sound practices reading club /34

Today we read the final article for the power amps (part 2) topic. It is also the first one that is purely about a single-ended amplifier. I had not planned that amps, part 2 would be about push-pull amps, it just happened to turn out this way.

We are reading B-52: VV52 SE with 7788 Driver by Grego Sanguinetti, from sound practices issue 15. Apart from the general story of the development of this amplifier, Grego goes into parafeed output stages; triode-connected pentode driver stages; NiCad battery kathode bias; power supply tuning with a small first cap; VR tube regulation and alternative tube options.

teaser quote: ‘I proceeded with a few primary goals in mind: 1. Elegance by simplicity, i.e. low device count, especially in the signal path. 2. Adequate power for wildly varied tastes in music in my 900 square foot family room. 3. Good visual and aural appeal. 4. Oh yeah, the chassis(s) had to fit between the rails of a standard EIA relay rack.’

my take
I was going to write that this article introduces a number of firsts—for the reading club, that is. But checking shows that  triode-connected pentodes have already shown up a couple of times; parafeed made an appearance recently; and so did NiCad bias. But here they are, combined in a modern package. This really cannot be called a retro amp. And Grego writes up each of them, these not-quite-firsts.

A real fist for this article is the use of VR tubes. This is the article that convinced me to try them out. The write-up, although not without its ambiguities, helped to get me going. It was also in combination with a interstage-coupled stage and a revelation. Already battery biased (on the grid, in my case), the removal of the B+ cap left just the tube and the inductor (the interstage transformer). Finally I could hear them, and just them. And thus was born my personal rule of ‘no capacitors around a choke loaded stage,‘ Grego got there years before me with this design.

Yes, there is the 4µF parafeed cap in the output stage. You could argue that in any conventional single ended amp, there is also a cap. It is the final cap in the power supply that connects to the OPT. All output signal current runs through it. You are always listening to it. No, the problem is not the parafeed cap, it is the ‘nasty filter,’ formed by the that 4µF cap, the 50H plate choke, the OPT inductance and the 40µF cap in the power supply. That is a fourth order filter (vs. second order for conventional single ended: PS cap + OPT inductance). The cartoons Grego talks about is the erratic frequency response you get when you do not tame this filter. Grego did this deftly.

I ran a simulation (assuming 300H for the OPT). It showed the first section of the output stage (a choke loaded tube with a PS cap) to be peaking (~6dB at ~4Hz) mainly because of an oversized plate choke (19H is plenty for a 600 Ohm rp tube). But then the second section (parafeed cap and OPT) rolls off early (~10Hz), because of an undersized cap, and this compensates for the peaking. The combined response shelves a couple of dBs below 20Hz, and then a 4 Hz falls of a cliff. What is impressive is how little phase shift the output section has at 20Hz and 10Hz. And Grego got himself a relatively small parafeed cap size (don’t even think about increasing it in this setup), where more quality can be found.

pop quiz
How is the VV52B biased? Ah, self-bias with that 680 Ohm, 25W on the filament transformer. Should that resistor not be decoupled? Well, yes. Data for the VV52B is scarce, out there, but I found some numbers saying plate resistance is 600 Ohm and µ = 4. Assuming that, the unbypassed bias resistor adds (4 + 1) * 680 = 3k4 to the 600 Ohm. That means driving a 3k OPT with a 4k impedance. That is not going to give ‘authoritative and punchy’ bass, no damping. Grego promised to get back to me to confirm the real hookup in the amps, he can still do so in a comment to this post. Meanwhile, my best bet is that the 4µF parafeed cap is not connected to ground, but to the filament-end of the 680 Ohm bias resistor, so-called ultra-path connection.

practical note: there was postscript by Grego in SP16, which is not on the CD. Being too long to type up, I scanned it.

That is it for amps (part 2). Next, it is time to build something: konstruktion korner!

Now go and read the article, see you next week.