Running 3.5 amps thru TRIAC causes it to overheat
Tested with HLG-480H-2100 charger

TRIAC (T810-600B) is rated to handle 8 amps, but that requires a much larger copper pour.  Looks like there's room on the PCB for that.  Look at "Figure 12: DPAK and D2PAK thermal resistance junction to ambient versus copper surface under tab" (page 6/21) for theta j-a values.  Looks like we'd need at least 4 square cm to appreciably improve performance.

Even after soldering a gullwing heatsink, the existing TRIAC was still too hot.
Potential replacement TRIACs:
BTA316B-600C,118

Another option is to add gull-wing heatsink:
V-1100-SMD/B (AE10774TR-ND)
V-1100-SMD/B-L (AE10775-ND)
ATS-PCBT1095 (ATS2137-ND)
573100D00010G (used on GG).  


Looking at "Figure 3: RMS on-state current versus ambient temperature" (page 4/21), the existing part isn't rated to carry more than 1.25 amps when the ambient temp is 55 degC.

That's what the existing grid charger pulls... so probably can't use a larger grid charger unless the TRIAC itself is replaced.  Tested with PLED200W-190-C1050-D (1.05 amp) grid charger and it caused 75 degF temp rise (147 degF in 72 degF ambient) on RevB PCB (when sinking at 120 volts into programmable load).

.........................

Also, need to add additional copper to forward biased diode D18... it runs a bit hot even with the standard 450 mA charger.  Existing diode is only rated for 2A continuous, and at that amplitude we're looking at 80 degC/Watt thetaJA.  At 2.1 amps, and with a 0.95 volt max diode drop, we're looking at 2.1 * 0.95 = 2 watts as heat in a single diode, which is 160 degC rise... I observed this in testing.

With the existing grid charger, we're looking at 0.85 volts drop * 0.450 amps = 383 mW as heat in a single diode, which is 31 degC rise.

Replace with diode: VS-4EGH06-M3 (slightly larger footprint)

..........................

Based on my testing today:
-Remove R331, to increase LED drive current.  This doesn't reduce temp, but will prevent LED aging from becoming an issue.  R59 stays 100 Ohms.
-Replace R73 (75 Ohm, large pads) with 100 Ohm resistor (0805).
-Removing R78/R337 pulldowns didn't affect temperature performance.  Can probably remove.
-Removing C33 didn't affect temperature performance.  Can probably remove.
-Can remove C32/R76... not needed based on extensive testing.

With the above changes, we'll have room for a larger TRIAC, with optional heat sink.  Don't put multiple TRIACs in parallel... they don't share current well.

.........................

Ordered the following parts for RevD:
BTA316B-600C,118
573300D00010G
S12MCV7GCT-ND

/////////////////////////////////////////////////////////////////////////////////

Solution:
-Changed R73 from 75 Ohm (large footprint) to 100 Ohm (0805).  This doesn't affect MOC3023 triac's phase angle.
-Removed R76 & C32 snubber
-Removed R331 to increase LED drive current.
-Changed D6 TRIAC footprint to D2PAK
-Added heatsink
-Changed D18 grid charger blocking diode footprint to DO-214AB (SMC)