Plugged in battery -> firework

kicad/mains_side.kicad_sch

@@ -1670,6 +1670,10 @@
     (effects (font (size 1.27 1.27)) (justify left bottom))
     (uuid 156df3ac-a2e5-464c-80fe-c4e2a3e4a24a)
   )
+  (text "Abs. max. Vbe = 6V" (at 140.97 138.43 0)
+    (effects (font (size 1.27 1.27)) (justify left bottom))
+    (uuid 33c073b7-f0bf-48b6-ab66-ac770ab161ee)
+  )
   (text "max. 0.5A" (at 148.59 104.14 90)
     (effects (font (size 1.27 1.27)) (justify left bottom))
     (uuid 759bcd2d-5441-493a-9f65-c7be38b5225c)

prints/V5 measurement 4 24-08-23/legend.txt

@@ -0,0 +1,4 @@
+CH1 - Voltage at the cathode of *D6  
+CH2 - Voltage across R5  
+CH3 - V1 Vbe 
+CH4 - V5 Vgs 

troubleshooting.md

@@ -102,29 +102,22 @@ mit Harry
 ![](prints/V5%20measurement%201%20V5%2023-08-23/DS1Z_QuickPrint1.png)
 (All measurements are with respect to gnd)
 
-The voltage across R19 (above optocoupler U1) was also measured during operation. 
-Channel legend on prints:  
+legend:  
 CH2 - V1, Vce  
 CH3 - V1, Vbe  
 CH4 - V5, Vds  
 
 ![](prints/V5%20measurement%203%2023-08-23/DS1Z_QuickPrint2.png)
 
 legend:  
-CH1 - Voltage at the cathode of *D6  
+CH1 - Voltage at the cathode of *D6  (because it defines the working point of V1)  
 CH2 - V5 Vds  
 CH3 - V5 Vgs  
 CH4 - V1 Vbe 
 
-What killed the npn? - We will never know
+The voltage across R19 (above optocoupler U1) was also measured during operation. 
 The secondary side can't influence the primary side however because the op-amps are not powered.
 
-Measured: 
-CH1 - Voltage at the cathode of *D6 (because it defines the working point of V1)
-CH2 - V5 Vds
-CH3 - V5 Vgs
-CH4 - Voltage across *R5 
-
 Conclusions: 
 The working point of V1 should be about 4.583 V according to the voltage divider.
 But a measurement shows that Vbe of V1 never exceeds 1 V (ignoring oscillations)
@@ -138,4 +131,43 @@ Resonant frequency
 
 f = 1/(2*pi*sqrt(L*C))
 
-with L = 9.31 uH / C = 10 nF gives f = 5216 kHz which is close to the measured 4880 kHz switching frequency -> C5/TR1_L1 must be the main resonant circuit
+with L = 9.31 uH / C = 10 nF gives f = 5216 kHz which is close to the measured 4880 kHz switching frequency -> C5/TR1_L1 must be the main resonant circuit
+
+### 24.08.23
+
+- Added two parallel 4.3 Ohm resistors in parallel to the already installed 1.95 Ohm for a total of 4.1 Ohms for the blown R5.
+   - subsequent power up: The voltage on the secondary side decreased further to ~ 0.75 Vrms. 
+- Replaced R5 with 1 Ohm resistor.
+   - subsequent power up: secondary side Vrms = ~ 1.2 Vrms
+
+   ![](prints/V5%20measurement%204%2024-08-23/DS1Z_QuickPrint3.png)
+
+   Observations:   
+   - Significantly less ringing  
+   - V5 switches off once Vbe of V1 reaches 0.890 V which now takes 1.48 us
+
+- Replaced R5 with 0.6 Ohm resistor.
+   - subsequent power up: Less then a second operation, death
+   - Desoldered V1, ttested: diode from 1 to 3 and from 2 to 3, failed.
+
+- Replaced R5 with 0.8 Ohm, replaced V1
+   - subsequent power up: Less then a second operation, death
+   - tested V1 in circuit, 7.5 kOhm between C/E (like previously), 17 ohm between b/e
+
+- Replaced Fuse (wasn't blown but looked burned)
+- Replaced V1 and R5 with 1 Ohm again
+   - subsequent power up: 1.15 Vrms at the secondary, but briefly up to 7 Vrms, might be something during the power up.
+   - Further measurements of the secondary side cap show that even though the uni-t UT61D showed ~ 100 mV DC the oscilloscope showed Vavg = 1.64 V and significant ripple
+
+- Plugged in a battery (while powered on): Bang, sparks, firework, death.  
+   - Damaged parts: F1 and R5 blew violently, C2 caught in R5s blast, R6 visibly damaged. V5 not visibly damaged
+   - R5 now measures 62 kOhm
+   - F1 is open line
+   - R6 now measures 10.5 kOhm
+   - V1 Rbe = 2.7 Ohm, Rce = 25.7 Ohm
+   - V5 Rgs = 22 ohm, Rds = 11.3 Ohm
+   - Pad of V5 gate lifted!!!
+
+   - Theory: The supply voltage is kept low to safe power while the battery charger is idling. Once a battery is detected by the micro controller it increases the secondary side voltage to match the desired charging current (cc charging phase) or if the voltage is already big enough cv. This is achieved via the optocoupler U1. By increasing the photo-transistors c/e-resistance the Vbe of V1 is decreased which again increases Vgs of V5, letting current flow for a longer portion of a period.
+
+