
Zajimavej to pocin http://www.groupdiy.com/index.php?topic=49282.0 .
Autor je znamej hiendovejma kompresorama a eq ( http://www.knifaudio.com/cgi-bin/view_eng.cgi?page=soma ). Obavam se, ze to nebude vubec levny jestli to nekdy bude prodavat.

Trochu cut'n'paste a drag'n'drop.
Hello folks,
I've spent several months with this project, and now I'm beginning to believe, that I will actually finish it. Missing from the photos are of course front panels, keys, power supply, all semiconductor stuff etc. Photos and early CAD drawings here:
https://picasaweb.google.com/jonte.knif ... 7611939218
https://picasaweb.google.com/jonte.knif ... 3748053346
Shortly: Oscillators, ring modulator, mixer, moog type filter, noise source, Vari Mu type VCA and output amp are tube units. Everything else is IC and some discrete. Don't ask for schematics, I spent full three months on the circuits alone.
Tuning systems can be changed with small plug in boards. Oscillators have saw, triangle, pulse and very crude sine. There is sync and FM modulation. Keyboard will have after touch but not dynamics. Three octaves of keys, in imitation ivory and imitation tortoise shell.
-Jonte
Ideas are mostly my own, nothing from Barbour, his designs are not refined enough to perform anything else than "hey, this makes cool noises", whereas I like to see this synth as a musical instrument in a very classical way. VCA for example is true symmetrical Vari Mu, with pentodes, and provides clean 40dB range. Output amp is in fact simplified version of my "therapy pre" which I published here a couple of years ago.
Noise source is pretty much based on this funny thing:
http://www.jptronics.org/radios/GR/GR1390.pdf
Filter is true Moog ladder filter, which is pretty straightforward to do with tubes. It requires variable mu - type tubes.
Mixer is a simple virtual earth unit with 12AX7 and ECC88.
Ring modulation is done with vacuum diodes and transformers, with SE-drivers. Behavior is far from ideal ring modulation, but provides a lot of different timbres depending on drive volume.
Power supply is external and includes small speakers and amp. There are +200V, +120V, -100V, +/- 18V, two 12V and two 6,3VAC supplies, one of which is floated to 100 volts.
I hope I have time and energy to put together a web page for this beast some day, complete with sound examples, a lot of pictures, text etc. But first I have to finish the synth.
That remains to be heard. I am not really a synth freak, (Well...perhaps I am, but I don't own other synths actually), so I will have to quote other peoples opinions then. Luckily my friend has a Mini Moog in a very good condition and it sounds better than the new ones in some way.
I have so far listened to the filter and ring modulator, and I do like what I hear. They both have very wide ranges of increasing nonlinearities, and so has the VCA, but that I only looked from the scope. It does *look* good though, and I think I know when I see a potent distortion waveform.
There is a lot of "iron" in the signal path, but I am not quite sure which of the transformers actually will have a lot of impact on the sound and in which ways. Three transformers are in very central roles in the oscillators and resulting waveforms are not "perfect", nor is the starting point, a sawtooth from thyratron oscillator quite perfect either. I personally believe that it usually requires quite complex waveforms and/or transient phenomena to make a transformer heard in all its glory, so probably the output transformer is quite central in the timbre, and as you can see from the photos, it is the only "high end" trannnie in the synth, a Lundahl 1671, which is not terribly colored. Other trannies are smaller, and will certainly create color in the bass at least, where they start to saturate. Overall the signal levels inside the synth are very high, starting with ca. 40V p-p waveforms.
So, I would say that at least this particular tube synth has all the possibilities to sound quite different from the discrete semiconductor classics and most certainly beats all the IC stuff, which is just too "perfect". A lot of lossy and saturating iron + tube nonlinearities should make a difference. I wouldn't be building this if I didn't expect something extraordinary. Too much work and expenses for "for the sake of it" -project.
-Jonte