Typical Single End 300B Design
For comparison I tried a typical 300B output stage in the SE Amp Cad, and made some calculations for the typical driver stage (6SN7 or 5687). The 300B has cathode bias and the B+ rail is at 480V. This leaves the 300B with 387V at 80mA. The load is 2.5kOhm.
The most interesting data are the output power (9.2W), the 2nd harmonic (6.3%), the 3rd harmonic (1.2%) and the output impedance (2.4 Ohm). At the more important 1W output the 2nd harmonic is still high with 1.6% and the 3rd is at 0.1%, This is nearly as good as the figures I get at full power in my design. From the loadline below you can see that this load is optimized for maximum power and that there is no margin for lower load than the stipulated. This can be a reason for the criticism often heard about these amplifiers with difficult loads.
The transfer function looks OK, but can be much better. If you wish to make a good push-pull design with 300B, I think a higher load is necessary because the transfer function is uneven. Most PP design use even lower load.
The typical driver stage in this kind of amplifier is a 6SN7 with 33 kOhm load.
The result 1.6% 2nd harmonic is probably a bit better than what is usually the case, because of the high (480V) B+. With a 5687 as driver the distortion will be higher, but the output impedance much lower. Probably 5687 is a better driver in this kind of amplifier.
Driver Stage Combined with Output Stage
The driver stage is connected in anti-phase with the output stage, which reduces the 2nd harmonic, and the result for the amplifier will be: 4.7% 2nd harmonic and 1.2% 3rd harmonic at full power (9.2W) and the output impedance is 2.4 Ohm. With a 5687 as driver the 2nd harmonic will probably be at least 1% lower. This looks decent, but the 300B needs 1mA at 0V on the grid which means that the input impedance is only 65k with a 330k grid resistor. The driver stage is then loaded with 22k instead of 33k, and this will cause distortion (mainly high order) in the driver stage.
This is not a bad amplifier. The distortion at lower power is low enough, but it is no power horse and difficult loads are out of the question. The output taps normally equipped (4, 8 Ohm) shall be considered the lowest possible, if your speakers gets below 8 Ohm use the 4 Ohm tap. If used on the 4 Ohm tap this is probably a very good amplifier, but the driver stage will need some more voltage at this load. The result is OK but I think it is possible to do much better.
Measurements on 'Typical' 300B SE Amplifier (Spark 743D)
The Spark 743D uses a 6SN7 cathode follower as driver, but this is in phase with the output stage leaving the preceeding stage for some 2nd harmonic cancellation. This is resistance loaded (transformer load would give much lower distortion) which means that the cancellation will occur as if it was connected directly to the output stage, but there is some feedback in the driver section of this amplifier and this alters the characteristics a little. I think it is interesting to look at the figures anyway. The cathode follower driver is capable of driving the 300B without severe problems, quite a good design. The review on this amplifier was very positive (High Fidelity 1/1999 Swedish edition). This amplifier works at lower voltage (360V) than my example, and this may explain why the behavior is a bit different.
The distortion (8 Ohm tap) is quite high and as you can see it is almost exactly what the SE Amp CAD predicts. Note the very high distortion into the 4 Ohm load (remember the margin for lower impedances?). The output impedance was measured to 2.6 Ohm at the 8 Ohm tap and 1.5 Ohm at the 4 Ohm tap.
The FFT plot shows the spectra at 5W into 8 Ohm. The 2nd harmonic is -33dB and the 3rd is unfortunately masked. The SE Amp CAD results at 5W is found below. The driver stage will probably still deliver 1.5% 2nd harmonic which cancels out some of the 2nd harmonic in the output stage, giving the result 2.4% (-32.3dB) which is spot on. The 3rd harmonic is at 0.5% (-46.4dB), not alarmingly high but can be better (simply switch to 4 Ohm tap).
Measurements on another 'Typical' 300B SE Amplifier (Cary 300SEI LX20)
Below is another example of distortion cancellation, the lower plot is actually at 10W. Output impedance was measured to 2.6 Ohm. The distortion cancellation at 10W is almost as large as in push-pull amplifiers (indicating high distortion in the driver stage), but the result is good even if the 3rd harmonic is too high, and at 1W the result is excellent. The used output tubes are the Kron KR300BLX20. This plot is taken from HiFi News April 1999. The review was very positive: 'Within its power compass the 300SEI LX20 is up there with the world's best'.
This site uses frames, but if you have found this page via a search engine and do not see any frames, you can use the button below to get to the main frame of the MT Audio Design site.