GoodSoundClub - Romy the Cat's Audio Site

In the Forum: Melquiades Amplifier
In the Thread: The DSET perspective examines the Herb Reichert article.

Post Subject: Opinions are like.... well, you know... everyone has one.Posted by floobydust on: 7/2/2009

I looked at the schematic for Herb's 300B SET many years ago. It's been copied and slightly modified by many.... personally, I would not use it for a 300B amp... then again I probably wouldn't use it for a 2A3 or 45 either (I think my input/driver design for the 45/2A3 is much better). Despite the 6SN7 being an excellent twin triode, I've concluded that you can't drive the 300B with an RC-coupled design based on this tube, at least not very well. The 300B is a real bitch to drive due to the high input capacitance. The testing I've done shows far too much dynamic compression as you drive the grid closer to zero and a sharp increase in distortion. I've abandoned RC-coupled for the time being (for the 300B) and will work on a transformer coupled design when I get more free time. I prefer to keep a lower cathode current, step up the plate voltage and run a higher impedance load on the tube (same for the 45 and 2A3). I also prefer a 5K load on the 300B however, with higher voltages of course.

You can bias a DHT for lower distortion at higher power levels and promote longer tube life if you don't push the cathode so hard and risk overheating the tube. Overheat it once and it's likely to be damaged.... especially if you cherried the plate in the process. Excess heat is still an enemy of the vacuum tube, last I checked. I'll run the plate dissipation up but I prefer to keep the cathode current below the maximum rating.

His ranting about No, this, No that, etc. on parts selection is just personal opinion... as he clearly states his design satisfies his personal taste in sound. I think most people develop a predilection for certain parts... resistors, capacitors, connectors, wires, sockets, etc. Again, simply personal choice based on what you like and what you can afford. I'm the same way for my personal gear. I still hate pretty much all connectors however... even some of the most expensive are mediocre for getting a pure connection. I also prefer NOS tube sockets and have a nice stash of them... EBY, Cinch, Amphenol and some others. All are silver-plated pins but most are not ceramic. When you're dealing with frequencies below 50KHz I doubt you'll find any real sonic difference... especially with a rectifier tube or a 4-pin UX4 for the 45, 2A3 and 300B. The newer stock of Chinese-made stuff doesn't cut it in my book. You can only get so close using metric equipment to make SAE spec parts and quality control problems make it worse.

I think his comment about prefering 10Y in push-pull can be interpreted two ways... one, that he prefers push-pull operation over SE, or two, that he prefers the 10Y triode in push-pull over the 10Y in SE mode. I think if he really prefered P-P overy SE than he wouldn't have spent much time doing SE designs. Again, just my $0.02.

As for the Wizard's cap... I don't want one... I'm not making magic or doing anything paranormal. When I design a circuit, I void everything in my mind and start clean sheet... plate curve charts come first... selecting component parts comes last. I do believe that the physical layout and construction can make or break any design however... taking a schematic and engineering it into a product is in most cases more difficult than doing the design itself... the reality of it all settles in and can bite pretty hard. To this day I still breadboard everything first. The real goal is to make an amplifier that has no sonic signature of it's own.

Still, you made some very good (and valid) points. I've not ventured into DSET territory yet.... but I have managed some very well optimized SE designs engineered to final products that perform very well... both measured and audible. Common sense should be your guide.

Regards, KM

Rerurn to Romy the Cat's Site