I did not believed it until I heard it with my own ears .Replacing the GZ34 with its equivelent 5AR4 on a300b SE mono bloc amp ,the high frequencies got metamorphosed,as if the tweeters were replaced by different type.
With this incontestable evidence ,you don't need to be a polytechnic graduate to understand that the choke inductor is the cupid . When the rectifier conducts to charge the input capacitor it grounds (AC) the inductor, paralleling to the output capacitor,hence high impedance tank notch gets created.Of course the 10henry and 47uf makes about 7 hertz but the rectifiers does not conduct continuously and the inductor is not 10 henry at high frequencies.By high impedance rectifier as tube, more damping is provided . Japanese even use serie resistors with tube rectifiers. (see Asno 2a3 Loftin White)
I did some trials (early 80's) using an "electrophone" amp (ECL82 EZ81) and the primary of a small ps trans. as choke, I reached to the conclusion that this phenomena occurs only if the output transformer is fed by low impedance ,triod or anode feedback, and above all ,without overall feedback .
You don't need to be the one who has invented the wheel to find that a doubler power supply is the radical solution for this problem . While the rectifier is charging one capacitor the other one is discharging in series with the choke , hence no more parallel tank resonator.
A few years later Jean Hiraga published a "hors-serie" of LED magazine untitled "Initiation aux amplis a tubes" .A total omerta about this subject , note that he does not use tube rectifiers in his realisations.But at the end of his book he added a precious collection of amp schematics, one circuit did use a doubler power supply in class AB amp.
Conclusion:
If you need to use choke inductor ,always use doubler power supply with semiconductor rectifier. http://www.ampslab.com/vintage_marantz9.htm
If you need to light a bulb in your power supply instead of a candle for saint Isidor ,stay in your faith to harvest a successful amplifier .
KOKORIANTZ
With this incontestable evidence ,you don't need to be a polytechnic graduate to understand that the choke inductor is the cupid . When the rectifier conducts to charge the input capacitor it grounds (AC) the inductor, paralleling to the output capacitor,hence high impedance tank notch gets created.Of course the 10henry and 47uf makes about 7 hertz but the rectifiers does not conduct continuously and the inductor is not 10 henry at high frequencies.By high impedance rectifier as tube, more damping is provided . Japanese even use serie resistors with tube rectifiers. (see Asno 2a3 Loftin White)
I did some trials (early 80's) using an "electrophone" amp (ECL82 EZ81) and the primary of a small ps trans. as choke, I reached to the conclusion that this phenomena occurs only if the output transformer is fed by low impedance ,triod or anode feedback, and above all ,without overall feedback .
You don't need to be the one who has invented the wheel to find that a doubler power supply is the radical solution for this problem . While the rectifier is charging one capacitor the other one is discharging in series with the choke , hence no more parallel tank resonator.
A few years later Jean Hiraga published a "hors-serie" of LED magazine untitled "Initiation aux amplis a tubes" .A total omerta about this subject , note that he does not use tube rectifiers in his realisations.But at the end of his book he added a precious collection of amp schematics, one circuit did use a doubler power supply in class AB amp.
Conclusion:
If you need to use choke inductor ,always use doubler power supply with semiconductor rectifier. http://www.ampslab.com/vintage_marantz9.htm
If you need to light a bulb in your power supply instead of a candle for saint Isidor ,stay in your faith to harvest a successful amplifier .
KOKORIANTZ
The rectifier resistance (plus transformer resistance) plus choke resistance will all play a role, but as the choke is only "grounded" for a brief time 100 or 120 times a second but its resonance with the output cap is at a much lower frequency you need to approach this in quite a different way. What actually happens is that the rectifier etc. plus the reservoir cap act as a resistance at low frequencies. This damps the subsonic resonance - so it will vary a bit with different rectifiers, but perhaps not as much as you might expect.
This small variation in the subsonic behaviour of the PSU will not, of course, have any material effect on the HF performance of the amplifier. However, ears and brains are easily fooled so a change in the way the amp treats a musical envelope may be perceived as a change in HF. This is especially true if the amp is a low (or zero) feedback SET, as then the amp gain will depend on the supply rail voltage so subsonic ringing in the PSU will create intermodulation. The degree of 'muddiness' may thus depend on rectifier choice. This IM effect will be absent from a more conventional PP amp with sensible amounts of feedback.
Doubler power supplies, other things being equal, have higher output impedance and may make PSU subsonic ringing worse so worsen IM. Best avoided, but I suppose some people might like the resultant sound.
One source follower with a Zener reference source is enough to remove transformer and rectifier from the equation. Now watch for cathode bias shunt time constant, and a coupling cap before the final output tube. However, don't forget about currents that charge filter cups, it is the common mistake causing rectifiers to "sound".
Doubler supplies can be highly satisfactory. Implementation details are key, regardless of PSU topology selected. The revered Marantz 8B, H/K Cit. 2, and some McIntosh models use doubler B+ PSUs.
Doubler PSUs have an advantage in the copper loss dept. Jim McShane leads the way in obtaining superior performance, including low impedance, from doubler PSUs. It's not surprising that TANSTAAFL applies. Large valued doubler stack caps. are needed to obtain equivalence with "routine" FW supplies. As I have indicated previously, large valued caps. at the I/P of a PSU filter come with some "baggage". A "hash" filter, along with a low DCR filter choke, between the large I/P capacitance and the reservoir capacitance deals with the matter.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.