Print PDF
Favorite
Email
Step 11Power Amp Stage
The simplest tube amp type is Class A, single-ended. Without going into too much detail, Class A amps are considered to be the richest, warmest sounding type of audio amplifier. By their very nature, they tend to emphasize even-order harmonics, one reason why they sound so good.
"Single-ended" means to drive the output transformer from one side only--contrasted with a "Push-Pull" configuration, in which power tubes are driving the transformer from both ends (with current flowing from a center tap.) Push-pull is more efficient, but is more complex--the audio signal supplied to one tube must have a twin "mirrored," or inverted signal for the other tube. Hence the "push-pull" name. That requires a "phase inverter" stage, a necessary complication.
And there are limitations to Class A. It's a bit harder to get max volume from a "classic Class A" design (single-ended, cathode bias, etc.) than a Class AB or Class B amp, however.
One way to increase volume, but keep the simple single-ended topology is to add a second tube in parallel with the first. Again, this isn't as loud as a two-tube PP setup, but it's simpler. It's also easier to keep a single-ended amp in "Class A territory."
Historically, there are some commercial amps which used the parallel SE configuration--the Gibson GA9, and the "Gibsonette," to name two. These, plus the Angela link (see: How did this project get started?) were inspirational.
Note that the plates are simply connected together at the output transformer primary. It's that easy.
Grid stopper resistors were added, simply because they are on the Angela project, and the old Gibson schematics, too. Although the Gibson plans usually had only one (I wonder if some additional asymmetry results?) Is the source of oscillation interaction between the two power tubes, so only one grid stopper is needed?
There's more cathode bypass capacitance that I normally care for. I wanted a pretty fat sound. I certainly got what I wished for.
Note the cathode bypass cap switch. An attractive alternative: change the hard-wired caps from 40uF and 15uF to 10uF for each. Then switch in an extra 15uF on both with a DPST switch.
The cathode-bias resistors MUST be rated for 5 watts.
Biasing
This is a standard cathode-biased Class A setup. My biasing voltage is slightly less ("hot" bias) than is noted on the datasheets. Here it's a 150 ohm bias resistor.
The datasheets recommend 180 ohm for 200V, although one datasheet used 160 ohms. We'll stick with 150 ohm for now. There's no sign of red plating or any other problems. If it lessens the life of the power tubes considerably, I'll change it to 180 ohms...
Load Resistance Based on the Datasheet
Power tubes have a characteristic called "load resistance," which specifies a recommended output transformer impedance. The load resistance is listed on the datasheet:
Voltage of 6DG6GT : load resistance
110V : 2000 (ohms)
200V : 4000
(Again, sorry for the lost formatting.)
With B.1 voltage close to 190V, a load resistance around 3666 ohms is recommended. However, this value is for one tube.
Load resistance for two tubes is half the value of one, or about 1833 ohms. This is the theoretical value of the primary impedance for our output transformer.
Note: this is a guesstimate, based on the datasheet. In the next step, we'll actually find the load resistance mathematically...
Maximum Power Output
(The load resistances discussed here are for one tube--since this project uses two, then 1/2 these values are equivalent within the circuit.)
I originally estimated the power output from the datasheets as approx 7+ watts. But the example values in the datasheets are for polite amplifiers, where accurate sound fidelity is more important than volume. But guitar amps need distortion, so we push this one pretty hard.
So lets look at the load resistance vs. power output chart. The red line represents our load resistance, somewhere near 3500 ohms (remember, for two tubes, that's 1700 ohms.) Where the red line crosses the Po curve is our power output.
For a "driven" amp, the max output is close to 4.4 watts. In fact, any load resistance values between 2600 and 6000 ohms exceed 4 watts per tube. These values depend on a high p-p signal, a decent bias and a plate voltage of 200v.
We're at 190V, so it'll be slightly less than the chart. We don't really know the p-to-p output of the preamp stage, but the preamp is definitely hi-gain. And we are running the power stage with a "hot" bias....
We'll never know for sure unless it's bench-tested, but I suspect this amp is running above 4 watts per tube, over 8 watts total. It's safe to say this is an 8 watt amp.
