Quote:
Originally Posted by SwizzyMan
Well to start they are both totally different tubes. The as7 is a simple low mu twin triode while the 6L6 is a high mu and high power pentode capable of 30 watts of plate dissipation (on a single plate) while the as7 only dissipates a grand total of 26 watts (keep in mind 13 per plate). The as7 is a lower power tube and is not nearly as efficient when used as an audio amp tube. For example, running a 6AS7 in a class A1 amp only gives an amplification factor of 2 while a 6L6 in the same setup has an amp factor of 8. We must take into account that a single pentode 6l6 puts out more than double the amplification factor than using two triodes thus making the 6l6 a more efficient and powerful tube with out the manufacturing costs of two separate triodes in one tube. RCA intended for this tube to be used in lower power circuits and had high power audio triodes such as the 6sn7gtb's which at less plate and screen voltage have an amp factor of 20 in a class A1 amp which is perfect for audio use. The as7 was never really meant to be used for audio circuits, its main use was as a series regulator in scopes and other test equipment such as signal generators ans supplies. Nowadays the audio fools see a nice big twin triode and automatically assume it would be great for their amps, none really do their research and look to see if this tube is really ideal as an amp tube.
Just my 2 cents....or more
|
I'm well aware of the fact that the 6L6 is a beam power tube and the 6AS7G is a dual power triode.
If we directly compare a cathode biased pair of 6L6s and a single 6L6 working in push pull with itself, the difference between the valves is not as terrible as you seem to suggest. Assume we want an output stage of approximately 10 clean watts. Assume the 6L6s are cathode biased with a plate voltage of 266 volts, a screen voltage of 266 volts, and 16 volts are being dropped across the bias resistor(s). For zero signal, the total screen and plate current is 130 mA, meaning you're drawing 34.58 watts from your 266 volt rail. Assume the 6AS7G is in push pull with itself with a plate voltage of 375 volts and 125 volts are being dropped across the cathode resistors. For the zero signal condition, you would be drawing 100 mA from the 375 volt rail, or 37.5 watts. That's only a difference of 2.92 watts. So in terms of plate/screen dissipation, you really aren't gaining much by opting to use the 6L6s.
Now let's consider the max signal condition:
For the 6L6s, total screen and plate draw from the 266 volt rail is 145 mA, or 38.57 watts. For the 6AS7G, total plate draw is 106 mA from the 375 volt rail, or 39.75 watts. Now the difference is only 1.18 watts. The 6L6 now looks even less attractive.
The 6L6s are theoretically giving about 13 watts, the 6AS7G is theoretically giving about 11. Both stages at 2% THD. The total heater current drawn by the 6L6s is 1.8 amps at 6.3 volts. The 6AS7G is drawing 2.5 amps. So you're giving up another 4.41 watts by using the 6AS7G over the 6L6s.
So what are you getting by using the 6AS7G? More benign distortion characteristics, and MUCH better damping factor. In fact, a 6AS7G in push pull with itself has a plate resistance far lower than the 2A3. That's impressive as hell. The plate resistance of a 6L6 is 22,500 ohms. If I'm designing a Hi-Fi amplifier, you better believe I'd trade away 1-3 watts of B+ draw, and another 4.4 watts of heater power, for excellent damping factor and intermodulation distortion that won't make one's ears bleed. If you had bothered to peruse the articles I had posted from RCA's Ham Tips, you would have found all of these advantages laid out in black and white by J.H. Owens, editor of Ham Tips, and D.P. Heacock of RCA's Tube Application Engineering Group. In fact, it was the incredibly low plate resistance that was exploited to make a better driver for class B modulators.
This all comes at a cost of requiring an ungodly drive voltage. The 6L6 needs ~36 volts peak grid to grid, but the 6AS7G needs ~250 volts peak grid to grid. That said, this is a small price to pay, and with such a high allowable grid leak resistance, a consequence of the design of the valve, you could probably coax a 6SL7 driver stage R-C coupled to the grids of the 6AS7 into swinging the required drive voltage. A mu of 2 isn't so terrible for an output triode. The venerable 2A3 has a mu of 4.2.
I really don't think you understand the concept of amplification factor. First of all, the amplification factor you quoted of 8 for a 6L6 is for when the valve is triode strapped. It is much higher when operated as a normal beam output tube. Output devices tend to have relatively low amplification factors. You aren't trying to get copious amounts of gain from the output stage; you can always get more gain from a 12AX7 or 6SF5 or 6SL7 gain stage out front. You're trying to develop a signal of high enough voltage at sufficient current on the primary side of the output transformer so that you can drive a 16 or 8 or 4 ohm voice coil. Consider a 16 ohm voice coil being driven by 10 watts. You need ~8.9 volts RMS at ~1.1 amps. The total power delivered by the device is not correlated with its amplification factor. If it were, we'd all be using 6SF5 single ended triode amps
As far as being used as a pass element in power supplies goes, yes, the 6AS7G was used in such capacity, but it was not the main goal in producing the valve. In fact, it is a rather marginal pass element. In some cases a 5881 or 6Y6G was used in pass service instead of the 6AS7G. In any case, the 6AS7G, like the 2A3 which preceded it in pass service, was rather quickly discarded in favour of even higher perveance valves.
The valve was built originally for TV damper service, this is known fact. This is what my original post was about, and indeed it was the point of this thread. The audiophiles are using it, and for a damn good reason too. Perhaps in time you'll figure out what it is.