When AOE 3rd ED to be published?

[John Woodgate]

I read in sci.electronics.design that Tom Del Rosso

In John Woodgate typed:

"Hydrometers for the use of boys"?

Yes. A serious permuted entry uses a stock phrase:

Hydrometers, boys, for the use of.

However, this breaks down in the case of a stock solution used (in those
far-off days) in junior chemistry lab:

Soap solution, Boys, boys, not for the use of. (;-)

 

[Costas Vlachos]

I read in sci.electronics.design that Costas Vlachos <c-X-vlachos@hot-X-


... unless Fermat's marginal note was true!

Looks like he probably had a proof in his mind when he wrote the note, but
it must have been false as he later proved a special case of it.

But what if it was true... Makes you wonder... It's such an easy problem to
state, you can't help thinking that there must be a simple solution...

Costas

 

[jmc]

I think your website should have a hierachically organised suggestions page
to keep track of what has been suggested and what you are keeping and
rejecting. That way, yours and the posters' time should be saved in the long
run. In addition, a ready-made forum in the form of Yahoo Groups or
http://forum.snitz.com/. I can see your publishers not being too happy about
this for your current sales, perhaps;-0)

jmc

Forget the suggestion. Pointless. I keep forgetting that it's gonna be an
update, not a complete rewrite and most of it will be recognisable from the
second edition.
jmc

 

[Costas Vlachos]

I read in sci.electronics.design that Costas Vlachos <c-X-vlachos@hot-X-


I think that there are always infinitely many integer solutions to x^2 -
k*y^2 = 1, for k an integer > 0. So it would be much 'more wrong'
(because that's mathematics) to claim that there is no solution than to
fail to find it (because that's just arithmetic).

Close, but not quite true. It can be proven that there are infinitely many
positive integer solutions to x^2 - k*y^2 = 1 for k an integer > 0, if k is
not a square. If it is, then there are none. But yes, failing to find a
solution is not nearly as bad as claiming there's no solution.

Costas

 

[John Larkin]

So, from a practical engineering perspective, it is reasonable to
assume that there are no such solutions.

John

 

[Robert C Monsen]

The bad circuits are one of my favourite parts of the book! They're
expecially handy when doing candidate technical interviews -- I have a
few of the bad circuits pages tagged, and I'll open the book, and ask
the candidate to pick one out and tell me what's wrong.

I like the fact that both the good and bad circuits have no explicit
pointers. It makes them puzzles, and reinforces the content of the chapter.
I'll agree that the 'typo' puzzles are annoying, but there is some utility
to them, since real datasheets (and, sadly, many of my own circuits) have
errors like this; learning to spot the obvious ones is an art of sorts.

Regards,
Bob Monsen

 

[John Larkin]

Wouter van Ooijen (www.voti.nl wrote...

Dan Marshall just pointed out to me that NSC's LMC6462 datasheet,
bottom first page, has an innocent-looking page 258 fig F howler.

http://www.national.com/ds.cgi/LM/LMC6462.pdf

Thanks,
- Win

whill_at_picovolt-dot-com

This is great. It keeps the appnote-copier type of engineers from
taking our livelihoods away from us.

John

 

[Terry Pinnell]

I read in sci.electronics.design that Costas Vlachos <c-X-vlachos@hot-X-


... unless Fermat's marginal note was true!

I've stumbled across incontrovertible evidence that it was, but
unfortunately don't have the time here to describe it...

 

[John Woodgate]

I read in sci.electronics.design that John Larkin <jjlarkin@highSNIPland
THIStechPLEASEnology.com> wrote (in <livd10lpcl8s8u115rs5vne7aebl64m874@
4ax.com>) about 'Bad Circuits', on Tue, 27 Jan 2004:

So, from a practical engineering perspective, it is reasonable to
assume that there are no such solutions.

I think 'practical engineering' has about as much to do with this as
embroidery does.

 

[John Woodgate]

I read in sci.electronics.design that Terry Pinnell <terrypinDELETE@dial

I've stumbled across incontrovertible evidence that it was, but
unfortunately don't have the time here to describe it...

Put it on your web site.

 

[Max Hauser]

We have recently had a TV mini-series where a group of year 11 students
were sent to a re-created 1950-style boarding school. They didn't like
it a bit! An initial maths test was failed dismally by all but one
(maybe two), and this test was for year 6 students in 1950.

However, I suspect that a group of year 11 students in 1950 might well
have been similarly disenchanted with the accommodation, discipline and
academic standards of a school of 1900. Especially the viva voce
translation from Latin and Greek.

Just so. And in the US too, where a well-prepared 1900 graduate of public
instruction at age 17 or 18 would be expected to have, along with a solid
four years of mathematics, two modern and two ancient languages (usu. Latin
and Greek). In those days a good US High School education was something
substantial.

Don't just believe me -- as Victor Spinetti said in _Help,_ "It's from
Harvard!"

 

[Wouter van Ooijen]

Dan Marshall just pointed out to me that NSC's LMC6462 datasheet,

bottom first page, has an innocent-looking page 258 fig F howler.

Imagine a catalog of design mistakes, so when faced with a design
error I don't have to convince the author, I just say 'this is
definitely a H&H p600 fig Q error'


Wouter van Ooijen

-- ------------------------------------
http://www.voti.nl
PICmicro chips, programmers, consulting

 

[Costas Vlachos]

So, from a practical engineering perspective, it is reasonable to
assume that there are no such solutions.

John

Change the 991 to 960 and the smallest solution is

x = 31
y = 1

Lesson: Drawing general conclusions by simply examining an equation
numerically can be dangerous if you're not careful. If the 991 was the
nominal value of a noisy parameter that changed slightly, your assumption
about the non-existence of solutions would be wrong.

Number theory problems like this one rarely arise in practical engineering
problems though, so this is not a great example, but just illustrates the
importance of proof as opposed to numerical observations.

Costas

 

[John Woodgate]

I read in sci.electronics.design that Wouter van Ooijen (www.voti.nl)

Imagine a catalog of design mistakes, so when faced with a design error
I don't have to convince the author, I just say 'this is definitely a
H&H p600 fig Q error'

But there are so many design errors. Think 'page 60 950, fig ZZ'. (;-)

 

[Tom Del Rosso]

In Winfield Hill typed:

Thanks, Terry for that great list. While I suspect a few of the
things on your list aren't in the book (e.g. Coils and Coil
winding?), it's a very useful list. Gulp!

Oh yeah, coils and coil winding should also be in the book.

 

[Tim Shoppa]

I read in sci.electronics.design that Tim Shoppa <shoppa@trailing-


Yes, many of these arithmetical methods are very old. In fact, does
anybody know of a 'new' one? Say, no earlier than 20th century.

If anything, these methods have atrophied over the last century. I
was probably near the tail end of kids taught how to use log tables,
I'm guessing. That's a shame, because we now have a generation of
kids who expect every number expressed to 8 decimal digits and don't
have a clue about accuracy vs precision. I see this showing up in
s.e.d as folks assuming that there's nothing between a RC oscillator
and a TCXO (see the 16F628 oscillator thread) when of course there's
several economical choices in-between, or circuits that need 1%
parts to work when years ago we were happy if our 20% parts actually
stayed in that band .

Tim.

 

[Tom Del Rosso]

In Tim Shoppa typed:

One neat trick is 9's complement to do subtractions. Do a
google search on "nine's complement". It's related to "casting out
9's".

Don't you use 10's complement for subtraction? Or you might call it 9's
complement and start with a carry.