Quick Question (Driving 6L6GC's)

[sorenj07]

Just a quick one. Is it possible to drive a 6L6GC with just one tube (I'm thinking a 12AT7) per channel? I've seen a couple implementations:
http://www.dissident-audio.com/PP_6L6/Sch-V21.gif
http://www.geocities.com/MyDesign_99/Pp6l6eng.htm

but was just wondering about opinions. Using a minimum of stages would be a priority for me. Maybe a 12AT7 LTP with a CCS?

 

[G]

A 6L6xx is one of the easiest tubes to drive. You could do it with a 12AX7 if you wanted. a 6SL7/6L6xx UL would be a good sounding amp.

 

[Wavebourn]

Not so easy as EL84. I'd still consider drilling couple of holes for 9-pin sockets...

 

[jon_010101]

A Dynaco style front end with a pentode driving a triode split load would do the trick. Some interesting pentode-triode tubes out there. I find this topology quite elegant, but it is best when driving "easy" tubes. EL84s especially. But 6L6 would work too.

 

[ray_moth]

I have to say that I agree wih Wavebourn on this. An amp should be as simple as possible, but no simpler! A solution that might work for little tubes with modest drive requirements, like EL84 or 6V6, may not turn out so well for more powerful tubes.

A 6L6GC requires about the same signal voltage to drive it as an EL34, but the required grid-ground resistance of the 6L6GC is lower, especially if you use fixed bias. High impedance, low current tubes like 12AX7 or 6SL7 might struggle to drive it.

A pentode-triode front end like the Dynaco isn't a good idea either, in my opinion, because of the very limited swing of a concertina splitter. There are several front-end improvement kits on the market for Dynacos because of this; if they weren't needed, nobody would buy them.

A 12AT7 LTP is far better, but without a front end you won't get much gain. Probably the simplest topology to be effective is a Mullard arrangement, with Voltage amp - LTP splitter (12AT7 is fine) - OP tubes. You will still want a CCS for the tail of the splitter, if you want good symmetry, but don't use the LM317 (like in the second of your examples), because it has a poor frequency response.

 

[SY]

I'd agree with Ray, but have a bit of a quibble about the Dyna circuit. The split-load inverter can have plenty of swing (way more than the 70V or so peak to peak that a 6L6 needs), but the problem is with the pentode section. The distortion performance is quite mediocre and distressingly variable between different tubes of nominally the same type.

Courtesy of burnedfingers, I did some measurements of several 7199 and 6GH8A in the Dynaco circuit (pentode to split load), and was appalled by the distortion spectra; high distortion, high order, and every tube looked completely different making channel matching somewhat problematic. Lots of feedback is the only savior, and that carries its own set of problems. Swing, however, was NOT an issue.

 

[sorenj07]

Thanks for all the info! I'm not about to build myself a new amp any time soon, but it's nice to know. Am I right in assuming that a simple grounded-cathode gainstage tacked to the input of a 12AT7 input seems like a good combo? I suppose that a preamp with some gain would also work instead of that extra stage, it's just in a different enclosure, right?

 

[Wavebourn]

Sunn used such topology in their guitar amps (6AN8, pentode VAS powered from 320V, triode Concertina with 47K loads powered from 425V), but what is good for a guitar is not necessary acceptable for Hi-Fi. I've found for myself that the optimal combination is 6N2P (one half for a VAS, one for Concertina), plus one 6N1P as a differential cascade to drive 6L6 tubes. Williamson used such topology, but he had in my mind too many RC coupling networks, so for better low end stability it is better to connect all triodes directly, like Altec Lansing did in 1659 amplifier. Probably something between 6N1P and 6N6P would be better, but I still can't select such a driving tube...

However, if you have no more space you may still use something like 6AN8...