6SN7 vs. 6SL7 in Darling amp

[MrCurwen]

I am planning on converting my EL84 DC to a Darling DC. My main interest is headphones, as such I have no need for the extra power, and in my experiments with 1G6G and a couple similar triodes I have found the true triode sound to be more to my liking. Also just for fun and trying out things.

My question is: for the preamp tube, should I use 6SN7 or 6SL7?

The desing is limited on the basis that there is one octal socket there, and not room for another.

Aside from the gain, are there any significant differences? How much drive does the 1626 need? I've heard good things on both of these tubes, which one is your favourite, and why?

 

[Miles Prower]

Aside from the gain, are there any significant differences? How much drive does the 1626 need? I've heard good things on both of these tubes, which one is your favourite, and why?

The 6SL7 one of those types that feature high DC and AC plate resistance. The purpose behind this is to make a type that can operate with a plate current of ~1.0mA, or even less. For the 6SL7, Pd= 1000mW; for the 6SN7, Pd= 3.5W (5.0W for the 6SN7GTB). That way, you can use very large plate resistors to get the gain up without having to drop an insane amount of DC across the plate resistors.

This also makes the type suitable for driving Hi-Z, Lo-C loads only. The control grid of a power triode certainly is neither. The design I did that used a 6SL7 (as an LTP phase splitter/voltage amp) operated them into cathode followers which got the capacitance down (no Miller Effect) and the input impedance up.

The 6SN7 has less gain, but also can stand up to much higher Q-Point plate currents. This makes the type more suitable for driving loads that might be too much for types like the 6SL7.

 

[MrCurwen]

So 6SN7 would be the choice. Under which conditions would I get sonically best results? Mainly at which current would I get the lowest distortion?

I'm probably going to have B+ at 350V and grid of 1626 at about 100V.

I seem to recall 8mA as being a good place to be with 6SN7, is this so?

 

[mach1]

I'm not sure that a 6SN7 would have adequated gain to drive a 1626 without a pre. A 6SL7 certainly does. Why not stick to the 8532 used in the original ?

 

[MrCurwen]

mach1, well because I'm not doing a scratch build, but rather experimenting with an existing, working device (which I have built myself). It currently has EL84 single ended directly coupled power section, and half a 6SN7 for each preamp. The build is rather compact, there's no way to fit in more holes on the chassis. (Of course the EL84 holes can be easily enlarged for the 1626 octal sockets.) So on the preamp I'm pretty much committed on the octals.

Is there anybody who has built a 6SL7 Darling? Any opinions on whether or not it's a good idea?

If I go with 6SL7, how should I set it up (again, mainly at which current level) for optimum distortion and sound?

 

[Cassiel]

A 6SL7 is a good idea as is the 6sn7, same pinout so it's pretty easy to try both and find out witch one is best for that design. DC coupled amp so you'll have to tweak the plate resistors to match the DC voltage for the grid of your output tube.

 

[mach1]

MrCurwen,

mea culpa, I didn't read your first post fully 😀. The answer is yes, it has been done; if you google around a bit you will find a few darlings using the 6SL7:

http://home.att.net/~abraxasaudio/projects/d2/d2r2.html

http://www.lebong.de/darling/index.htm

If you wanted to be really adventurous you could even use a 717A (pentode):

http://www.jogis-roehrenbude.de/Leserbriefe/Reinhoefer-Darling/darling1626.htm

Some other octal pentodes would also be suitable (6SH7 [as a pentode or triode], 6AC7 and 6SJ7 spring to mind). Audionote are driving 300Bs with triode wired 6SH7s, which would probably have just about enough gain to drive a 1626 without a pre.