +110,8 @@ class ParallelFIR(Module):
else: else:
self.comb += o0.eq(o + m) self.comb += o0.eq(o + m)
assert min(js) - tap >= 0 assert min(js) - tap >= 0
js = [j for j in js if (p - 1 - j - tap) % p not in cull_delays] js = [j for j in js
if (p - 1 - j - tap) % p not in cull_delays]
if not js: if not js:
continue continue
self.comb += q.eq(reduce(add, [x[j - tap] for j in js])) self.comb += q.eq(reduce(add, [x[j - tap] for j in js]))
@ -129,6 +130,7 @@ class FIR(ParallelFIR):
def halfgen4_cascade(rate, width, order=None): def halfgen4_cascade(rate, width, order=None):
"""Generate coefficients for cascaded half-band filters. """Generate coefficients for cascaded half-band filters.
Coefficients are normalized to a gain of two per stage to compensate for Coefficients are normalized to a gain of two per stage to compensate for
the zero stuffing. the zero stuffing.