Joy of birth - a new preamp brought to the world

[dirkwright]

Built and the circuitry is on PCB's, so little room for modifications. As I stated earlier the first one of these has already served for three years or so, but now that I built another one, I see room for improvement. Doesn't matter really, though, since this is what I've been asked to do. Maybe a retrofit MKII PCB kit in the future... This one is part of a set of a preamp and two KT88 monoblocks I've been building for a friend; copies of what I already had.

Well, the only thing I was thinking of was to put the volume control before the gain stage instead of after the follower stage. With class A triodes and no loop feedback, the distortion goes up with increasing signal levels, so as I understand your circuit, the input triode sees the full output from your source component. If the volume control was before the input triode, then I think the distortion would go way down. I mean, 24dB is a hell of a lot of gain. Many people only use a buffer or maybe 6dB of gain in a preamp. But, if this works for you then leave it.

 

[Adolf Corkscrew]

Well, the only thing I was thinking of was to put the volume control before the gain stage instead of after the follower stage. With class A triodes and no loop feedback, the distortion goes up with increasing signal levels, so as I understand your circuit, the input triode sees the full output from your source component. If the volume control was before the input triode, then I think the distortion would go way down. I mean, 24dB is a hell of a lot of gain. Many people only use a buffer or maybe 6dB of gain in a preamp. But, if this works for you then leave it.

That's the gain you get with a CCS loaded 12BH7 or ECC99... Gain goes up, distortion goes down. The point of having the volume control in the end is a compromise between distortion and signal to noise ratio. This way you get the best SNR at low volumes, and especially for the guy I built these for it is very important, since he's driving extremely sensitive Voxativ fullranges with this set, in a very small space - thus small listening distance. With a CD-player standard maximum output of 2V_RMS the distortion still stays below 0.2%.

The next preamp I was thinking of probably has lower gain and has the volume pot in the front. But that will be a totally different story.

 

[dirkwright]

That's the gain you get with a CCS loaded 12BH7 or ECC99... Gain goes up, distortion goes down. The point of having the volume control in the end is a compromise between distortion and signal to noise ratio. This way you get the best SNR at low volumes, and especially for the guy I built these for it is very important, since he's driving extremely sensitive Voxativ fullranges with this set, in a very small space - thus small listening distance. With a CD-player standard maximum output of 2V_RMS the distortion still stays below 0.2%.

The next preamp I was thinking of probably has lower gain and has the volume pot in the front. But that will be a totally different story.

If you are pleased with 0.2% distortion, then that's great.

Just so you know, with the simulation in spice I showed, distortion is only 0.0007% with 2Vp-p input. Maybe it clips at 2Vrms, I don't know, haven't tried that.

So, my guess is that your circuit is not optimum, though spice-land is a special place, sometimes full of fantasies about zero distortion...

 

[Adolf Corkscrew]

The distortion figures I stated are within the margin I could say to be near reliable with the measuring equipment I had at hand back then. It could still be lower.

Anyhow, if one wanted the lowest possible distortion figure, he wouldn't be screwing around with tubes, but opamps, wouldn't he? Instant low distortion with no effort... Still, the tube preamp with more distortion (have'nt actually compared these on FFT, thus this is based on published numbers) sounds clearer and cleaner than a SS preamp with an opamp.

I really think that when you get into the sub-single-digit figures in distortion it loses it's meaning. 0,2% THD should be inaudible if it's in the low orders. By 0,2% THD something else comes in to play...

Oh, the circuit I described clips around 3V_RMS. It can give a clean 56V_pp out with a bandwidth of 450kHz... Try that with a 5532..