I'm finally getting around to posting an update.
Here are my circuit modifications for using 6CG7 tubes with both headphones and small speakers:
Amp
- 330R input resistor
- 47R CCS resistor or quality 100R trim pot (recommend Bourns 3296)
- Edcor GXPP10-10K output transformer with 6 Ohm secondary
- 4.7uF film capacitor B+ to ground near output transformer
- 6.8R 5W non-inductive wirewound headphone positive to common (recommend Mills)
Power supply, LC filter version
- C2: 220uF 350V (recommend Nichicon PW)
- C3: 100uF 315V (recommend Nichicon PW)
- R1: 0.45mH/16R choke
- R2: 500R 5W
Power supply, RC filter version
- C2: 220uF 350V (recommend Nichicon PW)
- C3: 100uF 315V (recommend Nichicon PW)
- R1: 100R 5W
- R2: 47R 5W
Loaded B+ is 200-210V depending on the specific tube. To achieve such a high voltage with this power transformer, filter resistance has to be reduced to almost nothing, so I increased the capacitor values to help compensate. Measured ripple is less than 4mV for the RC filter version.
Late into my design, I lucked out finding some 0.45mH chokes that measure approximately 16 Ohms and are only 43x33x35.5mm LxWxD. I ended up hot gluing them to the inside of my power supply chassis since I could only squeeze them in after the fact. Ripple is practically immeasurable for the LC filter version, nearly a straight line on my scope's 2mV setting.
The chokes are a very nice upgrade; however, either configuration will allow the amp to perform well sonically. The trick is simply keeping ripple low while keeping B+ over 200V if possible, for 6CG7 tubes.
In the amp stage, reducing the input resistor allows slightly more input signal to pass. I installed trimmer potentiometers in place of the CCS resistors and while adjusting their values, measured distortion at the speaker terminals. Approximately 47 Ohms provided the lowest THD and more current has not proven necessary to drive my speakers without distortion. I am using a pair of EPI M60 which are 2-way speakers, 6-inch woofer, DCR 5.5 Ohms (8 Ohms nominal), and ~90dB sensitivity. I decided to leave the trim pots in place since I used high quality Bourns 3296 types and they provide some flexibility to calibrate for other tube types able to run in this voltage range.
The output transformers are Edcor GXPP10-10K with 6 Ohm secondaries for running speakers. I measured a slight benefit increasing the local bypass capacitor value on the output transformers with the RC filtered supply since it is a less efficient filter at these resistor values. The LC filtered supply makes raising this cap value mostly unnecessary.
A well-matched 6.8 Ohm non-inductive wirewound resistor is installed from each headphone positive to common at the jack. The load presented to the tubes is between 5 and 6 Ohms across the 32-600 Ohm headphone range. The jack has NC contacts to switch the resistors out of circuit, and speakers in circuit, when headphones are disconnected. 2-5W resistors should be more than fine, but I had Mills MRA-10 10W available. Metal film is a less expensive alternative here.
Headphones are plenty loud by 2 o'clock on the volume knob. Speaker levels, given 1-2V line level input, are what I would call night time listening volume, or quiet daytime listening volume, sufficient for my musical needs. Best for small speakers in a near field monitoring configuration. The tuner I am using has a variable output in addition to the line level output, which provides slightly more volume—I measured better than 60dB at 1 meter playing rock radio. The amp will easily handle 4-5V input, so professional gear or a line buffer/preamp will work fine into the inputs, and give more output. Distortion with line level input is very low at less than 0.039%.
So, how does it sound? Marvelous! Thanks again to all participating in this thread and of course, our fearless leader, Bruce.
Login
Register
FAQ
