add correct reason for why the other MOSFET was working during the first tests

README.md

@@ -141,8 +141,9 @@ Errata for V1.0
     with some fixup wires.
 2. CJ2301 and AO3004 have high input capacity. That seems to break the voltage converter
    for WS2812.
-  - It worked on all boards in the initial test (no DCDC, supplied by USB) but EMI from
-    the DCDC seems to be enough to break it on most boards.
+  - It worked on all boards in the initial test (no DCDC, supplied by USB) but this was most likely
+    because of the voltage drop on D5, i.e. the 5V were actually less than 5V, which is a known
+    method to make WS2812 work with lower voltage signals.
   - **Hotfix**: Replace by BSS138. They are a known-good choice and fxk8y has some at hand.
   - Fix is needed for Q7 but it would also be useful for Q3 and Q4.