Horowitz-Hill: Serious scholarly query

Here is a serious constructive inquiry or feedback about the Horowitz-Hill _Art of Electronics_ book, for W. Hill or P. Horowitz. I first raised this matter by email but saw no response to date. The issue remains, so I'll try posting it here.

Both the first (1980) and second (1989) edition of the H-H book contain a particular remarkable circuit design, for example Figure 9.73 of the first-edition copy that I've seen (9.90 in the second). This design is a pseudorandom noise generator with programmable spectrum, demonstrating several important electronics principles at once. What I found most striking about it was how closely it resembles the circuit published, if I recall, the year before the first edition, in _Electronic Design_ magazine by Michael Workman of IBM. The latter received an award by that magazine for best design idea of the year (with attendant publicity). This unusual design appears unattributed in both editions of the Horowitz-Hill book that I have seen. I _don't_ know or presume the exact history of the circuit. But whether Workman's appeared first (making it prior art), or the circuit in the book appeared elsewhere before Workman (originating with you, or with some other author), attribution in your book would be widely considered good form, eliminating any conceivable misunderstanding. (Workman, by the way, at my last word, was a vice president at IBM.)

(Again my original effort with this query was private, in a January 2004 email from my IEEE address to W. Hill's email address of record. I later posted a reminder of the email, spelling out the address I'd used, here on SED on February 1 as part of a technical thread in which W. Hill participated. I've seen no response to either.)

Max Hauser

Reply to
Max Hauser
Loading thread data ...

H&H need not attribute the idea to Workman because it predates him by a decade or two.

John

Reply to
John Larkin

Of course there's nothing didactical about AoE; electronic design is an art.

John

Reply to
John Larkin

Horowitz-Hill

raised this

I'll try

contain a

is a

demonstrating

if I

magazine

magazine

unusual

book that

circuit.

circuit

or with

considered good

way,

2004

later

here on

Hewlett-Packard published a version of the circuit in the Hewlett-Packard Journal in the early 1970s. I used it to as the basis for a one-off for a friend of mine in the late 1970's. AoE mentions the National Semiconductor IC - the MM5437 - which was a single chip implementation of the same idea, so it doesn't sound as if HP had a patent, which suggests they didn't invent it.

Workman's implementation is obviously not "prior art" and EDN is not in any sense a referreed journal - I myself have had the experience of seeing an equation published in EDN Design Ideas that I'd published in Measurement Science and Technology a few years earlier (November 1996)

- admittedly Measurement Science and Technology is a British jurnal (published by the British Institute of Physics

formatting link
) and thus not read by Americans.

If you want to raise issues of priority, you really should quote publication dates ....

-------- Bill Sloman, Nijmegen

Reply to
bill.sloman

I'm sorry Max, I received neither of those emails you mention, nor did I see the post you refer to. My long-standing home email address has not been functional for about a year now, and my work email should be used instead, edit this: hill_a@t_rowland-dotties-harvard-dot.s-edu

I do try to answer all my mail. However, at work I receive about 10x more spams than real mail, more than about 300 messages/day (at home this number exceeded 2000/day and actually reached 10000 one weekend). Now my email is automatically screened with a Bayesian filter before I see it. This makes living with spam more bearable, only a 2x ratio to go through, but it also means some legitimate mail may be rejected.

Hmm, another email I don't recall receiving. I'd be interested in seeing a copy, to see how you phrased your complaint to me. Perhaps it was more along the line of suggestions? :>)

All I can say for our book is we worked hard on it, and it has to stand on its own. We're making major modifications for the third edition, and are happy to hear suggestions, however it's not likely we will become a Tietze-Schenk look-alike. That's because our attitude and approach is completely different - for example, we prefer to talk about parts and part numbers along with theory, more along the lines of what your completed drawings will look like when ready for your technician to make. We're trying to teach you how to think about your circuit, live inside it, the art of the design, so to speak, rather than simply the dry theory.

One complaint that's grown in my mind is that we've gone too far in this direction, often spelling out a design thought-by-thought, without simply writing down the equations. So we're adding those into the third edition, and tightening the text in places. We're also adding a set of concisely- written dense advanced "X" chapters, where we can pour more good stuff. It'll continue to be presenting electronic circuit design as we see it.

