Hi Luis -- What a lucky break that you were able to obtain a KX-200 power transformer!
In a nut shell, if I understand your approach correctly, there are various problems, not the least of which is that you're trying to have a common transformer winding power two separate rectifier circuits that both have the winding reference chassis ground at different levels, which won't work. To make the winding of this transformer work will be tricky because as is, it has too much voltage for some of the applications, and not enough for others. The trick then is to produce the voltages needed with proper rectification, without having to drop a lot of unneeded voltage through large dropping resistors. If I may then, and if you are committed to using a dual rail supply, then I would approach your project this way:
1. There is no real need or practical result from regulating the heaters in this application. To power the heaters, the regulators would require heat sinking (likely to the chassis), transferring the heat of dropping resistors, to the regulators. Also, using just one-half of the winding to power the heaters with a bridge rectifier design will quickly overheat the winding, since it would be conducting current 100% of the time, whereas in the original design, each half only conducts current 50% of the time. The buffer amp does not require regulation either, nor any further reduction in ripple beyond that provided by the original filtering scheme because of the large common mode rejection factor the op amp has - particularly when operated from dual supplies. However, the regulators can still be quite useful to power the op-amp buffer (only) if a dual supply approach is taken, as will be shown.
2. To produce a proper dual (+/-) supply to power the op amp, the CT of the transformer winding must reference the CT of the op-amp supply to properly center the the 0 volt point between the rails of the supply. With that understanding then:
3. For the heaters, connect them in series/parallel pairs as shown for the KX-200.
4. For the heater supply, use the supply as shown for the KX-200 (without bias take off), including the filter network as shown.
5. For the other rail of op-amp supply, you would build a mirror image rectifier circuit (effectively making a full bridge), with the cathodes of these rectifiers (banded end) connected to the transformer winding. An appropriate filter cap would be connected between the output of these rectifiers (- to the rectifiers), and the transformer CT. The output of these rectifiers would also power the EFB™ Grid Bias Regulator.
6. The circuit now has the CT of the winding at (about) -32 vdc below ground (as in the KX-200), and the output of the additional diodes at (about) -64 vdc below ground.
7. Since this is too much voltage to power the op amp with, your + and - voltage regulators would be used to effective supply the op amp with +/- 12 vdc, but notice that the 0 volt point (neutral) of the op-amp supply is actually the -32 volt point (CT of the winding, as it should be), so that the + regulator is actually fed from ground, with its output representing -20 vdc relative to circuit ground, and the output of the - regulator (powered from the new diodes) representing - 44 vdc relative to circuit ground. As a result, input and output coupling caps must be used for the op-amp (like in the 400's buffer), because the 0 volt point of the buffer supply does not represent chassis ground with this scenario.
The only other option you have -- which is the best option in my opinion -- is to in fact ground the center tap, forcing the op-amp neutral to operate at true ground level. In that scenario, you need to "flip" the ground of the heater circuit so the the transformer CT is grounded, not the positive side. The filter resistor would be moved to the positive side as well, which prevents use of a common positive can for the heater filter cap. You still need the diodes representing the "other half" of the bridge to supply (in this case) about -32 vdc to the negative regulator for the op-amp, but the dual advantage of this approach is that the input and output coupling caps can be done away with in the buffer circuit, and, the heaters now operate positive with respect to ground, not negative -- which acts to further reduce any noise that might otherwise transfer between the heater/cathode junctions within the tubes. The down side of this approach is that now you need the use the special voltage multiplier circuit I developed for the original 400 receiver project, because the -32 volt supply that powers the -12 volt regulator for the op amp is not enough to power the EFB circuits. However, powering the EFB circuits from a source of about -100 vdc (which the multiplier circuit would produce in this case) really acts to make any line voltages invisible to the EFB regulator.
I know you were trying to get around using the multiplier circuit, but once again, it really is the best option, with the output cap of the circuit being the only one representing any significant size.
Finally, if you dispense with the idea of using dual supplies to power the op-amp, then things get pretty simple: use the existing KX-200 heater circuit as is (without the bias connection), power the op-amp with a faux balanced supply like it did in the 400 receiver scenario from the heater supply (in/out coupling caps required), and add on the two extra diodes and filter cap to create a -64 volt point to power the EFB regulator.
Good luck with your project!
Dave
) and all the transformers reamains good and working. A pair of 7591 the transconductance is a quite low but allows yet to adjust bias.
