Mike,
Just two common cathode stages direct-coupled (cascade). A cathode follower would not result in 43dB of gain. No totem-pole... a single 5814A and no CCS or LED biasing, just high quality resistors and two film bypass caps. The first stage is the input amp... but it has little to no load beyond the plate load resistor as it's direct-coupled to the second stage... which is the driver, i.e., it drives the 45/2A3 grid.
Regards, KM
Just two common cathode stages direct-coupled (cascade). A cathode follower would not result in 43dB of gain. No totem-pole... a single 5814A and no CCS or LED biasing, just high quality resistors and two film bypass caps. The first stage is the input amp... but it has little to no load beyond the plate load resistor as it's direct-coupled to the second stage... which is the driver, i.e., it drives the 45/2A3 grid.
Regards, KM
Ah, of course the gain should have tipped me off. Sounds simple which I rather like. It sounds like careful selection of resistor values and operating points along with quality parts was your guiding principle.
Not to want to belabour,
but being inquisitive, I googled up some Ia-Va curves for 12AU7 and 6SN7 from the same source. At almost A4 size, I looked at behaviour with a load of 100K at +300V and bias about -5V. With 6Vpp input the 12AU7 showed 14% of non-linear distortion and the 6SN7 = 3,7%. The bias was about optimal for both.
I do not have a 6SN7, otherwise I would have tested in-circuit. Also a shock to me, but there then. (I have used several 6SN7s decades ago in power amplifiers, but never a 12AU7 at the time.)
As a typical 'audio tube' (?) the 12AY7 seems to be a better 'medium mu' (40) type; not sure why it was never really popular. RCA used neither in their typical application examples; GEC used mostly 6SN7.
but being inquisitive, I googled up some Ia-Va curves for 12AU7 and 6SN7 from the same source. At almost A4 size, I looked at behaviour with a load of 100K at +300V and bias about -5V. With 6Vpp input the 12AU7 showed 14% of non-linear distortion and the 6SN7 = 3,7%. The bias was about optimal for both.
I do not have a 6SN7, otherwise I would have tested in-circuit. Also a shock to me, but there then. (I have used several 6SN7s decades ago in power amplifiers, but never a 12AU7 at the time.)
As a typical 'audio tube' (?) the 12AY7 seems to be a better 'medium mu' (40) type; not sure why it was never really popular. RCA used neither in their typical application examples; GEC used mostly 6SN7.
The EICO ST70 uses the 6SN7GT in long tail pairs and works well.
Question is does it matter what brand to use?
I have tried Sovtek, as bad as its reputation I think even it works in this application where the signal level at the grid is large.
Tom
Question is does it matter what brand to use?
I have tried Sovtek, as bad as its reputation I think even it works in this application where the signal level at the grid is large.
Tom
Hi Tom,
If you don't want to invest in NOS tubes, the Elektro Harmonix EH6SN7 is a pretty decent tube.
If you don't want to invest in NOS tubes, the Elektro Harmonix EH6SN7 is a pretty decent tube.
hey Ray,
I'd offer another alternative, go shopping at 'other-than-usual-retail' outlets, such as Hamfest's. 6SN7 are quite reasonable there, even less than the usual current production cost...🙂
cheers,
Douglas
I'd offer another alternative, go shopping at 'other-than-usual-retail' outlets, such as Hamfest's. 6SN7 are quite reasonable there, even less than the usual current production cost...🙂
cheers,
Douglas
Hi Douglas,
Yes, I tend to forget that there are such things in the USA and Europe. Where I live, one is restricted to a very few local suppliers, plus what one can order from websites. Nevertheless, I could recommend EH6SN7s.
Yes, I tend to forget that there are such things in the USA and Europe. Where I live, one is restricted to a very few local suppliers, plus what one can order from websites. Nevertheless, I could recommend EH6SN7s.
None of those would make my list these days, how about 5687, 6CG7, 12BH7 or the Russian 6N6P? The 6SN7 as Douglas has suggested is another great choice. Of these I would most likely choose the 6N6P.
I'm using a cathode follower-buffered (2 sections) diff amp with constant sink for the diff amp to allow equal valued plate resistors on it. Works well for me, and balance is very good.
Responded because a newbie dredged up this long dead thread, and I did not check previous posting dates.. LOL Going forward I think we'll have a feature to retire necro threads after a certain point.
I see the "This thread has not been posted in for months, do you really want to poke it??" warning function does not happen here? That's certainly a first step to wake-up people (not just newbies) who come into a thread perhaps via Search and think they hit a hot topic ripe for posting.
I assume this new software has that basic function?
Yes, once Bumble1 had tagged a stale thread (perhaps with good reason), there's no real way to alert kev or eg that they might be wasting finger-energy.
I'm not real fond of trashing threads just because they are old. The old stuff may be the good stuff. I know several threads (elsewhere) that unhibernate every few years, refreshed. And Bumble1 does have a valid opinion about LTP tubes.
I assume this new software has that basic function?
Yes, once Bumble1 had tagged a stale thread (perhaps with good reason), there's no real way to alert kev or eg that they might be wasting finger-energy.
I'm not real fond of trashing threads just because they are old. The old stuff may be the good stuff. I know several threads (elsewhere) that unhibernate every few years, refreshed. And Bumble1 does have a valid opinion about LTP tubes.
No harm done! Look what Jan Didden achieved in this 11-year dead thread. https://www.diyaudio.com/community/threads/30mb6-data.165383/#post-6888562
I agree with PRR. A warning flag is a good idea but sometimes a revived thread just happens to have just what I need at the moment. Sometimes it is a good thing to dig again the wells our fathers dug in a sense.
Sorry . I seem to have caused issues opening this old thread. My fault did not check the dates.
Don't worry about it too much. If it is uninteresting to folks they can ignore it. BTW. ECC88 seems to model quite nicely in a LTP.