Classic Computer Magazine Archive ATARI CLASSICS Volume 2, Issue 5 / October 1993 / PAGE 10

the 8-Bit Alchemist

     AUdio BypASS
UpGRAdE fOR ThE XL/XE




Multiple Personality Syndrome
    Welcome to an occasional column by your Managing Editor, whose secret ambition since 1984 has been to have some kind of Atari 8-bit magazine dedicated to hardware hackers. I had hoped ComputerSmyth would be such a publication when I first began publishing there in 1986, but alas!, that magazine went bellyup in 1988. I now know there will never be a "hacker's journal" as such. But we have AC, so maybe there's still hope for those of us with a passion for tinkering. Or, in my case, for improving things.
    There may be some among you who think I get off on being "Mr. Managing Editor". The fact is, I hate it. My emotional constitution was never made for being the "Manager" of anything. There are now about 25 people involved with this publication in one way or another, and keeping tabs on everyone is extremely stressful for a fellow as ungregarious as me. I do the Editor Thing because I sort of fell into it. But give me something that captures my fancy and I'll drag it off to my own quiet corner, far from the madding crowd. And that "quiet corner" is packed with all kinds of scientific toys.
    In real life I'm a research chemist. I work for one of the world's largest pharmaceutical companies, and my specialty is natural products chemistry. If you saw the movie "Medicine Man", you have the Hollywood version of what I do for a living (Sean Connery makes everything look easy!). Basically, my job is to look for new medicines in sources from Nature: mostly exotic plants, microbial fermentations, or slimy dead things from the ocean. Although I get involved with hi-tech stuff like reversed-phase high pressure liquid chromatography and gated spin-echo nuclear magnetic resonance, it isn't as glamorous as Hollywood makes it out to be: some of the stuff I work with would make a billygoat puke. And the daily environment I inhabit in a corporate chemistry lab is about as far as you can get from the congested serenity of ACs Editorial Offices.
    My hobby is electronics. I was- and still am- an audio freak. The subject of psychoacoustics never fails to fascinate me, and I have an ongoing love affair with monster power amplifiers. During the 1980's I achieved the status of a minor guru by publishing a number of articles in audio electronics. These papers, along with a dozen or so contributions to professional scientific journals, constituted my rites of passage in the publishing world. Nearly all those articles were written on the same trusty 800XL I'm using to compose this article right now.

Birth of The Alchemist
    In 1983 I read a book called The New Alchemists (by Dirk Hanson; Little, Brown & Co., 1982, now out of print) in which I first glimpsed a philosophical perspective for my tekkie interests. In one of the most spellbinding narratives I've ever read, Hanson describes how the modern science of microelectronics (the "New Alchemy") was founded upon scientific discoveries in chemistry, physics, and mathematics in the late 19th century. From the strange adventures of Edison, Babbage, and Tesla, Hanson weaves a tale that works up to the glory days of Silicon Valley- including the founding of Atari by Nolan Bushnell and the startup of Apple Computer by an ex-Atari employee named Steve Jobs.
    Given that my work involves equal measures of voodoo and sophisticated science, I've always felt a certain kinship to the alchemists of old (one wall of ACs editorial offices is adorned with a quote from the great German alchemist Johann Joachim Becher who lived in the 1600's). Along comes this guy Hanson who proclaims my hobby to be a new form of alchemy. And of course, I'm a sworn 8-bitter. There you have it: The 8-Bit Alchemist. The logic is inescapable.
    Although I had submitted several articles to Current Notes magazine in early 1990, in September of that year I was officially recognized as that magazine's 8-Bit Editor with the publication of the first installment of my column titled "The 8-Bit Alchemist". Those of you who subscribed to CN back then will recognize my familiar logo. I'm very grateful to Joe Waters at Current Notes for allowing me to use my old logo in this magazine. I departed CN in January 1992 following a fire which severely damaged my home in late 1991. (I wish I could say the fire was caused by something honorable, like a 300-watt audio amp blowing up. Instead, it was something stupid: sloppy masonry work in my fireplace started a fire inside the wall.) During the hiatus imposed by repairs to my house, I got caught up in the Mail Campaign begun by Jeff McWilliams to start this magazine. The rest, as they say, is history.
    Ever since the founding of AC, the Alchemist in me has been at odds with my roles as Managing Editor and Circulation Editor. At long last the Editors in me succumbed to the Alchemist, casting aside the pen in favor of the soldering iron. Nearly two years of pentup frustration was suddenly cut loose, and The 8-Bit Alchemist went on a mouth-frothing binge of hardware hacking. These labors yielded a rich harvest of hardware info and improvements which I hope to present in future issues. For the present I'll content myself with showing you a neat little hack that improves audio in the XL/XE machines. It's altogether fitting that the first hack presented by The 8-Bit Alchemist in AC has to do with audio.

