Thoughts on tube "magic".........

Don't intend to spread any voodoo or nothing but I was wondering if anyone has experienced with certain tubes that seem to portrait sound "better" than others?

From my own limited experience, I have found that certain tube seems to alter or in my case "enhance" the stereo image(eg. make the sound bigger and wider) and I guess it would have something to do with the frequency response of the tube. However, I have sweep the tube on the CRO and the signal indicated a fairly flat freq response with hi freq rolloff around 14khz.

Which lead me to believe that it might be the non linear charracteristic of the gm curve, however I tried to bias the tube at different point to see if that'll change the audible perception and got the same result from the same tube. So far I have yet to experience the same "enhancement" from other tubes, can anyone suggest any other testing method to test the tube??

First of all, what are you doing? Is this a stereo line amp, mic preamp, what? If so, what's the source whose sound is being enhanced?

And why is there a 14kHz rolloff? That's a bit on the low side.

Peace,
Paul

[quote author="pstamler"]First of all, what are you doing? Is this a stereo line amp, mic preamp, what?[/quote]

Does it make any difference if its a line amp or mike pre?

If so, what's the source whose sound is being enhanced?

And why is there a 14kHz rolloff? That's a bit on the low side.

Peace,
Paul

Click to expand...

Just music being played through it, 14Khz is the freq response the tube I'll try to change the bias to see if I can get it to roll off higher.

Go to any tube supplier website and you will notice that a Telefunken, Mullard, Amperex or equivalent brand 12AX7 cost 20 times more than a modern Chinese or Russian made 12AX7. Why? The sound....no amount of scope jockeying can explain this. Craftsmanship, tube materials, quality control and so forth. That's why I like to use JAN (joint army navy) spec tubes, and the ones developed for computers back in the day. When you venture into the world of tubes you will find a little bit of peculiarity in the way they behave. Goodluck.

Analag

[quote author="analag"]Go to any tube supplier website and you will notice that a Telefunken, Mullard, Amperex or equivalent brand 12AX7 cost 20 times more than a modern Chinese or Russian made 12AX7. Why? The sound....no amount of scope jockeying can explain this. Craftsmanship, tube materials, quality control and so forth. That's why I like to use JAN (joint army navy) spec tubes, and the ones developed for computers back in the day. When you venture into the world of tubes you will find a little bit of peculiarity in the way they behave. Goodluck.

Analag[/quote]

Hi Analag,
Yes, I am quite aware of the price difference and your explaination makes perfect sense. However, I would probably think that because these tubes are no longer in enough demand for production that'll probably also inflates its value.

OTOH, I am neither in the market for expensive audiophile tubes nor interested in rare audiphile tubes but more of trying to figure out and in search of exactly what is the reason that makes tubes sounds better. Of course, the definition of "better" is a matter of my liking.

I am looking for tubes that has an audible characteristic, it might be some unknown chinese tubes, russian tubes or whatever......... May be some dodgy made unbranded tube has some bizzare bulge or dips in freq response and that might create wonders in tracking some instruments. :shock:

I don't look for linearity and flat freq response building amp with tubes because I think that's trying alter the audible characteristic of the tube and defeats the purpose of using tubes in audio, since ICs are much linear, cheaper, simpler and performs much better in freq response why bother doing it with tubes?

To discover the variety in audible characteristics of tubes and its application is my primary objective, then trying to recreate that characteristic using IC circuits is what I am really aiming to do.

I am looking for tubes with ultra unlinearity and and with gm curves as f*cked up as bizzare as possible, looking for that individual characteristic of personality.

The tubes I use are cheap, it's the popular tubes that are expensive. Notice how tube designs are centered around the 12Axx family it's almost as if these are the only tubes they know. Here is an example...the 5687WA is a tube that kicks ass as a buffer it sells for $3.00 and is plentiful, yet you would rather use something from the 12Axx family that costs a lot more.

