“The project we swore we'd never do... A Valve Preamp”
“A 12AX7 valve audio preamplifier” by Jim Rowe
Silicon Chip, Nov 2003
And by May 2005 after many letters and e-mails have flowed back and forth and the dust has finally settled a bit it is still no clearer why SC decided to do what they swore they would never do. And that's only the start of the puzzle.
I'll be right up front in my opinion that they should have left it alone instead of making a fish-nor-fowl hash of it. Some ideas are brilliantly different, and some are just different.
“That's a novel idea. You must have read it in a novel.”
In case you don't recognise the name, Jim Rowe has had a long career as a technical journalist and editor at the now-defunct Electronics Australia and has many significant projects to his credit. He is also a Ham and no stranger to the odd valve and his clear explanation in this article of how to do a loadline for a 12AX7 is well worth saving.
The quote in the heading above is taken from the banner that almost covered the front page, so right from the start there was something a bit odd about this project. Clearly there was pressure from somewhere (the publishers teenage son?) to do it and the editor Leo Simpson was strongly opposed. It should be noted that Leo is associated with some outstanding solid-state amps, both in power and low distortion.
In the second paragraph Jim Rowe says “Leo finally gave in and approved the development of a valve preamp for guitar and other instruments.”(my emph)
It's just a damn shame that they didn't bother to talk to any working guitarists, search the Web, nor it seems bother to properly inform themselves of the background and history of the area and what exactly guitarists expect from the valve preamp they were designing for them. How they might want to apply it on stage, in a rack mount, stomp, or as an amp build-in. This was a really fundamental project engineering oversight.
“That's a lovely bicycle guys, but the client actually wanted a boat.”
Thus Leo would later write me dismissive e-mails that would suggest he thought the whole grubby business of guitar amplification was unworthy of his attention as an engineer; he has his standards of reproduction “anything over 1% distortion is gross”, take it or leave it, and it's time guitarists got real and stopped fooling themselves about valves. He doesn't seem receptive to the idea that he might be missing something.
It's not so much that this project went off the rails somewhere, as it never had a track to begin with. I'd love transcripts of the editorial meetings leading up to this project; I'm sure they would be really illuminating. [and no sooner written, than Elektor March 2005 published an insight into their own editorial process in a leader and article on staff disagreements about valves and transistors for audio]
Right from the start, and despite the explicit mention of guitars, the implicit object of the project was the lowest possible distortion, and when the first attempt produced 0.9% THD Jim commented:
“At Silicon Chip we have always tried to produce the best available audio performance, so we decided to try a different approach...” In short applying negative feedback. (33k, 33k, 220nF)
By this point they had seriously lost the plot.
This invites the obvious question, “why then are you doing a valve preamp at all?” What's the point of using a valve for nominally valve sound, warmth, or characteristic curve distortion, call it what you will, then using NFB to iron it out flat? Pardon me if this seems not just contradictory, but perverse.
“To a man with a hammer every problem is a nail.”
The key word that tries to sneak past is “best”. What they forgot is that “best” depends entirely on what you are trying to do, and somewhere very early-on this project somehow got permuted from their un-baked perception of a preamp for guitar into some kind of equally ill-conceived Hi-Fi preamp and ended up falling into the bloody great gap between the two areas.
For some odd reason Leo's “1%” comment puts me in mind of thumping great motor-driving servo-amps I once worked on - brutal - and what THD they had, and did it matter? Did anyone dare even ask? In my mind a guitar amp is much closer kin to those servo-amps than any Hi-Fi. 'Ballroom Dancing' is down the hall, this is 'Weight Lifting'.
“Who is going to use this preamp?” I asked Leo. “It has far too much distortion for Hi-Fi purists, and not nearly enough to interest serious guitarists.”
Jim Rowe himself makes the point in the article (p29) that even though modern valves out-perform older ones, the noise and distortion performance of this preamp is well short of what can easily be achieved using solid-state, and I would be the first to agree with him. If our “best” is 0% THD then valves, old or modern, aren't even a starter. Even an LM833 will walk all over a 12AX7 any day. That's why I used them in my keyboard amp.
But what about for guitar?
At some point Leo told me that negative feedback was normal on preamps of the period. After considerable checking I will re-assert my reply; feedback was indeed almost universal on Hi-Fi preamps, also called control units, of the period.
But with the occasional exception of Baxandall tone controls or an un-bypassed cathode resistor, negative feedback in guitar preamps isn't just rare, it's unknown. If you know of even one exception, please e-mail me because I'm still looking and want to know.
[ Since writing the above I have been informed of one (obsure) brand of American guitar amp that used NFB around its preamp. There had to be at least one out there, and its very obscurity reinforces my point that guitarists don't want the “benefits” NFB in their preamps. ]
“Don't fix what 'ain't broke”
After my initial technicians interest at the front cover my guitarists heart sank when I saw the photo of the project at the head of the index page and took in the RCA connectors, the totally exposed “Hi-Fi” construction, then the inverter of which Jim remarks “Most of the noise is a low-level 'frazzle' from the 33kHz switching hash from the DC-DC inverter.” Gee, how unexpected! How welcome! Leo has his prejudice against distortion, and I have mine against inverter hash around audio.
It may surprise the guys at SC to learn that not all stage amps are bandwidth-challenged valves. Initially my new preamp for keyboard behaved more like an IF strip, flat to 3MHz (no, that's not a misprint), and prone to take off at 450kHz. Until I limited the bandpass to, coincidentally, 33kHz. So if any of Jims inverter 'frazzle' got into my amp I'd expect 'frazzled' tweeters.
As as Soundie I can't imagine anything less popular on stage than a “frazzling” inverter. Sorry Jim, bring it over here and I'll stand on it for you.
The RCA connectors alone totally give the game away. Standard in Hi-Fi practice they are almost unknown in stage work, the equally problematic 6.5mm jacks being the standard, as every guitarist knows. And we all know how long that input RCA is going to last with an RCA-to-Jack adaptor and guitar lead hanging off it, don't we?
The 'proud valve' format is also very Hi-Fi and wouldn't last one gig in the real world where even well-protected valves get broken. Physically, compared to stage gear that works, this is just a fragile toy.
Far from going down as a “must have” classic, it's an opportunity squandered. Direct from the design bench into the dustbin of guitar amp history.
http://www.ozvalveamps.org/scpreamp.htm | Last update: 19:35 02/03/06