We think if you could only have one electronics book, ours wouldn't be a bad choice, but we also think you should absolutely not own just one book. Along these lines, I have one suggestion for Ban, get hold of a copy of Peter Dunn's book, "Gateways into Electronics," ISBN 0-471-25448-7.

--
 Thanks,
    - Win
Reply to
Winfield Hill

As I mentioned before, I didn't receive your email. However, we can address your concern here if you like.

I remember working with Paul on that drawing, which would have to have been sometime in 1977 or 1978. The inspiration for our design was an HP instrument which came out probably in the early 70s. This instrument fully laid out the technique, although it was more complex than our design, and I thought we could devise a more simple version. I don't recall the Workman article in ED, although I may have seen it - could you send me a copy? It would be interesting to compare and consider the differences.

If you examine our book, you'll see that we have very few references. We did not write, nor did we claim to write a scholarly book. Early in the writing process we were overwhelmed by the mass and scope of what we hoped to accomplish, and it was clear we could not take the time to find good or accurate references for everything we were presenting, and therefore we didn't even try. Here and there an idea struck us as being so inspired or clever that we mentioned its source, such as the "atta-boy" we bestowed on HP for an RF detector circuit, but such mentions are few and far between.

While our lack of references does make it hard for interested readers to explore more (and has sometimes even made it hard for us to tell someone now where to research an issue, for example we haven't been able to find the issue of the HP Journal in which the RF detector appeared), it was a necessary part of our effort, allowing us to finish without burning out.

We didn't often copy material directly in the book, but if we did, such as drawings or photographs, we were careful to get and print the permission.

--
 Thanks,
    - Win
Reply to
Winfield Hill

Max, I made the same experience. After buying AoE I was disappointed with it and wrote an Email to Win about his claim to teach electronic design with this book. In contrary to Tietze-Schenk AoE represents an unconcise collection of circuits from different sources, but missing is the red thread going through. It might be of value for the average hobbyist, but there is no didactical value IMHO.

There was no reaction to this mail as well.

--
ciao Ban
Bordighera, Italy
Reply to
Ban

Not at all. It is indeed a "science". The issue here is that many haven't really given much thought to just how they design. They usually waffle on about how nebulous it was to come up with an actual design, but the bottom line is that design are always just a matter of the Darwinian axioms, i.e. replication, random variation and selection.

Kevin Aylward snipped-for-privacy@anasoft.co.uk

formatting link
SuperSpice, a very affordable Mixed-Mode Windows Simulator with Schematic Capture, Waveform Display, FFT's and Filter Design.

Reply to
Kevin Aylward

Analysis is science; invention is art.

That's not the way I do it; that would take *way* too long, and miss the really revolutionary ideas.

John

Reply to
John Larkin

No, but it improves with practice, and I can teach it. Just like, say, tennis (which I can't teach.)

Design is holographic, massively parallel. I think that's the way a brain has to work, given an unlimited supply of slow gates.

Schitzophrenia is correlated with creativity. I think of both as being a breakdown in boundaries between things usually kept apart.

In my experience, there's only a modest correlation between serious analytical skills and the ability to invent new ideas and architectures. That's why AoE is a good book... it breaks away from the rigidity of theory (which one certainly should have as a baseline) and plays with ideas.

If electronic design were all equations, you might as well be an accountant.

John

Reply to
John Larkin

Where do you get the student manual?

...Jim Thompson

--
|  James E.Thompson, P.E.                           |    mens     |
|  Analog Innovations, Inc.                         |     et      |
|  Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
|  Phoenix, Arizona            Voice:(480)460-2350  |             |
|  E-mail Address at Website     Fax:(480)460-2142  |  Brass Rat  |
|       http://www.analog-innovations.com           |    1962     |
             
I love to cook with wine.      Sometimes I even put it in the food.
Reply to
Jim Thompson

The HP instrument I referred to was the 3722A. It's not in my 1967 HP catalog, but it appears in 1969 with a broad 4-page entry (they also had a 8057A digital pink-noise generator in that catalog).

I scanned and posted the four 3722A pages to a.b.s.e. You'll see the cost of the 3722A was $2650, which I circled back then, while drooling. The last appearance I see was in my 1980 HP catalog, when it cost $4185.

