Make sure about the load current is covered like allowing more from the CCS initially. Because it kinda responds now better than before that was burning out right away. Check no small BJT is damaged first.
Changed the trimmer.
Add 10R as T.P. to adjust CCS Jung mod
CCS set 45mA
Vout 20.6V
1st heatsink very hot.
Add 10R as T.P. to adjust CCS Jung mod
CCS set 45mA
Vout 20.6V
1st heatsink very hot.
![WP_20141220_007[1].jpg](/community/data/attachments/423/423685-c86ac3be44051a4b866ad902a484927d.jpg?hash=yGrDvkQFGk)
Can't make out, its overworked. Anything could go wrong on that board. Why not building the whole circuit clean on a matrix board like you did the modded CCS, so to have control.
Needs a ballast series power resistor if using only the parallel part. 1K-1.2K-1.5K 10W should do in your case (60mA or 50mA or 40mA flow).
![WP_20141224_001[1].jpg](/community/data/attachments/429/429286-d6032cead6cfeeadccf062806928d2dc.jpg?hash=1gMs6tbP7q)
I think of no reason than some practical error not to work when having a limiting resistor.
Also:
"Another test tip is, as the circuit is off power, Q5 between E,B pins on Ohm meter with red probe on E should be reading in the hundreds of Ohm and changing reading when turning the R11 trimmer. When a blown Q5 could read a couple of Ohm near short. Q4 should be reading in the hundreds of kilo Ohm on same test probing (the trimmer is not involved in Q4)."
Also:
"Another test tip is, as the circuit is off power, Q5 between E,B pins on Ohm meter with red probe on E should be reading in the hundreds of Ohm and changing reading when turning the R11 trimmer. When a blown Q5 could read a couple of Ohm near short. Q4 should be reading in the hundreds of kilo Ohm on same test probing (the trimmer is not involved in Q4)."