pcb: update errata

pcb/DemoBoard1/README.md

@@ -1,4 +1,4 @@
-# Demo Baord 1
+# Demo Board 1
 
 Learning how to use shift registers, darlington arrays, and mosfets. I don't know what I'm doing.
 

pcb/LedArray/README.md

@@ -7,7 +7,7 @@
 ### Rev A
 
 1. I'm an idiot and somehow managed to switch polarity of LEDs. Rows were supposed to be cathodes, and columns anodes, but it's the other way around. Fixable by a 90° rotation, albeit with less than ideal trace layout. Oh well.
-2. The notes about the XL-1608SURC-06 LED are incorrect, and were in fact taken from XL-1608SURC-04's data sheet. The correct info is:
+2. The notes about the XL-1608SURC-06 LED are incorrect, and were in fact taken from XL-1608SURC-04's data sheet. (Correction: It seems that the ealier version of XL-1608SURC-06 datasheet was incorrect). The correct info is:
    - Forward voltage: 1.9-2.3V (at 20mA)
    - Max peak forward current: 80mA
    - Max continuous forward current: 20mA

pcb/LedArrayMux/README.md

@@ -1,3 +1,18 @@
-# LED Multiplexer - WIP
+# LED Multiplexer
 
 50x50mm module for driving rows and columns of LED Matrix modules
+
+## Revisions
+
+### Rev A
+
+Manufactured 2024-04-06, delivered 2024-04-13.
+
+Errata:
+
+1. PMOS gates are pulled up to VLED (5V), while the shift register is powered from 3V3 and signals above VCC are not allowed. Ouch!
+   - NMOS gates (i.e. row drivers) are not affected
+   - This is not easily fixable on the board itself
+   - It doesn't appear that 74HC595 dies immediately because of this, and the 10k resistor would limit worst-case scenario current to 0.5mA, but probably bad idea to ignore this altogether
+   - It's possible to set VCC to 5V, and add level shifting on signal lines
+   - Alternatively, the shift register could be replaced with 74HCT595 so that 3V3 signals register as logic high at 5V VCC