ee's without math

On Sat, 9 Dec 2023 12:11:18 +0000, Martin Brown
<'''newspam'''@nonad.co.uk> wrote:

>On 08/12/2023 23:04, john larkin wrote:
>> On Fri, 08 Dec 2023 22:01:11 +0000, Cursitor Doom <cd@notformail.com>
>> wrote:
>> 
>>> On Fri, 08 Dec 2023 08:05:09 -0800, John Larkin <jl@997PotHill.com>
>>> wrote:
>>>
>>>>
>>>> https://cosmosmagazine.com/science/mathematics/ripple-effects-of-maths-crisis-spread-to-engineering/
>>>>
>>>> (The aussies call it 'maths')
>>>
>>> Maxwell's equations. That's where a lot of people decide it's not the
>>> life for them!
>> 
>> Of course, very few ee's ever use Maxwell's equations. I never have.
>> My fields instructor was a brilliant Japanese guy and we couldn't
>> understand anything that he said. He graded on the curve.
>
>We had the same problem with our High Energy Physics lecturer. He was 
>Italian and a bit vivacious - his English wasn't great to begin with and 
>when excited he slipped into fast Italian. His lectures were almost 
>incomprehensible to us. HEP always looked like stamp collecting to me 
>and it still does. Or as another unsympathetic to HEP physicist put it 
>studying horology by smashing clocks together at ever greater speeds.
>
>> I haven't actually used calculus in about 20 years. I have to have a
>> feel for differential equations and initial conditions and such, but I
>> don't actually have to do it. I use Spice. Anything interesting is
>> nonlinear anyhow.
>
>That doesn't mean that you can't model it mathematically and have a cute 
>cubic or gulp quartic equation to solve analytically and give you a feel 
>for what is actually going on (or a good starting guess to refine).


We do polynomials all the time, in code or an FPGA, to linearize
things, like digitizing a resistor-thrmistor voltage divider into
temperature. Or yesterday, padding a digital capacitor to tweak its
effect on an LC oscillator. But that doesn't involve solving
equations; we run simulations, apply instincts, and tune.

I used to apply fancy current sources to capacitors to make linear
timing ramps. Or bootstrap. Now we use an RC and math out the roughly
exponential curve as part of instrument calibration. 4th order poly is
pretty good.

Once in a while I delegate actual symbolic equation solving to a
bright kid, but there is computer software to do that too.


>
>> Being able to do higher math is a kind of mechanical skill. It doesn't
>> necessarily create instincts for circuits or system dynamics.
>
>Only at the enough to pass exams stage. Higher maths is all about 
>intuiting an answer and then doing the formal algebra to prove that your 
>initial guess was right and communicate it to others unambiguously.
>
>> Some people, like Phil H, can see through the math to the reality, but
>> I think most EE students can't.
>
>That statement I agree with. I've often wondered why so many EEs find 
>Einstein's special relativity so completely impossible to understand.
>
>> Nowadays, computer programs can even do symbolic math and solve
>> equations.
>
>Nowhere near as well as a human can yet. But they can do brute force 
>algebra manipulations that would take humans forever and then be full of 
>errors (and have been doing so in some specialities since the 1980's).
>
>Human intuition and computer algebra (or other computer implementation) 
>to avoid silly mistakes is still the optimum for now. I'm not sure that 
>will hold for very much longer as general AI is getting frighteningly 
>good at more and more abstract and thought to be impossible problems.

The electronic design job is still inventing architectures and
circuits to solve problems. One needs ideas and then coarse instincts
for whether an idea could actually ever work, and then analysis tools
to see if it's practical. That can be solving differential equations,
or simulating, or breadboarding. Whatever works.

Having an EE degree, having all that theory pounded in to become
instinct, is almost mandatory, even if one uses the academic
techniques seldom or never.

The valuable part is design, having the ideas for circuits and
products. There are lots of ways to delegate the grunt work.

On Sat, 09 Dec 2023 06:54:17 GMT, Jan Panteltje <alien@comet.invalid>
wrote:

>On a sunny day (Fri, 08 Dec 2023 08:05:09 -0800) it happened John Larkin
><jl@997PotHill.com> wrote in <5jf6nidoovd0ki276ngnfdato3qsutmns4@4ax.com>:
>
>>
>>https://cosmosmagazine.com/science/mathematics/ripple-effects-of-maths-crisis-spread-to-engineering/
>>
>>(The aussies call it 'maths')
>
>I almost never use maaz
>Even counting change no longer is necessary with all those money cards.

I don't think I can subtract any more, and certainly can't do long
division. But I'm very good at guessing results to 5 or 10% accuracy,
standing up a whiteboard, sort of a human analog computer. That's good
enough to decide, say, if a pcb trace capacitance is worth worrying
about.



>But really, I almost never use maaz and all stuff works.
>Filters and stuff - plenty programs or online calculators where you can just enter data.
>But I am just a neural net.
>Lost of experience designing and building thing and seeing and repairing designs from others.
>Net is trained very well!
>Last time I uses S parameters was in my school days.
>Much much much more (enter more much-es, not math-es) is UNDERSTANDING what them electrons are doing.
>Same for other stuff in fishsicks.
>
>Same for programming.
>Big problem is units perhaps, US wants to be different, so one of their Mars orbiters crashed because they used the wrong units...
> https://science.nasa.gov/mission/mars-climate-orbiter/
>But then again if the tinkerers had a clue they would notice the different between pounds and Newtons.
>
>Bit of algebra is usually all you need...

