Hi all.
I've decided to silence my beloved 6AS7 SE amplifier. I favor simplicity and can live with the little ripple noise I now have, but reading all these posts regarding the issue has got the better of me...
To keep me on the simple end I have a simple CCS feeding a capacitor to about 20volts below Vb. (Another benefit of this Vreg is the slowly rising V+). This voltage is then buffered with a MOSFET source follower. So the question is, where do the 'audiphile' capacitor (33uF polyprops) go? At the MOSFET's supply (drain) or at the MOSFET's output (source)?
Should I place one of the large electrolytics after the MOSFET as well?
I have no problems understanding the MOSFET providing a well buffered Voutput, but how about the dynamic performance? This is where I wonder if the polyprop cap might help if it is on the output end of the Vreg, so transient currents is supplied by this cap.
(I also have 0.47uF orange drop foil caps right at the OPT V+ node which I intend to keep as is).
I've decided to silence my beloved 6AS7 SE amplifier. I favor simplicity and can live with the little ripple noise I now have, but reading all these posts regarding the issue has got the better of me...
To keep me on the simple end I have a simple CCS feeding a capacitor to about 20volts below Vb. (Another benefit of this Vreg is the slowly rising V+). This voltage is then buffered with a MOSFET source follower. So the question is, where do the 'audiphile' capacitor (33uF polyprops) go? At the MOSFET's supply (drain) or at the MOSFET's output (source)?
Should I place one of the large electrolytics after the MOSFET as well?
I have no problems understanding the MOSFET providing a well buffered Voutput, but how about the dynamic performance? This is where I wonder if the polyprop cap might help if it is on the output end of the Vreg, so transient currents is supplied by this cap.
(I also have 0.47uF orange drop foil caps right at the OPT V+ node which I intend to keep as is).
I plan to use a MOSFET source follower regulator in pretty much the way you describe, i.e. using a current source (made of cascoded BJTs) to develop a voltage across a resistor, instead of a high voltage zener string, from which a 1 Meg resistor will be connected to the gate of the MOSFET. There will be a a 2uF cap from the gate to ground, to give a "slow start" and to minimize ripple. At the source of the MOSFET (i.e. the output of the regulator) I plan to use a 50uF smoothing/reservoir cap to ground.
I've more or less decided the cap goes on the circuit side of the MOSFET, as close to the OPT as pussyble. But perhaps someone has a good reason for letting the MOSFET do all the work without a reservoir in between?
The CCS I use is a BJT current-mirror with a resistor to gnd setting the current, 4mA in this case, and this current being mirrored into another resistor shunted by a 40UF cap. It takes a few minutes for it to be fully charged, which allowes the tubes to get nice and comfy before V+ has settled.
The CCS I use is a BJT current-mirror with a resistor to gnd setting the current, 4mA in this case, and this current being mirrored into another resistor shunted by a 40UF cap. It takes a few minutes for it to be fully charged, which allowes the tubes to get nice and comfy before V+ has settled.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.