From d64cfcd6a238cd95e868dc5364883b936de10fb2 Mon Sep 17 00:00:00 2001
From: Eric Kooistra <kooistra@astron.nl>
Date: Thu, 18 Jul 2024 17:30:06 +0200
Subject: [PATCH] Clarify type 2 (advances) polyphase and Transposed Direct
form.
---
applications/lofar2/model/pfb_os/dsp_study_erko.txt | 8 +++++++-
1 file changed, 7 insertions(+), 1 deletion(-)
diff --git a/applications/lofar2/model/pfb_os/dsp_study_erko.txt b/applications/lofar2/model/pfb_os/dsp_study_erko.txt
index 72eea03d20..4aa7d564a5 100644
--- a/applications/lofar2/model/pfb_os/dsp_study_erko.txt
+++ b/applications/lofar2/model/pfb_os/dsp_study_erko.txt
@@ -1269,9 +1269,15 @@ c) s-plane and z-plane
where:
rp[n] = h(Qn - p), -p for clockwise with advances z^(+1)
- Rq(z) = E_{Q-1-p}(z), so flipud phases, but keep coefficient order
+ Rp(z) = E_{Q-1-p}(z), so flipud Eq phases, but keep coefficient order
per phase
+ ??? Is Type II polyphase same as Transposed Direct Form ???, because both still
+ implement H(z), they only yield different implementation structures. For Type 1
+ and Direct-Form the delay line goes from top to bottom in drawings. For Type 2
+ and Transposed we also want to draw the delay line from top to bottom (Fig 4.3.4,
+ Fig 4.3.11, Fig 4.5.6), but then we need to flipud the Eq phases.
+
. Phase q of h(n) with Q - 1 zeros, so [VAIDYANATHAN Fig 4.3.1]:
h[n] --> z^(-q) --> Q:1 --> eq[n]
--
GitLab