Wednesday, June 13, 2012
Ju-Jutsu: Ultimate Monoblock (Pt 4) Output Stage
While we're at it, lets look at the real output stage:
Again we see balanced-input complimentary Mu-Follower circuits,
each providing 1/2 of the P to P voltage across the transformer.
The nice thing is, there is no DC flowing through the OT.
The tubes get their current from the Power Supply through the top tubes,
and the OT route is superfluous.
The blocking cap I used was 10 uF, 600v polyprop, the only expensive cap in the whole amp.
It's also bypassed with a fastcap low esr, just in case.
The cap should not face any voltage other than the output signal.
If there were a slight imbalance between the tubes,
and a small DC voltage difference between the two tapping points,
who cares? its blocked.
The beauty of this is that the tubes don't even have to be balanced, as in a std Push-Pull.
The DC offset can't dump any current into the tranny, so there's no concern at all.
The tubes would have to be woefully out of wack to have any effect on the signal.
Since this is a constant-current style circuit, operating high in Class-A,
there is no question at all of 'blocking-distortion' from grid current in the bottom tubes.
But still the input grid-leak resistors should be as recommended by tube-makers,
e.g., for a 6550 say 100k and an appropriate input cap, say 1 uF.
Again with the Hafler cathode circuit, which minimizes Push-Pull IM.
What can I say?
These output tubes have to be self-biased, because of the huge high voltages across the circuit.
But thats okay, because the Mu-Follower forces linearity on both tubes,
even before they have to cancel out small non-linearities from each other.
Win win again.
Use large wattage cathode resistors, and adequate ventilation.
Don't forget sensible grid-stoppers and screen-grid stoppers on all power pentodes/tetrodes in Triode-mode.
Look at recommendations for the specific tubes you're using.
I left out a 100-200 ohm parallel resistor across the output secondary, to prevent spikes
in case of speaker-failure or disconnection.
The top tube in the WWII model (5998) cannot be substituted with a 6AS7G,
without significant modifications.
Try to put 1/2 to 2/3 of the B+ across the bottom tubes.
If you're using an SE transformer you can try omitting the blocking-cap in the primary!
Again, critically important to have a separate floating or DC-biased heater-supply for the top tubes.
Stay within 100 volts of the cathode, or as per tubesheet recommends.
Not only is any PS noise blocked by the heavy topload Mu Followers,
whatever signal gets through is near-perfectly cancelled across the transformer primary.
Remember that these Output stages are nearly CC,
so whatever small PS noise is left from the voltage-divider action,
is almost always near-perfectly cancelled.
noise free. hear the difference.