Well, the first thing that comes to mind, is that if you can hear EVERYTHING that's going on before the regulator, then what you have at hand is something other than a regulator.
Could you provide some example schematics of what you've tried in experiments?
I completely agree! Output impedance of 30 or 20 ohms is very good - output impedance of 2 ohms is much much much better, sonicly, in my experience.
Hi Ian,
No, I was simulating a simple triode gain stage, as per Jeff's KISS plan. The idea was to point out limitations of doing so.
No question the boostrapped follower is a very good solution. For low plate voltages as required here it's probably the best (without using any SS parts..).
Cheers,
GB
I get that. However, I think it's good to be open minded, and regs are a really great tool to have in a toolbox, when looking to get the very best results.
We can debate about sonic effects of different not-100% feedback loops; (I myself make only no gNFB amps) however in the case of modules with a 100% feedback loop, there isn't realistically anything to talk about regarding sonics.
Hi Jeff,
with these parameters, if I simmed right, HF roll off at 20 kHz will be around 2.0 dB (before any additional roll of from the OPT).
If you want to go single triode input/drive stage in this application, it seems you should look somewhere else than 12ax7/6sl7. Their impedances are too high.
GB
I also appreciate regulators, CCS loads, source followers and all that. Allows you to squeeze out the very best of triodes as voltage amplifiers.
The discussion here however was how to make a great amp in absence of all these? With a minimum count of active parts, overall and in the signal path.
A source follower is 100% FB, a regulator probably not. If the amp doesn't draw a constant current at any time, the regulator becomes part of the audio path.
Or use 12ax7/6SL7 with bootstrapped follower approach... Its easy to try out in any case since the 2nd triode is just sitting there in the envelope waiting to be used.
Hi Bigun,
The amp needs to be wideband no matter WHAT upper frequency you think you can hear !!
Its a falsity to equate your top frequency response with the audio amp's required response !!! And the speakers need to be wideband too, with tweeters !!!
What you CAN hear, below 10 KHZ, or even below 5KHZ, contains combinations of higher-than-what-you-can-hear RESULTANTS, which play down into the range where you do hear. The amp and music playback will always sound MOST REAL, when it plays back ALL the original information, into the areas where you do hear !!!
Jeff Medwin
Member
Joined 2009
Paid Member
Hi Jeff,
Yes indeed, I wouldn't want to build a sub-par amplifier just because I no longer have the ears of a teenager !
However, a couple of dB down at 20kHz doesn't concern me one iota.
However, I'm wondering if the 12AT7 wouldn't be worthy of consideration here, for the single driver option it would eliminate any concerns of h.f. roll-off for those that have them - which is especially true for the L-W topology where the input cathode is un-bypassed.
Yes indeed, I wouldn't want to build a sub-par amplifier just because I no longer have the ears of a teenager !
However, a couple of dB down at 20kHz doesn't concern me one iota.
However, I'm wondering if the 12AT7 wouldn't be worthy of consideration here, for the single driver option it would eliminate any concerns of h.f. roll-off for those that have them - which is especially true for the L-W topology where the input cathode is un-bypassed.
Last edited:
Well, I owe GoldenBeer an apology - for not looking sooner at a EC86/PC86. I also owe you, Mach1, a big thank you. Woke me up. The 6GK5 looks to be better, far more linear than a 12AT7, and I LIKE the high mu. Thank you, to both of you guys.
Time to do the direct couple and work out the amp's values. TAC for LW hum cancellation should be easy to achieve I am guessing !!
For my B+ filtering, I made minor changes in cap values, to account for the extra current over the 0.0007 A. I was going to use in the 6SL7.
I also prefer 9 pin miniature tubes to octals, all else equal. A late night with the calculator and graphs tonight will ensue !
Jeff Medwin
Last edited:
The Sound of the Machine
Rupert Neve had something to say about this
RUPERT NEVE QBE
see the article in Audio Technology Magazine's Issue 1, March/ April 1998
our basis for determining "what we can hear" is clearly bunkum.
Jeff,
you should first decide on your power tube. The 6AH4 doesn't require as much swing on the grid for full power, and you can choose from a much larger selection of medium mu tubes for you input.
If you want to go with the 45, it requires even more swing on the grid than a 2A3, and you're pretty much stuck with hi mu tubes. Otoh, it's input capacitances are very small, and you can probably get away with a single half 6SL7 (or maybe 12AX7).
you should first decide on your power tube. The 6AH4 doesn't require as much swing on the grid for full power, and you can choose from a much larger selection of medium mu tubes for you input.
If you want to go with the 45, it requires even more swing on the grid than a 2A3, and you're pretty much stuck with hi mu tubes. Otoh, it's input capacitances are very small, and you can probably get away with a single half 6SL7 (or maybe 12AX7).
Member
Joined 2009
Paid Member
There are many really nice triodes in the RCA manual last time I looked - but I thought you were trying to find something in current production - maybe I'm mixing up your projects ?
Member
Joined 2009
Paid Member
This reminds me how the SRPP driver seems to come in and out of 'fashion' - I've never tried/heard it myself and feel my education is lacking until I do 🙂
AES usually has 6GK5 for around $7. 2/3/4GK5 are even cheaper if you can find them. You can get a lifetime supply for pocket money.
Another possibility is the E180F/6J9P, triode strapped. Mu =100, Ia 10mA with Va=150V.
I've been doing some simulations for screen drive and it looks promising with a current source in the anode.
I've used it in the past in this configuration driving a triode strapped 6P41S.
I've been doing some simulations for screen drive and it looks promising with a current source in the anode.
I've used it in the past in this configuration driving a triode strapped 6P41S.
Very nice triode wired, but I remember a lower mu; somewhere around 60 I think. The D3a is higher.
I think you are right on Mu being closer to 60.
I was going on simulation results as I've seen no datasheet for triode mode.
I was going on simulation results as I've seen no datasheet for triode mode.
Member
Joined 2009
Paid Member
Does the Loftin and White circuit compensate for filament hum or are you implementing DC heating? On Correlation Between Residual DHT Filament Hum and AC Frequency. Distortion-induced hum in directly-heated triodes.
- Status
- Not open for further replies.