[CButterworth]

With my AKSA 55 (standard Nirvana upgrade) and new (to me, bought used) Polk RT600i speakers, I am really impressed by my system.  In fact, every time I add a new component, the sound is so noticeably improved that my spouse's jaw even drops in amazement.

Yes, I am aware that the Polk speakers are real high-end, or maybe even close by other folk's standards, and the CD-player - the cheap Toshiba that other forums have been raving, is likely not up to a higher-end player (I have my eyes on the Njoe Tjoeb).

Anyway,  I have been musing with the idea of adding some active component to my homebrew passive preamp.  One thought of mine is to take a popular headphone amp design (the CMOY op-amp) and build two on individual boards which could be powered individually.  This would ensure the same channel separation as the rest of my system.

So, besides going for a tube preamp such as those designed by Hugh, which add warmth to the sound, would I really gain anything from building op-amp preamp circuits into my passive preamp.  I would likely chose to buy one of the better Burr-Brown op-amps.

Sorry for the long preamble, although opinions may be more accurate when folk know a little more about my system.

Regards,
Charlie

[PSP]

Hi Charlie,
While I was building my TLP-N, I listened to my AKSA 55N in a passive configuration (I used a DACT attenuator and a source selector inside the AKSA enclosure, so signal paths from the input RCAs to the selector, to the DACT, and then to the AKSA amp boards were very short, say 2-3 inches total for the CD input.

I really liked the way this system sounded, but when I put the TLP in the chain I liked it a lot more.  I suppose it is accurate to say that "tubes add warmth" but that is like saying that "tubes add fuzz, distortion, and junk to the sound".  When I listened to my system in a passive configuration all sounded correct and polite.  When I put the TLP in the system it had more balls and was much more involving... I often read while I listen, and after I put the TLP in the system I found myself putting the book down to listen to the absolutely glorious sound.  I didn't do that nearly as much with the passive setup.

Of course, YMMV and each of us likes and listens to different things in music.  You may well like the passive configuration better, but you will have to hear it for yourself and decide.

good luck,
Peter

[AKSA]

Charlie,

I'd like to affirm Peter's comments.  It's actually quite a strange comparison!    

I was dragged screaming into designing an active preamp.  I didn't want to do it.  My feeling was that it would simply add distortion with all that additional signal processing.  Of course, this is true, but what I didn't count on is the human perception of music.

Anyway, with doleful intent, trying to keep the peace, I set off on this epic journey.  First came the TLP, then the GK1.  If it hadn't been for Darl Singh in New Zealand I probably would never have finished it as quickly as I did.  Thanks Darl!     I would lie awake in bed thinking of how to do this preamp thang properly.  

I knew tubes added something 'organic' to the sound, a sniff of something 'human' and 'sentient' to the presentation.  I'd done a fair bit of work with tubes, notably 6SL7/6SN7/5687/6DJ8/12AT7/12AU7/6BQ7A, and decided to use a frame grid tube from the 6DJ8 family, as it was readily available, ran a highish current, worked well as a cathode follower, and was known to sound good.

I settled on the cathode follower configuration, because short of using a very much larger tube I could not achieve the low source impedance necessary to drive long cables.  Since a CF has unity gain (or very close to it) I needed a gain stage.

I didn't want to use a tube for the gain stage, and it too had to have a very low Zout in order to minimize the invariably bad effects of the attenuator.  Another tube might have been too much of a good thing.  So, with Darl doing exhaustive testing, I did a highly refined version of a discrete opamp, modelled on both the AKSA and the famous Jensen 990 from the eighties.  I'd done work on this circuit block back in the late nineties, and remembered how transparent and fast it was - absolutely no coloration at all.  The difficulty was always in passing good imaging information through this circuit block;  something Darl worked on tirelessly with me until we got it right.  Using just this SS gain block, we eventually jagged it in about four months, then moved to the trick CF GK1 output stage.

We discovered that it was possible to sculpt the tube coloration with novel circuitry.  This was particularly important in the bass, where tubes traditionally fall down.  I found it was possible to set up a special drone circuit which compensated the tendency towards 'floppy' bass.  The circuit worked well, so well in fact that the bass of the GK1 sounded SS.  Heck, we didn't want that, so we backtracked!!  We realized that in truth most people like a little 'wetness', or 'floppiness', or warmth, to their bass, and it all seems to depend on your terminology and how much whisky you've just consumed....  Anyway, we carefully selected precisely 60% compensation below 150Hz and in some pretty careful tests this was adjudged to be just right.

The result of all this was the GK1;  it's more sophisticated by an order of magnitude than the TLP, which is still pretty good, but the bass on the GK1 is much tighter, just as warm, but with more authority.  Imaging is superior as well.  The sense of 'being there' is greater;  all this from additional circuitry - it really goes against the grain, but there it is.  We humans like our music with makeup, no doubt about it......   After all, why should music be any different to the other great pleasures in life?    

