VRef is working OK, as you can see in mine post 4.292 third pic. I don't understand why can't get 6.3V across filaments using 2R2 +3R pot in series???
R1=240R, R2=1k Vout=~6.5V P(ic)=3.5W based on powering one 6V6 heater from a 13.6VDC supply. Heatsink should be large enough or internal protection circuitry will prevent the LM317 from working properly... I would burn voltage in an RC filter before the LM317.
2R and 3R3 in series is way too small in value.
2R and 3R3 in series is way too small in value.
If it was 15V it could even allow for two CCS driven heaters in series
Here's mine original dual heater CCS. Been working for ages. You can see dual big bridge and 22.000uF raw DC sources fed from two separate AC windings. Going to the LM317s with 0.1uF MKT & 10uF Tantalum decoupling. Three resistors to tune Rset value.
Next to it my version of Mosfet M-Cap Filter for B+ after Maida died. CCS heat PSU outlived Maida which became erratic, but after many good working years, due to Mosfet, 317, and Vout set Rdivider, all at relatively high dissipation. The filter PSU ended up with its output cap also lytic. MKP output was bit stressing HF with its very low ESR. Couldn't fit an SSHV2 in there which is superior to both. There wasn't provision since the 6V6 build preceded its invention.
Next to it my version of Mosfet M-Cap Filter for B+ after Maida died. CCS heat PSU outlived Maida which became erratic, but after many good working years, due to Mosfet, 317, and Vout set Rdivider, all at relatively high dissipation. The filter PSU ended up with its output cap also lytic. MKP output was bit stressing HF with its very low ESR. Couldn't fit an SSHV2 in there which is superior to both. There wasn't provision since the 6V6 build preceded its invention.