"Single-ended" means to drive the output transformer from one side only--contrasted with a "Push-Pull" configuration, in which power tubes are driving the transformer from both ends (with current flowing from a center tap.) Push-pull is more efficient, but is more complex--the audio signal supplied to one tube must have a twin "mirrored," or inverted signal for the other tube. Hence the "push-pull" name. That requires a "phase inverter" stage, a necessary complication.
And there are limitations to Class A. It's a bit harder to get max volume from a "classic Class A" design (single-ended, cathode bias, etc.) than a Class AB or Class B amp, however.
One way to increase volume, but keep the simple single-ended topology is to add a second tube in parallel with the first. Again, this isn't as loud as a two-tube PP setup, but it's simpler. It's also easier to keep a single-ended amp in "Class A territory."
Historically, there are some commercial amps which used the parallel SE configuration--the Gibson GA9, and the "Gibsonette," to name two. These, plus the Angela link (see: How did this project get started?) were inspirational.
Note that the plates are simply connected together at the output transformer primary. It's that easy.
Grid stopper resistors were added, simply because they are on the Angela project, and the old Gibson schematics, too. Although the Gibson plans usually had only one (I wonder if some additional asymmetry results?) Is the source of oscillation interaction between the two power tubes, so only one grid stopper is needed?
There's more cathode bypass capacitance that I normally care for. I wanted a pretty fat sound. I certainly got what I wished for.
Note the cathode bypass cap switch. An attractive alternative: change the hard-wired caps from 40uF and 15uF to 10uF for each. Then switch in an extra 15uF on both with a DPST switch.
The cathode-bias resistors MUST be rated for 5 watts.
Biasing
This is a standard cathode-biased Class A setup. My biasing voltage is slightly less ("hot" bias) than is noted on the datasheets. Here it's a 150 ohm bias resistor.
The datasheets recommend 180 ohm for 200V, although one datasheet used 160 ohms. We'll stick with 150 ohm for now. There's no sign of red plating or any other problems. If it lessens the life of the power tubes considerably, I'll change it to 180 ohms...
Load Resistance Based on the Datasheet
Power tubes have a characteristic called "load resistance," which specifies a recommended output transformer impedance. The load resistance is listed on the datasheet:
Voltage of 6DG6GT : load resistance
110V : 2000 (ohms)
200V : 4000
(Again, sorry for the lost formatting.)
With B.1 voltage close to 190V, a load resistance around 3666 ohms is recommended. However, this value is for one tube.
Load resistance for two tubes is half the value of one, or about 1833 ohms. This is the theoretical value of the primary impedance for our output transformer.
Note: this is a guesstimate, based on the datasheet. In the next step, we'll actually find the load resistance mathematically...
Maximum Power Output
(The load resistances discussed here are for one tube--since this project uses two, then 1/2 these values are equivalent within the circuit.)
I originally estimated the power output from the datasheets as approx 7+ watts. But the example values in the datasheets are for polite amplifiers, where accurate sound fidelity is more important than volume. But guitar amps need distortion, so we push this one pretty hard.
So lets look at the load resistance vs. power output chart. The red line represents our load resistance, somewhere near 3500 ohms (remember, for two tubes, that's 1700 ohms.) Where the red line crosses the Po curve is our power output.
For a "driven" amp, the max output is close to 4.4 watts. In fact, any load resistance values between 2600 and 6000 ohms exceed 4 watts per tube. These values depend on a high p-p signal, a decent bias and a plate voltage of 200v.
We're at 190V, so it'll be slightly less than the chart. We don't really know the p-to-p output of the preamp stage, but the preamp is definitely hi-gain. And we are running the power stage with a "hot" bias....
We'll never know for sure unless it's bench-tested, but I suspect this amp is running above 4 watts per tube, over 8 watts total. It's safe to say this is an 8 watt amp.
| « Previous Step | Download PDFView All Steps | Next Step » |
1
comment
|
Add Comment
|
3
Oct 2, 2011. 12:27 AMLenny24
says:
Cool Instructable, but due to Electrical safety you should connect one lead of the Secondary side of the Output Transformer to ground. In case it fails, the Fuse will be blown instead of your Body.
Reply
 |
Add Comment
|
