When Andy Caspar, the good guy in Po Bronson's second novel, has maxed out his credit cards but still needs cash to bring his revolutionary computer ideas to market, he finds it in his closet.

For years he's been buying from the L.L. Bean catalog, and he's just remembered that Bean wear comes with a lifetime money-back guarantee. Sweaters, shirts, every pair of pants he owns and, bingo, he's got three months' rent.

It's just the sort of work-around you come to expect from Andy, one of the smartest, sweetest, most determined guys you'll ever want to meet, except when these characteristics don't aid the plot in "The First $20 Million Is Always the Hardest." He works at La Honda Institute, a Silicon Valley think tank where "ironmen," the best and brashest engineers, devise the next generation of computerphernalia, relatively free of corporate suffocation.

They may not have to wear ties at La Honda, but that's not to say there isn't corporate back-stabbing and intrigue. The institute's funding comes almost entirely from computer makers, and its main project is the 686 chip, the next bigger, faster chip that is going to allow its single sponsor, Omega Logic, to keep pace with Intel.

Andy, like any other self-respecting ironman, wants in on the project; it's why he renounced his Stanford scholarship and dropped out midterm when La Honda beckoned. But Francis Benoit, La Honda's brightest light and as expert at scheming as Andy is at schematics, has a private agenda. He not only dupes Andy into declining the 686 slot he's been offered but he also steers him toward a project no one wants: to design a $300 computer.

Smaller and simpler is the antithesis of what ironmen are about, but Andy decides that he's going to build the best $300 computer possible. First, he must recruit a team from among the handful of La Honda engineers wounded by not making the 686 cut. They're skeptical, not only of Andy, whom they know nothing about, but of the project. By coming to La Honda, they've shown they don't care about money; what they want is the jolt of nodding modestly when another engineer says, "Wow, you were in on that?"

Of course, when it comes down to it in a novel about the first $20 million, money does matter. These characters want the credit, but also the cash. Benoit, for example, is driven by resentment born during his days at Omega, when his brilliant work was undermined by its president, Lloyd Acheson, and his craven marketers. But when Benoit steals Andy's design, he turns to Acheson because he's the sort of guy who makes things happen.

Even Andy the choirboy, when he and his team separate from La Honda, conspires to withhold some revolutionary software they've developed while on the institute's payroll, within hours of pledging not to.

Of course, a computer that sells for $300 exists only in fiction, and trying to slip it in under the radar of suspended disbelief is one place Bronson stumbles. Andy declares triumph when he gets the list of parts down to $300, but what about marketing? What about profit?

Not only that, but Andy really gets the list down to only $400, shaving off the last hundred by applying "some general rules about the economies of scale in manufacturing, which suggested that for every doubling in unit volume of hardware built, your costs went down 25 percent." Now, that may well be true; I don't know. But my question is, how does Andy know?

What we know about him is that he's 29, he worked in Omega's marketing department before his disgust for corporate culture overwhelmed him, and he was going to Stanford before he moved to La Honda. How does he come to blithely apply economic theorems?

It's a problem that pervades Bronson's Andy, whose sterling attributes appear to be endless: He's a brilliant manipulator of men who somehow retains his decency, is a top-flight engineer even though he didn't study it in college, and can weld a pile of pipes into office furniture. And, of course, he's "tall with thick hair and good skin."

Just when we're willing to accept that Andy is in all things a savant, he shows himself to be a bonehead about venture capital: "He figured, hey, venture capital? and got out the yellow pages to look under the Vs." How has he lived in Silicon Valley for at least a half-dozen years and avoided even basic knowledge of start-ups?

Despite these bugs, it's not hard to get with Bronson's program. "The First $20 Million" is real and enjoyable enough to warrant spending a weekend with Andy and the gang. Who would have thought that a bunch of pocket-protected engineers could be this much fun?