From 5568dec4ad8f0420f0a8605406fb6b3a0e4567c8 Mon Sep 17 00:00:00 2001
From: jim <jim@ddd99763-3ecb-0310-9145-efcb8ce7c51f>
Date: Fri, 5 Dec 2008 02:13:46 +0000
Subject: [PATCH] notes

git-svn-id: https://bucket.mit.edu/svn/nilm/zoom@7010 ddd99763-3ecb-0310-9145-efcb8ce7c51f
---
 notes/20081104.txt | 55 ++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 55 insertions(+)
 create mode 100644 notes/20081104.txt

diff --git a/notes/20081104.txt b/notes/20081104.txt
new file mode 100644
index 0000000..a777d1c
--- /dev/null
+++ b/notes/20081104.txt
@@ -0,0 +1,55 @@
+Setup
+-----
+
+Total input current Ic (amps)
+
+LA-55 with ~200 turns, 1k burden.
+Voltage output is Vc = (Ic / 1000 * 1k) = Ic
+LA-55 output goes to trimmable offset opamp,
+  trimmed to 0 when DAC command is 32768.
+
+  -> NOTE: DC offset changes a ton (200mV!) when the LA-55 is
+  saturated one way versus the other, then returned to zero.
+
+  Will try implementing a degauss feature, to see if the DC offset
+  always settles to the same amount.  Result: seems to work.
+
+R36 = 1k (shouldn't matter)
+IC3A:
+  R37 = 1k, R38 = 3.3k -> scales up by 4.3
+IC3B:
+  R39 = 100k
+  R40 = 100k
+  R41 = 100k
+  R42 = 100k
+Reference: 
+  R43 and R44 are a pot, trimmed so R43 == R44 -> VREF1.25 = 1.25V.
+  
+So
+  ADC12V+ = 1.25 - (4.3 * Vc)
+          = 1.25 - 4.3 Ic
+
+PIC input value (PICV) is:
+  0    = 0V
+  2048 = 2.5V
+
+  PICV = (ADC12V+ * 2048 / 2.5)
+  PICV = (1.25 - 4.3 Ic) * 2048 / 2.5
+  PICV = 1024 - 3522.6 Ic
+  Ic = (1024 - PICV) / 3522.6
+
+Computed by adc12_to_current()
+
+Scaling
+-------
+DAC can do about ±20A.  DAC command (DACC) is:
+  0     = -20A
+  65536 = +20A
+
+  DACC = 32768 * (1 + Ic / 20)
+
+Computed by current_to_dac()
+
+Code
+----
+