--
 Thanks,
    - Win
Reply to
Winfield Hill

But you've said you don't know how you do it. ;-) However, even if one assumes that random variations and that most of the selection is done subconciously, I suspect that truly random variations would lead to FAR too many dead ends, and it has more to do with pattern fitting, visual and structural metaphors if you will, perhaps from unrelated previously experienced patterns. That would also go some way to explaining the apparent connection between genius and insanity- more like the parameters being somewhat out of whack than a defect in the mechanism.

Best regards, Spehro Pefhany

--
"it's the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com
Reply to
Spehro Pefhany

I understand about the references, and I concur with the policy. AoE is a lore book rather than an analytical textbook--it's like having RoboWin at your elbow to answer questions. Lore books are rare (and precious) in most fields, but electronics is unusually rich in them, a style probably rooted in the hobbyist tradition of early radio. I consciously adopted this approach (including the attaboy-style references) in writing about electro-optics. (Optics isn't the very last bastion of Edwardian stuffiness, but it's close.) A bit more mathematical rigour and especially physical insight in spots (e.g. putting the Transistor Man in concrete galoshes, and explaining why I_s has a huge positive tempco) would be useful. It's great that you seem to be heading that way.

I'm looking forward to AoE 3rd very much. (Don't take up skydiving until after you send back the galleys.)

Cheers,

Phil Hobbs

Reply to
Phil Hobbs

The underlying theory is rather old, almost 175 years, see "Evariste Galois".

Best regards, Spehro Pefhany

--
"it's the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com
Reply to
Spehro Pefhany
[...]

[...]

Win,

You can kill all spam very easily. Get a free account at

formatting link

and generate a suitable number of email addresses. Use them for temporary site registration, posting to newsgroups, and give them to your friends. Keep track of where they went so you can retire one when a friend gets infected with a mail virus.

Select one and enter it with the "mailto:" prefix in

formatting link

Put the encoded address on a web page somewhere. Add the url to your sig.

People who want to contact you can click on the url, which will take them to the encoded email address on your web page. They click on that and can send you an email.

You can add the address to your filter and know it is from someone you may wish to answer. This speeds filtering. My program filters between 100 and

600 messages/sec using this technique.

Spambots ignore the web page since there are so many different ways to encode data. After a while you can pretty much assume any mail to your previously posted address is spam and can be ignored. Here is an example:

formatting link

I have used this method for over a year and have never got a single spam to the encoded address.

However, I have received unwanted commercial messages from vendors. After warning them to stop, which had no effect, I simply went back to spammotel and cancelled that email address, then entered a new one on the web page.

Problem solved.

This is guaranteed to stop your spam problem. Try it!

Best,

Mike Monett

P.S. The reason for not using

formatting link
is they have a quota and simply ignore any email messages above the limit. They do not tell you they are discarding your emails.

Reply to
Mike Monett

I read in sci.electronics.design that Winfield Hill wrote (in ) about 'Horowitz-Hill: Serious scholarly query', on Thu, 23 Dec 2004:

I agree, if you buy the student manual as well. Without it, the largely non-mathematical approach can be frustrating for those who can cope with maths.

--
Regards, John Woodgate, OOO - Own Opinions Only. 
The good news is that nothing is compulsory.
The bad news is that everything is prohibited.
http://www.jmwa.demon.co.uk Also see http://www.isce.org.uk
Reply to
John Woodgate

I read in sci.electronics.design that Winfield Hill wrote (in ) about 'Horowitz-Hill: Serious scholarly query', on Thu, 23 Dec 2004:

I understand your problem. I suggest you include that (suitably edited) statement in the next edition.

--
Regards, John Woodgate, OOO - Own Opinions Only. 
The good news is that nothing is compulsory.
The bad news is that everything is prohibited.
http://www.jmwa.demon.co.uk Also see http://www.isce.org.uk
Reply to
John Woodgate

Last time I checked there were no actual axioms in Darwinism as it's still just a theory. What are you saying here Kevin - that design is random or accidental?

Regards, Mike.

Reply to
Mike Page

If they can cope with the math then they won't be frustrated by the lack thereof but relish the challenge to produce their own calculation. Who the hell wants the insult of having everything spelled out for them? The whole idea of that book is to develop insight into circuit operation without getting bogged down in the distraction of minute mathematical detail of unnecessary precision.

Reply to
Fred Bloggs

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.