R2 location in the 800XL

Treble Serendipity
    Rarely do I get a chance to present a hack that does two good things at once and is almost absurdly simple to perform. Most hardware hacks are difficult to do and require lots of soldering skill. Not this one. The hardest part of this hack is just getting your computer open without wrecking its innards. I'll diverge for a moment to describe how I stumbled on this hack.
    Several of my 800XL machines, especially one that contains Chuck Steinman's GUMBY upgrade, seem to put out various levels of annoying background noise that varies considerably from machine to machine. The noise is on the SIO audio line and is easily audible on my monitor speaker. It's composed of a hash of lowlevel squarewaves inhabiting the lower 8 octaves of the audio spectrum. I'd characterize it as a raspy, buzzsaw kind of sound. After 15 minutes of this I suddenly feel a vague urge to kill something. Sound familiar?
    I pulled out my 800XL schematics, popped the hood on my 800XL, scooped up a handful of capacitors, and got down to it. Initially it was my intention to hang a filter cap across the offending audio line. This didn't work so well. I quickly found that capacitance values large enough to suppress the noise also suppressed most of the legitimate audio signal. Throwing out the baby with the bathwater isn't my style, so I kept looking.
    Audio in the 600/800XL (but not the 1200XL), and in the XE machines, is handled by a garden-variety LM358 dual op-amp chip. It seemed to me the noise was being generated by instability in the op-amp circuit. In the XL/XE machines, the layout of the audio circuit almost guarantees you'll have some noise, due to the messy array of inputs and outputs which contribute a lot of stray capacitance to the high impedance inputs of the op-amps. The usual way to tame op-amp instability is to bypass the feedback resistor with a small value capacitor to prevent RF oscillation. I tried this in both halves of the op-amp circuit, with disappointing results.
    OK, I knew I had to stick some capacitance in there somewhere. If feedback bypassing didn't work, how about a feedforward bypass? There are two 2000-ohm coupling resistors in the 800XL op-amp circuit: R2 and R7. I tried my capacitor bypass on R7 first. That was the wrong one, it trashed the sound. So I tried bypassing R2 with my capacitor. That worked. The annoying background buzzies died down considerably.
    I got a pleasant surprise when I played some music through the modified circuit: not only was the background noise level much lower, but the treble tones got a nice little boost! I ran the audio self-test routines while listening to the output through the speaker in a Commodore 1902 monitor as I alternately connected and disconnected my bypass capacitor. Yup, I wasn't hearing things. The extra sparkle in the treble was plainly audible, and most welcome. I suddenly realized how dull the audio on a stock 800XL really was.

R111 location in the 600XL

R2 location in the 130XE

Doing The Mod
    You'll need a 0.luF capacitor, preferably a mylar type but any non-polarized cap will do. Don't worry about the voltage rating on this capacitor, it doesn't matter. Even if you buy this cap at your local rip-off Radio Shack, you'll get by for well under a dollar. Trim the leads on the cap to about 1/4-inch, and solder it across the appropriate 2000-ohm (red-black-red-gold) resistor on the component side of the board. In the 800XL and 130XE, this resistor is designated R2. In the 600XL, it's designated R111. You folks with 65XE's will have to be on your own for this upgrade, as I don't have either a 65XE or diagrams for one. I'd be willing to bet the appropriate resistor in the 65XE is designated R2, if you can find it. (If anyone tries it on a 65XE, write in and let us know!) Use the diagrams to locate the proper resistor in your machine.
    I haven't done this upgrade in my 130XE, but I performed it on an 800XL and a 600XL with satisfactory results on both machines. Even if you don't have a noisy audio line, you might want to add the audio bypass capacitor anyway just to get the enhanced treble response. Try it, you'll like it!
    OK, so much for audio. What about video? Atari butchered the video in all the 8-bits, especially the XL machines. Can we fix it? Heh, heh. Stay tuned, folks, The Alchemist is cooking... .

Postscripts
    In the August AC I grumbled and snarled about being tied to Radio Shack for label stock for the venerable 1025 printer I use to crank out all those mailing labels for this magazine. While I appreciate the various exotic suggestions I received from several readers, I'm happy to say I've since found an alternate source for 1025 labels, considerably cheaper than the Rip-off Shack. They're available from MISCO (1 Misco Plaza, Holmdel, NJ 07733; phone orders 1-800876-4726) item #GD8695, $10.99 plus shpg. per box of 2000 in lots of 3 or more ($30 minimum order).
    Also in the August AC, I reported I had been negotiating with MicroMiser Software to obtain rights to TurboFile, with the intention of making it available to the market once again. Sadly, I must now report those negotiations were a failure, and it looks like the most powerful database ever written for our machines is on its way to the dustbin of oblivion. The Alchemist learned an unhappy lesson and will henceforth stick to the things he knows best.
    In an "Editor's Note" in the August AC I mentioned that my 1088K Newell RAM upgrade was "still flaky". Editors may not know much about malfunctioning RAM upgrades, but Alchemists do (oh gosh, keeping track of these multiple personalities is confusing!). Suffice it to say things quickly got put right when The Alchemist took over, and now my 1088K 800XL is performing smartly. Having 128K of XE memory plus a 960K RAMdisk is pure heaven for Editors who have to ARC up 200K of text files at a shot. So then, shall we see an article on the Alchemist's adventures-in Newell Mega-land? All in good time.