I want to build a single ended amp with EL 34's as the output tube. I can not decide whether to use 6sn7 or 12ax7 as the input tubes. I have schematics for both. I've also noticed some manufacturers favour one or the other, such as Icon Audio (in the UK ) use 6sn7. Unison Research use 12au7 or 12ax7.
So my question really is, "does one sound better than the other?"
They are similar in price and readily available (in the UK) so that is not an issue.
I will be using the amp with an Arcam CD17 and Kef IQ5 speakers.
So my question really is, "does one sound better than the other?"
They are similar in price and readily available (in the UK) so that is not an issue.
I will be using the amp with an Arcam CD17 and Kef IQ5 speakers.
Dudleyperkins: Avoid use of the 12AX7 in a power amplifier. You don't need the high mu triode there. I would suggest using a 12AT7 (6201) input, a 12BH7 phase splitter, and a 12AU7 (5814) low gain driver direct coupled to the EL34's. I used this to replace the PWB in a Dyna stereo 70 and it cleaned up the sound immensely. I built the new board on a small piece of aluminum so it would directly fit in place of the Dyna PWB, (Equivalent military versions of tubes in parenthesis.) I didn't like the way the 7199 in the Dyna circuit affected the audio. If you like pentode input, try a EF86 (6267).
The 6SN7 is a rugged medium-mu triode, and can drive most of the standard push pull output tubes (6L6GC, EL34, 6550, etc.) operating in AB1 faithfully. You can improve audio quality by direct coupling stages and minimizing the use of plate coupling capacitors. If you modify a stereo 70, all the voltages are there including a negative bias supply. The 6SN7 can handle higher plate voltage and dissipation compared to a 12AU7. In brief, you want plenty of clean voltage swing at the output tube grids while the lower level tubes are operating well within the linear region of the operating curves.
I am assuming you know something about amplifier design...some of the old tube and output transformer manuals (i.e. Acrosound, UTC) have some good schematics.
The 6SN7 is a rugged medium-mu triode, and can drive most of the standard push pull output tubes (6L6GC, EL34, 6550, etc.) operating in AB1 faithfully. You can improve audio quality by direct coupling stages and minimizing the use of plate coupling capacitors. If you modify a stereo 70, all the voltages are there including a negative bias supply. The 6SN7 can handle higher plate voltage and dissipation compared to a 12AU7. In brief, you want plenty of clean voltage swing at the output tube grids while the lower level tubes are operating well within the linear region of the operating curves.
I am assuming you know something about amplifier design...some of the old tube and output transformer manuals (i.e. Acrosound, UTC) have some good schematics.
The OP is talking about building an SE amplifier.. The 6SN7 is an excellent choice based on its reasonably low rp and very good linearity. The 12AX7A will have some difficulty driving the miller capacitance of an EL34 should triode operation be intended even as an option.
The additional gain of a 12AX7A would allow for the application of some global feedback due to the higher gain if a single stage driver was contemplated. (Roughly 6 - 10dB depending on the required input sensitivity)
Another tube well worth considering would be the 5842 if you can get your hands on a few. Very linear, and with a low rp (<2K) that allows them to drive a capacitive load effectively.
The additional gain of a 12AX7A would allow for the application of some global feedback due to the higher gain if a single stage driver was contemplated. (Roughly 6 - 10dB depending on the required input sensitivity)
Another tube well worth considering would be the 5842 if you can get your hands on a few. Very linear, and with a low rp (<2K) that allows them to drive a capacitive load effectively.
Some comments, if I may.
Don't use the 12AT7 in SE circuitry, unless it's CCS loaded. It's a matter of too much 2nd harmonic distortion.
Kevin is spot on about the 12AX7 being a wimp, with next to no drive capability. That fact can be dealt with easily enough. Use a DC coupled ZVN0545A source follower as shown in the phono preamp schematic I've uploaded.
The 5842 and the somewhat similar Russian 6С45П (6s45p) can be quite useful, BUT they can drive you crazy with parasitic oscillation. My jest is that those types come close to oscillating stored in the cardboard box. Carbon composition stoppers of substantial value on each grid connection, a 100 Ω carbon film stopper on the plate connection, and ferrite beads on the heater wires are in order. The "in box oscillators" were engineered for use in grounded grid topology at RF.
