An alternative idea....
The low current draw from the LV 225v (26ma) invites another solution that might be worth looking at. Instead of a separate LV supply coming off the transformer C.T. use a series pass tube like a 6L6 off the 500V B+. So omit the choke, add a tube, then on the grid (G1) of tube use three 75v 1 watt zeners in series to get 225v (to gnd), and add a dropping resistor from 500v to the zener string cathode (about 150K), and bypass the zener string with a cap to gnd, maybe 4.7uf. Then triode connect the tube (G2 to plate through 10 ohms). The cathode goes to the 100uf and bleeder.
I'm sure your saying why? Well this approach is pretty simple (just adds a few more parts, and omits the 2nd choke). Here are a few pros and cons;
pro's:
~ 3 second turn on ramp of the LV as the 150K-4.7uf time constant ramps up
The extra load on the HV supply will allow the HV bleed resistor to pass less current (higher ohms and lower wattage) for the same % regulation
The 150K - 4.7 uf R-C will eliminate nearly 100% of the ripple in the LV output
The LV supply no-load to full load regulation will be good since the Rs will be the Rp of a triode connected 6L6
The voltage will be close to 225v as determined by the zener diodes.
Cons:
A short circuit will likely kill the 6L6
The HV supply has to provide 26ma more
The zeners will drift up ~ 5 volts as they warm up
The pass tube will dissipate power that was not dissipated before. (500-225) * .026 = 7.2 watts
Need to add an extra filament transformer for the 6L6.
I know, I just violated the KISS principle....
Jim
WD5JKO