A few years ago I built this amp based solely on Andrea Ciuffoli's design (audiodesignguide.com). Over the years I have been swapping parts in and out trying to improve it. I have now finally settled on what I believe is the finest 300B amp I have ever listened to. It is certainly the quietest. I have Klipsch La Scalas at a rated 105db spl and you can not even tell the amp is turned on at 3/4 volume. I am publishing here the schematic and the parts I used in case anyone wants to build a sure thing as far as performance goes. If anyone lives in Oregon Near Corvallis you are welcome to come listen to it. This is not my design. I only made some component choices.
Jeff
Jeff
This is Schade negative feedback (feature, not a bug). I am wondering whether it is really necessary.
Attached is the original schematic of post #1.
TS changed the schematic of post #1 after my post #2.
TS changed the schematic of post #1 after my post #2.
1. For Newbies, Be careful if you wire according to the schematic . . .
Inside the glass envelope of the rectifier, it shows a very unusual drawing of the plates and the cathode.
The cathode is drawn as the traditional straight line of a plate.
It also draws the plates as slanted straight lines.
Look up almost any 5AR4 data sheet for a more proper and traditional way of drawing the elements.
2. Schade negative feedback:
When using a triode wired pentode (as this circuit does), or when using a true triode for the input tube . . .
If you bypass the cathode bias resistor (as this circuit does), the plate resistance, rp, is reduced; That reduces the amount of negative feedback.
That is why most Schade negative feedback circuits do not bypass the cathode bias resistor of the input tube.
My conclusion is that there is effectively very little Schade negative feedback in this amplifier, and I am not surprised that it sounds excellent with very little or no Schade negative feedback (after all, it uses a very low distortion output tube, the 300B).
Just my opinions.
Your Mileage May Vary
Inside the glass envelope of the rectifier, it shows a very unusual drawing of the plates and the cathode.
The cathode is drawn as the traditional straight line of a plate.
It also draws the plates as slanted straight lines.
Look up almost any 5AR4 data sheet for a more proper and traditional way of drawing the elements.
2. Schade negative feedback:
When using a triode wired pentode (as this circuit does), or when using a true triode for the input tube . . .
If you bypass the cathode bias resistor (as this circuit does), the plate resistance, rp, is reduced; That reduces the amount of negative feedback.
That is why most Schade negative feedback circuits do not bypass the cathode bias resistor of the input tube.
My conclusion is that there is effectively very little Schade negative feedback in this amplifier, and I am not surprised that it sounds excellent with very little or no Schade negative feedback (after all, it uses a very low distortion output tube, the 300B).
Just my opinions.
Your Mileage May Vary
Last edited:
Interesting, thanks for sharing.
Replacing R1 with a carbon composition resistor and C2 and C3 with some film capacitors (DC link) would be on my list were I to build this.
Replacing R1 with a carbon composition resistor and C2 and C3 with some film capacitors (DC link) would be on my list were I to build this.
The schematic that now is attached to post #1 is the correct schematic.
The schematic in post #6 is the one that was attached to post #1 before but TS changed it (on this forum a topic starter can change his/her first post; there is no time limit for those changes).
To me it's weird that some call the way R8 is connected to the plate of the 300B in the schematic in post #6 Schade feedback (or however it's officially named). Yes, they have in common that there is a resistor going from the plate of the power tube to the plate of the preceding tube. But as far as I know the preceding tube also has to have it's own plate load to be able to work like with Schade feedback. That plate load is missing in the schematic in post #6. Imagine what effect that has...
The schematic in post #6 is the one that was attached to post #1 before but TS changed it (on this forum a topic starter can change his/her first post; there is no time limit for those changes).
To me it's weird that some call the way R8 is connected to the plate of the 300B in the schematic in post #6 Schade feedback (or however it's officially named). Yes, they have in common that there is a resistor going from the plate of the power tube to the plate of the preceding tube. But as far as I know the preceding tube also has to have it's own plate load to be able to work like with Schade feedback. That plate load is missing in the schematic in post #6. Imagine what effect that has...
Yes, Post # 1, does not have any feedback resistor, instead the input plate RL comes from B+.
I missed that, or else it was changed / or corrected.
But Post # 2, does have the Schade feedback resistor that comes from the output tube plate back to the driver tube plate.
Sometimes I see that Post @ 2 style of schematic with no RL; and sometimes I see a schematic that adds a driver tube load resistor, RL, that goes from the driver plate to B+.
Without the 3rd resistor (driver resistor, RL), the difference is that the Schade Feedback resistor becomes Both the feedback resistor, And the driver's plate load resistor RL.
Yes, that changes things:
1. The currents, when all of the driver's plate current is coming from the Schade feedback resistor, versus some of the current coming from the 3rd resistor, RL.
2. The 3 impedance ratio, that determines the input stage gain and feedback ratio is different:
A. The driver plate resistance, rp; the Schade feedback resistor; and the driver plate load resistance, RL; . . .
Versus
B. The 2 impedance ratio, that determines the input stage gain and feedback: the ratio of the driver plate resistance, rp, and the Schade feedback resistor.
That seems to me that it still conforms to the basic idea of using Schade negative feedback.
Just my opinions.
I missed that, or else it was changed / or corrected.
But Post # 2, does have the Schade feedback resistor that comes from the output tube plate back to the driver tube plate.
Sometimes I see that Post @ 2 style of schematic with no RL; and sometimes I see a schematic that adds a driver tube load resistor, RL, that goes from the driver plate to B+.
Without the 3rd resistor (driver resistor, RL), the difference is that the Schade Feedback resistor becomes Both the feedback resistor, And the driver's plate load resistor RL.
Yes, that changes things:
1. The currents, when all of the driver's plate current is coming from the Schade feedback resistor, versus some of the current coming from the 3rd resistor, RL.
2. The 3 impedance ratio, that determines the input stage gain and feedback ratio is different:
A. The driver plate resistance, rp; the Schade feedback resistor; and the driver plate load resistance, RL; . . .
Versus
B. The 2 impedance ratio, that determines the input stage gain and feedback: the ratio of the driver plate resistance, rp, and the Schade feedback resistor.
That seems to me that it still conforms to the basic idea of using Schade negative feedback.
Just my opinions.
Last edited:
So what will be the overall gain of this amplifier when the load resistor of the first stage would be connected to the plate of the power tube?
It doesn't have to be like this. When Gary Pimm experimented with Schade feedback in his 47 amplifier, he found that combining the two resistors, pentode driver plate load and Schade feedback, into one, makes better sound than two separate resistors.
FYI as drawn J4 is bridging a grounded short. Cathode current meter?
edit: how do you like the Intact?
edit: how do you like the Intact?
Thanks for posting the schematics. The first one is the earliest version of the 47 amplifier, which still has the two resistors. The later version used 6BA6 driver instead of 6SN7, and one resistor, essentially as in the Tabor.
At the page I linked he writes that he switched to a 6AU6 driver, perhaps that is what you meant. Updated schematic: