300B Pushpull Power Amplifier Project
UPDATE 1: 8/1/98
My next amplifier project will utilize a pair of 300B per channel in pushpull configuration for about
40Wrms per channel output. Driver circuitry will employ 6SL7 and 6SN7 in a differential direct coupled
driver circuit. Tube voltage regulation, and fixed bias will be employed. Operation with and without
loop feedback will be possible..
Design will start shortly...
This amplifier will employ the new Svetlana 300B DHT, which became available recently.
Visual inspection reveals a tube that appears to be virtually identical to the WE300B in construction.
Quality appears to be excellent..
UPDATE 2: 3/20/99
Over the last week I have gradually assembled the chassis, a few days ago I disassembled the power transformers, added new lead wires, and painted the end bells.. I also prepped and painted the Citation II output transformers during a warm week day..
The schematics are nearly finished, the basic design decisions have been made, and the amplifiers themselves are roughly 50% complete at this time. Still waiting for a lot of materials, anticipating their arrival this week.
On the technical end for simplicity I have chosen to use a fairly conventional cascaded differential driver circuit with a 6188 in the front end direct coupled to the 6SN7 in the driver stage...
The Svetlana 300B's are to be operated with fixed bias at a level of 70 - 80mA at +400V, each tube has individually adjustable bias current.
Individual center tapped 5.0V windings are provided for the filaments, and at least initially AC operation of the filaments is intended..
No global feedback will be employed, and the amplifer features both balanced and unbalanced input options.. (Some very limited local feedback is used however.)
The positive supply is tube rectified using a single 5AR4, pi filtered with a pair of 47uF 630V film caps and a 5H, 400mA choke. This feeds a 6336 series pass regulator tube with 12AX7A cascode error amplifier..
The -200V negative supply is solid state rectified, filtered by film caps and series pass regulated using a 6U8A with pentode as pass element, and triode as error amplifier..
The negative supply uses a dedicated winding on the transformer to allow grounding of the positive regulated output, and use of the negative rail as the (mirrored) regulated supply output..
Output tube protection against bias supply failure is included.
Regulated 6.3Vdc is used for heating the input and driver stage filaments..
Some surplus iron was used in this project:
- Citation II output iron...
- Citation II Chokes..
- Chicago PC8413 power transformers..
Some New Iron was developed for this project:
- DHT Filament Transformer: 2 X 5VCT @ 3A, or 2 x 2.5VCT @ 3A
- Bias and Filament Transformer: 220V @ 20mA, 8V @ 2A, 6.3V @ 300mA, 6.3V @ 600mA
A lot of thought went into providing adequate underchassis ventilation in what is a relatively cramped amount of space..
The chassis were custom made for me at great expense in time and effort as you can even tell by looking at the admittedly mediocre quality pictures..
UPDATE 3: 3/23/99
Quite a lot of progress has been made, most of the parts are finally here, and the one remaining pair of power tranformers will be in by the middle of
next week.. Hopefully I will be able to fit them in... I used film capacitors everywhere in the power supply except the raw bias supply filter capacitors,
and that was simply because I was running out of under chassis room..
A custom set of power transformers and choke will be available from KTA after development is complete, and will consist of a main plate supply transformer which will
source 250mA @ 800VCT, 6.3V @ 6A, and 5V @ 4A, a bias and filament transformer which will provide 2 x 5VCT @ 2A, 6.3V @ 600mA, 6.3V @ 450mA, 10V @ 2A, 10V @ 250mA, and 220V @ 30mA..
The choke will be 5H, have less than 50 ohms dcr, and have a 250mA rating.. Prices will be $510.00 for the complete set for two mono blocks...
I estimate that the amplifiers will be largely complete within 10 days, and then the debugging process can begin..
UPDATE 4: 4/11/99
The amplifiers are finally complete, and installed in my listening room where they drive my Magneplanar MG1.6 Quasi-Ribbon loudspeakers with ease.
The amplifiers functioned immediately, but the driver stage initially did not perform quite to my expectations, the problem was finally eradicated with a little thought and a few hours with
my solder iron. Performance now exceeds my expectations, and the amplifiers sound very good. The complete article with design will be published in the issue 12 of
"Vacuum Tube Valley." Please order a copy of VTV Issue 12 if you are interested in seeing the schematic.
UPDATE 5: 11/9/99
About seven months have passed since completion of these amplifiers, and I am still very pleased with their overall performance.
I have made some minor improvements to the amplifiers since their completion, the most significant of which would be the substitution of
Jensen paper in oil copper foil coupling capacitors for the RTX originally specified as well as a value change from 0.1uF/600V to 0.25uF/630V.
Other important improvements included (oddly) a small reduction in the 400V regulator loop gain in exchange for somewhat greater loop bandwidth (reduced error amplifier
plate load to 162K) and shunting Rfeedback with a 0.47uF film capacitor. The amplifiers have been completely reliable, and problem free since their completion.
UPDATE 6: 6/23/2001
More than 2 years have passed since the completion of these amplifiers, and you can see the entire Vacuum Tube Valley article at this location: Svetlana Tech Note # 57 Link is gone - see update 8 for exciting news!
The amplifiers have functioned well over the last few years with only a few minor modifications, including the replacement of the coupling capacitors with REL TFT types which significantly improved the perceived resolution of the amplifiers. In recent months I have used these amplifiers quite successfully with Silverline Sonatinas and currently use them with a pair of JBL Rhodes C37 with the annular ring horn tweeter. (Not the potato masher)They have been run in balanced and unbalanced mode with equally good results.
UPDATE 7: 10/28/01
I recently upgraded to the JJ300B after repeated and unrelated failures in the original output tubes specified for this project. They sound excellent in this design, producing a quicker, more detailed sound which I prefer. The factory matching was so close that I have no audible hum on the output.. Sound great!
Please note that I can no longer provide any of the power transformers listed in the article as my OEM supplier has gone out of business.
UPDATE 8: 12/20/05
This is the first new article to appear on this site in over 5 years. Please bear in mind that this article was written 6 years ago and the
opinions I expressed in this article may have changed. Read hopefully somewhat matured.. Something about failing at something
gives a guy more humility than expressed here..
This design depends heavily on its regulated power supplies for proper operation. Numerous copies have been built and the only one that failed
to function correctly was built without the regulated supplies. The driver stage is dc coupled and is optomized for 400V dc and because of the
large swings required at the 300B grids there is not a lot of margin to spare.
The original Svetlana 300B DHT mentioned in this article to my knowledge is no longer made, which given the lack of reliability is not as bad as it
may seem. There are many excellent alternatives including the JJ which I recommend and the WE300B which I have also used with excellent results.
This particular pair of amplifiers has proven to be extremely reliable over the last six years, only the original pairs of SV300B have been replaced.
Book version of the article originally published in VTV.
Link to schematics and article: Buy 300B Construction Article at LULU
Click Back to navigate back to this page.
The design shown on these pages may not be used for commercial purposes without the express
agreement of the author and without a licensing agreement in place.
This page created 8/1/98. ©1998, 2001, 2005, 2008 By Kevin Kennedy