So here are a few arguably obsolete comments on DIY and chosen parts. I will finish with a commentary on tubes from a period point of view when most audio tubes were still in good supply. Writing this will be a diversion as I need a break from my Jekyll/Hyde show which just opened.
So preliminaries on DIY – I like Jason’s support of the spirit of DIY. I have thought of doing my own DIY project as well but two separate considerations disabuse me of the notion. The first is that modern digital products depend on surface mount. Not just small resistors and capacitors, but teeny tiny multilegged ICs with up to and more than 100 pins with pin spacing under a millimeter is some cases, with ground planes UNDER the IC. Ugh. This is not to mention the need to program firmware, which requires code portability and expensive proprietary hardware. The second is that I would have to support it. Those that have known me over the years are familiar with my meager customer service skills. (Oh yeah, well if you didn’t want to solder then why the phuc did you buy it?, etc) So it looks like I will not be doing this anytime soon.
So back to 1974, when I was just out of school and working at Douglas, a job I for the most part, hated. I worked for a genius who graduated from Cal Tech, wrote several graduate level textbooks on DC amps, grounding and shielding techniques, etc. He played in string quartets for relaxation and was quite the musical virtuoso. He was completely bald and his head had lumps. He looked kinda like a tall, lumpy, fireplug with warts.
He was the guy who taught me how to differentiate signal path resistors. Something called voltage coefficients (vocos for the really cool nerdspeak of the time). What is that? It is the amount that resistors change their ohmic value as the voltage across them changes. He was smart – that turned out to be the primary sonic determinant of resistors. Vocos kinda makes them spongy as the signal wiggles across them. Now back 50 or so years ago, there were carbon composition resistors, carbon film resistors, metal film resistors, and (really expensive) non-inductive wire-wound resistors, and finally really, really expensive) Vishay foil resistors. These are named in order of lesser voltage coefficients. This was before surface mount parts and are still available for amplifier DIYers. The carbon comp resistors sound like frozen twice and rewarmed ass. Foil resistors are stunning, simply a knockout. Lower vocos, better sound. Period. With today's surface mount we have thick film (ass) and thin film (much, much better). These are still not as good as non-inductive wirewound or foil, but certainly as good as legacy metal film. A hint is that tube designs swing many more volts than solid state, so resistor selection is very important for tubes.
Oh, and caps. Back in the day, film caps were the only choice for audio signals. The fire hydrant with leprosy and lumps told me the most important factor was DA. Huh? Dialectric absorbtion. That test goes like this: you charge the cap for 60 seconds, short it out for 60 seconds, wait 60 seconds and then measure the voltage across the cap compared to the original charging voltage. That percentage is the DA spec. Just like the voco spec for resistors, the DA spec for caps tells you how it’s gonna sound. Electrolytic caps suck balls; strictly for bad audio like radios. The film caps go something like this as they improve: mylar, polycarbonate, polyphenylene sulfide, polypropylene, polystyrene, and teflon. Teflon caps are bigger than a can of Australian beer and cost more than sending someone through college. Polystyrene are only available in smaller values and melt at higher (tube) temps. Polypropylene are kinda big, especially if film and foil and are the best practical bet. There are two primary ways to make tubular caps – metalized and film and foil. Metalized are wound together with two dialectric plastic rolls with a conductive spray bonded to the plastic on each. The wires are fastened to each foil layer. Film and foil use two pure foils and two dialectric plastic rolls all wound together. Leads go to each foil. Far more physically obese, but sound much better than metalized which have much higher levels of asstivity in each dialectric type. The best dialectric for surface mount caps is polyphenylene sulfide. Their construction is far more equivalent to film and foil. If you are building an insanely physically big tube amp, search out film and foil polypropylenes. If you are doing anything modern, polyphenylene sulfide surface mount caps are the way to go. Their much smaller physical size bring MUCH smaller values of distributed capacitance and inductance, which is a HUGE sonic advantage in itself. You can’t use these with tubes at all, because the voltages will launch them off of the board.
Tubes – someone just wrote a very interesting post over on Jason’s thread about sonic differences between tube and solid state gear, which, for the most part reflect my thoughts. I would add an extra viewpoint or two. One is that tubes are great at swinging lots of volts. They are, after all inherent voltage devices. (So are FETS, which sound quite different without even bringing up enhancement mode vs. depletion mode – tubes are more analogous to depletion mode – I will make this a topic for a theoretical vs practical post someday.) One challenge with tube circuits is they require generally more capacitors than solid state ones. Another is that they suck for low impedance outputs in general and high current ones in particular. Makes it harder to drive everything from poorly designed audiophile cables to headphones to speaker. The way around this is in a power amplifier is to use an output transformer which has low frequency limitations that solid state amps do not.
Tubes can vary in performance such as linearity, gain, and steady state levels which require matching. There are tubes designed for consumer audio such as 12AX7, 12AU7, and 12AT7s which are designed for garbage applications, such as portable radios, and portable record players. They feature awful performance, but are ubiquitous in audio design. I have harsher thoughts, but they can only be expressed elsewhere. Tubes that were designed for mission critical applications such as test equipment (6DJ8 families, ALL Western Electric audio tubes, etc.) are not only more reliable, but can far more linear as well. Mutual conductance tube checkers inform and test the parameters alluded to above to properly match tubes.
Emission tube checkers are generally looked down upon by current tube performance evaluators. A point is frequently missed. Tubes have a finite life. They have heater/cathodes or heaters flying solo which have a finite number of electrons to emit before the tube weakens and dies. As the tube ages, all of the measured parameters above change. The only way to conveniently albeit relatively gauge the remaining life of a tube is with an emission tester. They are cheap, usually less than a hundred bucks on ebay, and are the best predictor of the remaining tube life once worn and new tubes are compared. Also, as tubes age, their matching characteristics will change more quickly.
In product news, we have a working proto transport running. The remaining development work is an arm based processor to run the various traffic cop aspects of the transport, the operational buttons and display. Our current proto uses a Raspberry Pi zero which, of course is an arm based processor. It would be unfitting of us to include an overly complicated Raspberry Pi so we have begun our slimmed down Schiit Pie which Ivana will be baking. Speaking of Ivana, she has also prepared a second functional module to run alongside (as another choice) the gadget in our music processor. Wait until you hear it! We also have our own USB audio decoder done (for our inexpensive designs). Forty+ first article parts turntable parts have arrived for the Sol, and all but two were approved. We will see our foundry, and (given their) permission there may even be a Schiitferbrains video of the operators throwing a heavy duty pan full of molten aluminum into this smart car sized machine which then burps out a platter in a few seconds. Stay tuned.
Back to show news, it opened Saturday Night with an added matinee Sunday which resulted in two standing ovations both shows for a stunningly talented cast who really make me look good. It you are close to Newhall, come see it!