Art of Electronics, 3rd edition, corrections

Yes. There's an interesting story about this. My experience goes back to the circuit suggested by Harris Semiconductor, for their HIP5600 high-voltage regulator. It uses the same scheme, an LM317-type voltage-setting resistor to ground. The pot would be an 2-watt Allen Bradley type J, or similar. When you turn down the voltage, the pot must discharge the output capacitor. However, it seems to be perfectly happy enough to doing this.

...except when trying to figure out what actually happens inside the channe l during pinchoff. That place is weird. Shouldn't it go unstable and sta rt spalling off moving blobs of depletion, even spewing out millimeter wave s like a tiny little TWT amp? And I recall an IR microscope photo of a hor izontal mosfet channel emitting light as part of the usual breakdown regime . "Usual?" The FET channel in pinchoff is constantly breaking down? Yup, apparently.

On understanding BJTs, I can't resist here pushing my beginners' non-math a rticle on Ebers-Moll which doesn't discuss Ebers-Moll:

Hey, someone from cambridge.org contacted me about spreading the word on AOE-3, but then they never sent me any ads to put on amasci.com. Were none made?

Found it, Feb 3 Fran Robinson, no email response.

(((((((((((((((( ( ( ( ( (O) ) ) ) ) )))))))))))))))) William J. Beaty

nel during pinchoff. That place is weird. Shouldn't it go unstable and s tart spalling off moving blobs of depletion, even spewing out millimeter wa ves like a tiny little TWT amp? And I recall an IR microscope photo of a h orizontal mosfet channel emitting light as part of the usual breakdown regi me. "Usual?" The FET channel in pinchoff is constantly breaking down? Yu p, apparently.

article on Ebers-Moll which doesn't discuss Ebers-Moll:

Hi Bill, I read a page or so... I didn't get to bjt's

You know I said this the other day at work, and a smart electronics guy responded, "One elctron/hole from E to B base drags more along with it that fall into the collector".

And I thought, "No, it's Vbe that matter's. I have to pull some charge carriers out of the base to get Vbe lower. And that reduces the drift current, and so the carriers diffuse further.

It's more diffusion that gives more current. They are not pulled, they diffuse.

At least that's my current model. (Does that work for you?)

George H.

I don't understand how diodes or BJTs or fets work. Doesn't seem to matter. I do understand pretty much how they behave.

Somebody in marketing, no doubt. Dunno, haven't seen any ads.

I'll try emailing again.

We should all plaster some AOE banner ads all over our blogs!

Yes, a square root in the feedback loop is called for if the integrated error signal controls both the raw supply voltage, and the PWM output. In such a case, a 50:1 output control range can easily be extended to a 2500:1 dynamic range.

Tell us about Beautiful Layout Hunchback.

Two of my children have been helping out with various things, and we (or at least I) have a running joke about it being hard to get good hunchbacks these days.

Dashing Firmware Hunchback is my son Simon, who just finished his third year in Honours Physics at UBC. He spent two summers working with me building gizmos, which will be a leg up when he goes off to grad school. This summer he's going to be working at the local mini particle accelerator/laboratory astrophysics lab (TRIUMF) on a dark-matter detection experiment, where it sounds like he gets to be the main gizmo guy. (He'll be doing a bunch of FPGAs and other circuitry to look for correlations from 256 photomultipliers. He's really jazzed about it.)

Beautiful Layout Hunchback is my younger daughter, Magdalen, who recently graduated with a liberal arts degree. To fill in till the job of her dreams comes along, she's been doing PC board layouts, board stuffing, inventory, all that hunchback-type stuff, and studying physics and math part time (which she discovers that she loves).

It's a wonderful thing to be able to admire one's own children.

Cheers

Phil Hobbs

I used to use Liz, aka The Brat, for PCB layout. But she went and hired a layout guy, and signed up for an MBA, so I guess she's a manager now. Looks like I'll be working for her soon too. Suits me fine.

The project on the cover of the January/February 2016 edition of _QEX_ uses digitally tunable capacitors. The good book, AoE, mentions digital capacitors on page 1082. Once again, the enigmatic Chapter 3x appears. ;)

"My world is coming nearer."

I've used the 3-step Maxim "Flecap" as the coarse tune of a VCO, with a varicap for the fine tune. Worked fine. Other people make apparently better digital caps now.

Maxim has EOL'd it of course. Never Buy Maxim.

The _QEX_ project uses a 32 step Peregrine PE64102 [1].

Note.

1. formatting link

That part has a specified operating frequency range of 100M to 3G Hz. I wonder why there is a minimum.

A couple of other peaple make digital caps, but they seem to use MEMS technology. That's usually bad news.

Get a lifetime supply, 6k at Mouser and 13k at Rochester, 75 cents each.

I'd better. I don't want to spin the board layout or, even worse, rewrite the code, which is 68K assembly.

I'm designing a new oscillator now, and prefer to use a small surface-mount trimmer cap to center the frequency. One could tweak that with a screwdriver in 30 seconds. The kids that I employ want to do everything digital and scorn any sort of trimpot-like thing. It's a good thing that I'm still the boss.

I always hate doing lifetime buys- it feels like betting against success on the upside and throwing money away on the downside.

--sp

I had that attitude until fairly recently. I used a comparator and a trimp ot in an application that could have been served by a SOT-23 microcontrolle r or something similar, on the grounds that it was silly to use a dedicated processor to drive a status indicator LED. Real Men still do analog, righ t?

I soon realized that since nobody buys trimpots anymore, nobody bothers mak ing them very well, at least not at a price I want to pay. The second lot of trimpots I bought (presumably legitimate Bourns 3223W-1-102E parts, purc hased from Mouser) exhibited a 25% failure rate as soon as they were solder ed to the board.

So your less-experienced engineers may be right on this one. It seems that trimpots and trimmer capacitors are now considered weird parts. I'm learn ing not to use weird parts when I don't absolutely have to, even if it make s me feel like less of a Real Man.

-- john, KE5FX