I also understand your interest is as scientific as it is audio related hopefully you can scope jockey some answers for the rest of us.
I make tube gear that is as if not more linear than the chips and dips you aspire to....and I'm happy to say nothing can do it quite like a tube can.
Analag

[quote author="analag"]
I also understand your interest is as scientific as it is audio related hopefully you can scope jockey some answers for the rest of us.[/quote]

Well, math used as reference in scientific principle is the most commonly established "language" that everyone can agree on and communicate with. I wish and hope to find some meaningful answers but at this particluar stage I can only wish and hope...... :?

[quote author="Learner"]However, I have sweep the tube on the CRO and the signal indicated a fairly flat freq response with hi freq rolloff around 14khz.[/quote]
A flat frequency response won't tell you much - but if the circuit rolls off at 14kHz, the circuit sounds "faulty".

To find out what's going on, try looking at the harmonic/distortion spectrum. The distribution of the different harmonics etc.

Best regards,

Mikkel C. Simonsen

Just want to add the signal starts to attenuate at around 14Khz, its not the 3db roll off point.

[quote author="Learner"]Just want to add the signal starts to attenuate at around 14Khz, its not the 3db roll off point.[/quote]

That is a significant information, because it can be only a small attenuation at 20K. You should always give the cutoff freq correctly (-3dB) or people will misunderstand your info.

chrissugar

[quote author="chrissugar"][quote author="Learner"]Just want to add the signal starts to attenuate at around 14Khz, its not the 3db roll off point.[/quote]

That is a significant information, because it can be only a small attenuation at 20K. You should always give the cutoff freq correctly (-3dB) or people will misunderstand your info.

chrissugar[/quote]

Sorry about that guys....

I guess its a scope jockey syndrome?? :shock: Since its easier to see the starting point of attenuation on the CRO than measuring the 3db roll off point..

[quote author="Learner"][quote author="chrissugar"][quote author="Learner"]Just want to add the signal starts to attenuate at around 14Khz, its not the 3db roll off point.[/quote]

That is a significant information, because it can be only a small attenuation at 20K. You should always give the cutoff freq correctly (-3dB) or people will misunderstand your info.[/quote]

I guess its a scope jockey syndrome?? :shock: Since its easier to see the starting point of attenuation on the CRO than measuring the 3db roll off point..[/quote]

It's actually easy to find the 3dB rolloff point on a scope. Use your variable gain knob to set the signal at an un-rolled-off frequency to 10 divisions, then increase the frequency until the signal occupies 7 divisions. (Well, 7.07 if you want to be more precise, but 7 gets you close enough.)
That's the 3dB rolloff point.

So you're amplifying two channels of signal through something that rolls off beginning at 14kHz? And it's adding width to the image? Questions:

What tubes are you using, with what plate & cathode resistors?

What is the load, including length of cable connecting to it?

And, most critical, do the two channels roll off at the same frequency?

Peace,
Paul

...

The circuit is mono, so I guess it was a bad choice of word to use stereo image I guess it should have been mono image???(Don't think I have heard of that one before) or the phrase Sound picture. Sorry about that once again...

Before I carry on with some more of this tube "magic" non sense and make a fool out of myself which I have done many times over, I think I'll need to build a few more circuits based on the same tube biased at couple of more points for listening comparison.

Winston's quote from 'MUSIC, SOUND and SENSATION' by Fritz Winckel, 1967: "With high non-linear distortion, the sound picture appears widened; this is of particular interest in single-channel recordings". has redirected me back to my initial suspicion, so I think I should definitely do a lot more thorough experiment in terms of comparing the audible differences between tubes biasing at linear vs non linear region.

Paul,
If you too are curious about the myth of this "magical" effect, I guess you can find out by doing the same with any tube that displays a linear and non linear region on its gm curve.

...

Cathode.

[quote author="CJ"]Cathode.[/quote]

Gesundheit.

Peace,
Paul

Yes, healthy cathode.
:grin: