More on SET & tube verses SS

[artto]

I see the SET thread got locked. Wouldn't you know it! Just when I decide its time to respond.

OK, now Im going to throw another stone in the wheel, so to speak.

<?xml:namespace prefix = o ns = "urn:schemas-microsoft-com:office:office" />

A while back I had some correspondence with Mark Deneen regarding this. Perhaps (hopefully) Mark will recall and elaborate on this as I'm still not sure why this works. It has to do with the gain level settings of the source component, preamp and power amplifiers. Depending on what kind of equipment you have, you may or may not be able to perform this little experiment.

The other night I decided to revisit this experiment again. Its something I noticed a very long time ago, when I bought my first Klipsch (Cornwall) speakers. I decided to use a worst case scenario for this test so I employed what may cause some of you may freak out, or at the very least gag. I used my everyday, always on stuff that I dont have to hassle with re-routing signal paths or waiting for tubes to warm up, etc. The source was a Yamaha T-2 tuner and Shanling SACD player (using its solid state outputs), both have line level gain controls. Preamp/control center is an old Crown DL-2 circa 1978 flagship ($3000 msrp). Crown D-60 on each of the three speakers.

Try these various combinations of extreme gain levels. First turn the preamp all the way up. Set the gain level of your power amps to a low gain level and set the overall gain at the source component. Listen to how it sounds. After a while, try reversing the situation, turn the preamp down but run the power amps with the gain all the way up. I notice a BIG difference in what I would call the tonal quality of each of these extremes. If your setup (room included) is good enough you should also be able to hear differences in resolution, transparency and detail, whatever.

I tried this again the other night and was quite astounded. With the Crown DL-2 turned all the way up, and the Crown D-60s turned about a third of the way up, using the volume controls on the source components to control the overall gain, the sound really opened up. The bass became more full with more bloom, almost tube-like but not soft, wooly or mushy. More detail, openness in the midrange and treble. Now keep in mind that the Crown DL-2 has not a discrete component on its chassis. This thing is about as op-amp integrated circuit as anything Ive ever seen (it actually looks more like a PC inside). And yet, my guess is that most people, if not everyone, would not even know they were listening to a non-tube system if I had the tube gear lit but not connected.

So what Id like to know is why this happens? How does just changing the gain levels between the source, preamp and power amp alter the sound quality so much? IMHO, its conditions like this that are a determining a large part of what we hear. Why? Because not every group of components is going to let you do this. Some power amps dont have gain controls. Many line level components dont. Many phono preamp sections dont either. So I wonder how much of ones opinion about an amplifiers sound is formed by what one has to live with (or without) in regards to gain level control between various components.

And just in case anyone is wondering, NO, Im NOT canning my Luxman triodes, however I am considering getting some SET amps for comparison. These are simple amps and shouldnt cost an arm and a leg as most seem to. And since I dont like sending other people to the bank with a happy smile on their face with my money I think my best bet right now is a trio of Bottlehead 2A3 Paramount. Other suggestions welcome.

[Duke Spinner]

d-60/75

just like my Phase Linear 400...

have a sweet spot 'bout 1/3 power on the SOA curve

[NOSValves]

Artto,

The just of what your finding here is system matching. The implementation of many amplifier level controls is not the ideal place to be adjusting the signal. When I'm rebuilding many vintage amplifiers that have an input level control one of the first requests I'm ask for is to eliminate or upgrade the input level control. These amps had some of the worst implemantion of the controls and also tended to use the lowest quality control to boot.

When designing an amplifier without a input level control the designer has to walk a fine line also. Does he want his amp to be easily driven with say a passive preamp or directly from a CD player? Or does he want the amp to absolutely have to be used with an active preamp? Trade offs they can be frustrating.

Craig

[artto]

Thats an interesting point Craig. But lets say as Duke pointed out, that youre running Crown D-60/75 and youve found that the sweet spot is with the gain turned about 1/3 up (this is true BTW, thats where I usually keep mine set too). Or my Luxman MB3045, those of us whove had these for a while in original condition agree that they sound best with their input level controls set at 12 oclock, about half way up, not with the input controls removed. I think there may be something more to this than simply the insertion of gain controls into the circuit. Its almost as if you can use the gain controls between the components as a tone control. It certainly can affect noise throughput, and at the extremes, even the more subjective things like detail and transparency.

[NOSValves]

Artto,

Not surprising your not getting the just of what I said as intended since my writing skills have always sucked [] I'm not saying that "all" input levels controls on "all" amps are poorly implemented but sadly with many amps that is the case.

Of course there is going to be a sweet spot and if I was designing them into an amplifier I would try to make that spot at about 50%. With that said you can't possibly rule out that any given amp with this sweet spot could not improve with the control removed the the sweet spot made exact and not adjustable (although the end difference may be small).

The problem with the sweet spot is it doesn't always match the amplifiers source component sweet spot! (preamp, CDP or whatever it maybe)

Controls on any kind have losses by nature they are not perfect devices. The 2nd largest modification request I receive is to upgrade the main volume control on preamps and integrated amps. Sadly with most vintage amps this is not an easy task since these controls often serve multiple functions beyond just volume control.

[Duke Spinner]

, the series R will interact with the grid capacitance to form the Miller Effect and reduce HF response of the stage.

Aaaaahhhh ....

the Miller Effect ......only on ... Triodes ...right ..???

[oldtimer]

Hey Duke pass me one of those Miller's, I'm gettin' dry.

[artto]

Ah Hah! Thanks Mark. It's good to know I was actually hearing things! (as opposed to just thinking I was hearing things).

Subjective evaluation has it's place but it's always nice to be able to quantify things if possible.

I hadn't actually tried this for a long time until last week. For anything serious with CD or SACD I usually just bypass everything and run the Shanling right into the power amps. With LP you obviously can't do that so easily. But regardless, as I mentioned, the Luxman MB3045 seems to sound it's best with a middle setting of the input control so it "set" me a wondering again.

[artto]

Im not sure exactly when the MB3045 were designed. If my memory serves me right it was around 1975. From what I know, theyre supposedly based on older McIntosh designs, albeit with different, custom made output and driver tubes, and output transformer. The input is a 2 stage differential amplifier utilizing 12AX7 and 12AU7. Maybe this has something to with them sounding best with the input gain control turned half way up? Also different from early Mac designs (or any other designs Im aware of) is the quadrafilar wound output transformer which apparently is very expensive and difficult to make. I asked another MB3045 guru friend of mine about adapting some 2A3 or 300B to it but he said the Luxman has too much high voltage and the variety of tubes available would result in low power. But Im thinking if it results in something like a very robust Brooks 2A3 it just might be the ticket. On the other hand might this "too high voltage" also be responsible higher ME?