I'm designing a power supply for hybrid amplifier (tube + discrete semiconductor power amplifier). I need 30V for power amp and 60V for tube. Also I need 6.3V for the filament (6DJ8 dual triode doesn't allow series connection of filaments so I can't use 12V power supply).

Let assume I have a transformer with two secondaries rated to 15V each. I can connect windings in series and get 30V directly and 60V using voltage doubler.

To obtain 6.3V for the filament I can utilize transformer's midpoint and use a linear regulator. In this case all filaments' current will pass through single winding. Can I spread this load to both windings? Filament power supply may float.

Edit: I think something like this may work. Values are nominal, regulation is omitted.

enter image description here

  • 2
    \$\begingroup\$ If you're using a toroidal transformer, it's pretty simple to bodge on an additional winding for an extra voltage. \$\endgroup\$
    – Ian Bland
    Nov 1 '16 at 23:05
  • \$\begingroup\$ Can I estimate approximate turns number for starting point from transformer's power rating? In other words can transformer's rating say me something about primary turns number? \$\endgroup\$
    – e_asphyx
    Nov 1 '16 at 23:34
  • 2
    \$\begingroup\$ You can do it pretty accurately by winding on a few turns and measuring the voltage you get, then dividing that by the number of "test turns" to work out how many turns you need. Since you will be regulating the voltage anyway, you don't need to be super accurate. It's often a simple and effective way to get a necessary voltage compared to other means. And it's genuinely floating until you reference it to the circuit. Enamelled wire is cheap and freely available. It may be your answer. The power rating won't help with the turns as the voltage is a turns ratio, plus (some amount) for losses. \$\endgroup\$
    – Ian Bland
    Nov 1 '16 at 23:57
  • 1
    \$\begingroup\$ There is a large common mode noise on the floating 19V filament pre-reg voltage which can't be good for EMI with 50Hz and buzz up to 10kHz from the diode current spikes. It would inject a buzz into the valve or nearby wiring. I would go with well filtered SMPS supply. \$\endgroup\$ Nov 2 '16 at 0:24

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.