Don't use the 12AT7 in SE circuitry, unless it's CCS loaded. It's a matter of too much 2nd harmonic distortion.
Kevin is spot on about the 12AX7 being a wimp, with next to no drive capability. That fact can be dealt with easily enough. Use a DC coupled ZVN0545A source follower as shown in the phono preamp schematic I've uploaded.
The 5842 and the somewhat similar Russian 6С45П (6s45p) can be quite useful, BUT they can drive you crazy with parasitic oscillation. My jest is that those types come close to oscillating stored in the cardboard box. Carbon composition stoppers of substantial value on each grid connection, a 100 Ω carbon film stopper on the plate connection, and ferrite beads on the heater wires are in order. The "in box oscillators" were engineered for use in grounded grid topology at RF.
Attachments
What about using a 6SL7 for a SE amp? Shannon from DIYTube already has one designed called the Clementine. I have one that i absolutely love. Check them out .... diytube.com :: Index
What about using a 6SL7 for a SE amp?
The triodes in a 6SL7 are very closely related to those in a 12AX7. They are high μ, high RP, and very low gm. Those characteristics make them poor drivers of substantial Miller capacitance "finals". A pentode wired O/P tube, with its small CMiller, can be OK, but any sort of triode is out.
Better to have valves and circuits which add nothing. 'Personality' is fine in a guitar amp but undesirable in a music reproduction system.
Then why use valves at all? A transistor amp will have less colour at a fraction of the cost...
Hi!
You should select the input tube based on the requirements it has to full fill in terms of gain and driver capability. This heavily depends on the overall circuit. I agree with previous posters about 12AX7 or similar not beeing the best suited to drive an output tube.
Lower rp tubes unfortunately often have too low mu. I found the 6N7 (both halves paralleled) a good compromise. Considerable gain at reasonably low rp for small output tubes
Best regards
Thomas
You should select the input tube based on the requirements it has to full fill in terms of gain and driver capability. This heavily depends on the overall circuit. I agree with previous posters about 12AX7 or similar not beeing the best suited to drive an output tube.
Lower rp tubes unfortunately often have too low mu. I found the 6N7 (both halves paralleled) a good compromise. Considerable gain at reasonably low rp for small output tubes
Best regards
Thomas
In my case it is a combination of nostalgia, and finding that my attempt at designing and making a solid-state power amp failed due to both output oscillation and loop oscillation. Used properly, valves can do as good a job as SS - in the sense that the music is not noticeably damaged.MerlinB said:Then why use valves at all?
I have little sympathy for those who strain for 0.001% THD using SS, or for those who seem happy with 10% THD using valves. Both seem to miss the point.
My main reason for me wanting a single ended amp is because I once owned a Fatman 182, though it proved unreliable in the end, the sound was superb. Even my wife enjoyed it, she said it was like listening to a favourite cd for the first time. It had KT88 output valves and I think 6NP1 triodes. I want something that will sound as good but using a readily available triode here in the UK.
Why drive a '57 T-Bird? A new Honda will get you to the same place at a fraction of the cost.
Actually that supports my argument, rather than countering it.
If absolute fidelity is your goal (e.g., a fast efficient car), why use valves (e.g., T-Bird)? Answer: fidelity has surprisingly little to do with enjoyment...
Actually that supports my argument, rather than countering it.
If absolute fidelity is your goal (e.g., a fast efficient car), why use valves (e.g., T-Bird)? Answer: fidelity has surprisingly little to do with enjoyment...
I converted the ferrari to model T in my setup by resistor loading the mosfets .
4P1L driver and resistor loaded mosfets for output.
Makes any other tube amp look like energyclass A+ .
Member
Joined 2009
Paid Member
just build an amp to where they have both sockets and a/b it.
+1
Actually that supports my argument, rather than countering it.
If absolute fidelity is your goal (e.g., a fast efficient car), why use valves (e.g., T-Bird)? Answer: fidelity has surprisingly little to do with enjoyment...
When going to absolute fidelity I prefer to combine tubes with SS devices.
Such marriage produces the best results.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
- Home
- Amplifiers
- Tubes / Valves
- choosing input valve