Cheers,

Hugh

[mb]

...
Anyway, I have been musing with the idea of adding some active component to my homebrew passive preamp. One thought of mine is to take a popular headphone amp design (the CMOY op-amp) and build two on individual boards which could be powered individually. This would ensure the same channel separation as the rest of my system.
 ...

IME there will be a significant change in sound when adding a CMOY or PPA-style active stage or something like Aunt Corey's active buffer. I've found that even the very friendly 47k load of the AKSA benefits audibly when driven by a low impedance source (3 ohms in my case). A poorly implemented CMOY will sound grainy, harsh and fatiguing (that's how my first attempt sounded, even with battery supply). A carefully implemented opamp linestage or buffer can sound very clean, smooth and polished while adding drive the attenuator. Sounds very much like adding 4x power to the AKSA. FWIW almost everyone who's heard my buffered pre considers it very "tube-like". For the real thing go for the GK-1.

An active / passive toggle will give you best of both worlds and the chance to audition until you're convinced either way. I omitted that toggle because for my application battery-powered buffering was best to my ears. I had identical results adding buffers to the passive attenuator sections my gainclone chipamp and my Creek 4330.

[Guan]

I was using a DIY tube DAC with a built-in DACT driving my amp. It sounded clean, detailed and balanced but after a while, I noticed the sound just wasn't involving. After hearing a couple of tube pre-amps, I found one (a rather pricey ARC REF2MkII) that sounded more 'neutral' than most but added that intangible emotive factor and liquidity to the music. I'd love to hear the GK-1.

I knew tubes added something 'organic' to the sound, a sniff of something 'human' and 'sentient' to the presentation. I'd done a fair bit of work with tubes, notably 6SL7/6SN7/5687/6DJ8/12AT7/12AU7/6BQ7A, and decided to use a frame grid tube from the 6DJ8 family, as it was readily available, ran a highish current, worked well as a cathode follower, and was known to sound good.

Have you considered the 6H30 tube in current production? I read that it has good electrical characteristics and capable of putting out 6V of pulse current at each cathode.  My ARC pre uses 6 of these tubes and sounds quite clear and dynamic. Downside is that there is no 'flavor of the month' 6H30 choice for tube rollers as there is only one manufacturer.

[CButterworth]

Thanks for the opinions,  I certainly will try to convince my spouse about the benefits of a preamp (maybe the TLP or even the GK).  I will go back and investigate sonic differences using two of my currently constracted headphone amps.  I have two:

CMOY - a simple op-amp based amplifier.  It is relatively small
META42 a buffered preamp which is biased into Class A.  It is larger and uses premium parts - Elentec buffers, Wima and Black Gate caps and Vishay resistors.  I added a crossfeed board, although it can be bypassed.  It uses a fairly basic op-amp- Burr-Brown OP2132PA, which could be changed to one of the higher quality Burr-Browns such as the OPA627.  At present, the sound is crystal clear when listened through Grados.

My concern about using either of these, is that the L and R signals are amplified using a single power-source, hence, they're not totally independant.  However, I suppose that the L and R channels don't necessarily have initial source amplification in the CD player, and the "real" amplification that takes place in the AKSA probably totally masks any channel interaction that occurs earlier in the system.  DOES THIS SOUND REASONABLE?

Thanks,
Charlie

[AKSA]

Charlie,

You are right.

Modern op amp designs have huge power supply rejection ratio, often over 100dB, and this means that the amplifying elements are blissfully unaware of peturbations on the power supply.  So, if two opamps are powered from the same rail, there will be no interaction as their PSRR ensures one can't 'see' the other.

It is not so easy to do this with power amplifiers, because the speaker currents run into the amperes and interactions at this level of current tend to cause serious voltage sag on the rails, which affects adjacent channels.

By way of explanation, the GK-1 phono uses a dual opamp with a bipolar, regulated power supply.  But, the PSRR of the opamp is so high that there is absolutely no interaction, and those who have heard the GK-1 phono will testify that the channel separation and imaging are exceptional.  So you need not worry on that score.

Mervin (MB, from Singapore) has done a lot of work with opamp buffers, and may like to comment here on your META42.  He really likes them for their slam and transparency.  However, as Nelson Pass once said on DIY Audio, 'If you want it to sound like a tube, use a tube.'  This has been the tautology behind the GK-1!!

Cheers,

Hugh

[mb]

Charlie, it might be interesting to use just the output buffer section of the Meta42 to buffer your passive. That's essentially my config except I use the Intersil buffer instead of Elantec (cheaper...). There should still be plenty of volume as the AKSA is sensitive enough.