top of page

The 807 in P-P AB2 at about 40 watts

bob6542

Updated: Jan 7

I always liked the look of the very old 807 tube and so decided to create a design based the 807. The 807 has a reputation of being a "difficult" tube to design a high quality amplifier with. Some of this seems deserved and some may be due to the 807 being such a old part available in the early days of audio design when many troublesome design were created.

Part of this project is to see what can be done with a pair of 807 in a more modern design and part of the project is just to create something to look different as I personally enjoy the "romance" part of the audio experience and well as the music.

This project is still a "in play" as I seems to have far too many projects to finish.


The 807 plate supply is about 500V loaded at 6.6K ohms PP.

The 807 screens are run in UL mode to reduce distortion.

This allows 40 watts RMS in class AB2 mode.



How can the screens run in UL mode with a 300V maximum screen voltage on the 807 and a 500V plate supply?


First the 807 300 volt maximum screen voltage is a conservative rating so a small amount of rating push is in order. I set the screens at a regulated 350V DC.

This regulated screen supply also locks in the bias current with line voltage changes.

At 500V on the plates we are working these 807s so bias stability is important.

I set the screen to 350V by capacitor coupling the UL AC signal from the transformer to allow a voltage shift between the plate DC supply and the screen DC supply. I also use a voltage divider to scale the size of the UL AC signal to the lower DC screen voltage to get a best amount of negative feedback on the screens. This lets you "tune" the UL level in the amplifier without needing different transformers. The voltage divider resistors also work as a effective screen stopper helping to the keep the 807's under control. This was effective at reducing some of the 807's "bad habits"


The 807 are driven by a 6N5P cathode follower with high S to allow a minimum of distortion to be introduced by the screen current when the 807 enters class AB2 mode.


The 6N5P cathode follower is driven by the second gain stage and phase inverter stage using a 6N3P. This tube has good gain and reasonably low plate impedance. This long tail pair stage does not have a current source in the cathode so the match between the two outputs is not perfect. A trim pot is provided to allow adjustment for drive matching of the 807s. The result is the balance in the output stage can be dialed in for minimum THD at lower power levels. Of course good matching of the 807 pair is very helpful as well in reducing THD.


The input is a single ended pentode stage using a 6AK5. This allowed high gain with good linearity in a single gain stage and no coupling capacitor to degrade low frequency stability was required.

Negative feedback from the transformer secondary is applied to the cathode of this stage. This means a low leakage inductance transformer is required. I used a Hammond part with the old school secondary arrangement for impedance selection.



Here is were it is today.



Assembled PCB ready to test. Great looking tubes.

Blank PCB. Note the large holes for the old school tube sockets.

No PCB mount for these sockets!

The test bench ready for testing. Fingers crossed.



THD at 1 watt is 0.0427%. Pretty good for 1930s technology.


THD at 22 watts is 0.109% and the residual THD is pretty smooth. We are still in class AB1 at 22 watts.

At 40 watts THD is 0.241%. The residual is looking a bit choppy as we have entered class AB2 at this power level. There is a shift in current gain as the 807 enters positive grid voltages and this leads to additional THD. Still this if far better than staying in AB1 and simply clipping. Not bad for a tube designed in the 1930s!

There is the tested PCB mounted in the chassis.

The small tubes with a plate cap are 6C10P Russian damper tubes being used for the main rectifiers in the power supply. They are low cost (1$), have low high voltage drop like the 5AR4 and unlike current production 5AR4's are very tough. You need two for full wave and they need a 6.3V filament supply so this is a complexity with a 5V winding being standard.

Bottom view. Note there are a number of regulated power supplies PCB around the edge. It is a bit complex.

Top view. Love the look of the 807! Note the large ferrite beads on the plate wires.

The circuit uses 3 low level tubes per channel.



Circuit Diagrams




 
 
 

Comments


ABOUT FEEDs & GRIDs

I'm a paragraph. Click here to add your own text and edit me. It’s easy. Just click “Edit Text” or double click me to add your own content and make changes to the font. I’m a great place for you to tell a story and let your users know a little more about you.

SOCIALS 

SUBSCRIBE 

I'm a paragraph. Click here to add your own text and edit me. It’s easy.

Thanks for submitting!

© 2023 by FEEDs & GRIDs. Proudly created with Wix.com

bottom of page