Right. A*B = C*D needs to be untangled now and then.



>And logic reasoning, this is for the mamaticians here:
> https://arstechnica.com/ai/2023/12/the-real-research-behind-the-wild-rumors-about-openais-q-project/

On Sat, 9 Dec 2023 07:13:48 -0800 (PST), Fred Bloggs
<bloggs.fredbloggs.fred@gmail.com> wrote:

>On Friday, December 8, 2023 at 11:05:59?AM UTC-5, John Larkin wrote:
>> https://cosmosmagazine.com/science/mathematics/ripple-effects-of-maths-crisis-spread-to-engineering/ 
>> 
>> (The aussies call it 'maths')
>
>This may come as a surprise to you, but engineers were NEVER good at math. 

I have a couple that are. But I have more crazy ideas - that become
products - than they do. Is that a correlation somehow?

>Hence all these charts, graphs, nomo's, table lookups, handbooks, standards, arithmeticization of transcendental math ( transforms) and whatever else it took to get them *numbers* in least time, if at all. Engineers used to be exceptionally good at arithmetic. They used to be ridiculously pathetic programmers totally lacking in analytical and organizational skills, and probably still are AFAIK. Then don't even let them near singularities. Heaviside's so-called analysis is mere symbolic arithmetic. Author of article is a case point, a complete idiot.

Einstein almost invented a few things, like the laser, but didn't.
That's curious.

Calculators erased the need to be good at arithmetic. Slide rules
didn't add or subtract or work to 9 places.

On Saturday, December 9, 2023 at 11:11:06 AM UTC-5, John Larkin wrote:
> On Sat, 9 Dec 2023 12:11:18 +0000, Martin Brown 
> <'''newspam'''@nonad.co.uk> wrote: 
> 
> >On 08/12/2023 23:04, john larkin wrote: 
> >> On Fri, 08 Dec 2023 22:01:11 +0000, Cursitor Doom <c...@notformail.com> 
> >> wrote: 
> >> 
> >>> On Fri, 08 Dec 2023 08:05:09 -0800, John Larkin <j...@997PotHill.com> 
> >>> wrote: 
> >>> 
> >>>> 
> >>>> https://cosmosmagazine.com/science/mathematics/ripple-effects-of-maths-crisis-spread-to-engineering/ 
> >>>> 
> >>>> (The aussies call it 'maths') 
> >>> 
> >>> Maxwell's equations. That's where a lot of people decide it's not the 
> >>> life for them! 
> >> 
> >> Of course, very few ee's ever use Maxwell's equations. I never have. 
> >> My fields instructor was a brilliant Japanese guy and we couldn't 
> >> understand anything that he said. He graded on the curve. 
> > 
> >We had the same problem with our High Energy Physics lecturer. He was 
> >Italian and a bit vivacious - his English wasn't great to begin with and 
> >when excited he slipped into fast Italian. His lectures were almost 
> >incomprehensible to us. HEP always looked like stamp collecting to me 
> >and it still does. Or as another unsympathetic to HEP physicist put it 
> >studying horology by smashing clocks together at ever greater speeds. 
> > 
> >> I haven't actually used calculus in about 20 years. I have to have a 
> >> feel for differential equations and initial conditions and such, but I 
> >> don't actually have to do it. I use Spice. Anything interesting is 
> >> nonlinear anyhow. 
> > 
> >That doesn't mean that you can't model it mathematically and have a cute 
> >cubic or gulp quartic equation to solve analytically and give you a feel 
> >for what is actually going on (or a good starting guess to refine).
> We do polynomials all the time, in code or an FPGA, to linearize 
> things, like digitizing a resistor-thrmistor voltage divider into 
> temperature. Or yesterday, padding a digital capacitor to tweak its 
> effect on an LC oscillator. But that doesn't involve solving 
> equations; we run simulations, apply instincts, and tune. 
> 
> I used to apply fancy current sources to capacitors to make linear 
> timing ramps. Or bootstrap. Now we use an RC and math out the roughly 
> exponential curve as part of instrument calibration. 4th order poly is 
> pretty good. 
> 
> Once in a while I delegate actual symbolic equation solving to a 
> bright kid, but there is computer software to do that too.
> > 
> >> Being able to do higher math is a kind of mechanical skill. It doesn't 
> >> necessarily create instincts for circuits or system dynamics. 
> > 
> >Only at the enough to pass exams stage. Higher maths is all about 
> >intuiting an answer and then doing the formal algebra to prove that your 
> >initial guess was right and communicate it to others unambiguously. 
> > 
> >> Some people, like Phil H, can see through the math to the reality, but 
> >> I think most EE students can't. 
> > 
> >That statement I agree with. I've often wondered why so many EEs find 
> >Einstein's special relativity so completely impossible to understand. 
> > 
> >> Nowadays, computer programs can even do symbolic math and solve 
> >> equations. 
> > 
> >Nowhere near as well as a human can yet. But they can do brute force 
> >algebra manipulations that would take humans forever and then be full of 
> >errors (and have been doing so in some specialities since the 1980's). 
> > 
> >Human intuition and computer algebra (or other computer implementation) 
> >to avoid silly mistakes is still the optimum for now. I'm not sure that 
> >will hold for very much longer as general AI is getting frighteningly 
> >good at more and more abstract and thought to be impossible problems.
> The electronic design job is still inventing architectures and 
> circuits to solve problems. One needs ideas and then coarse instincts 
> for whether an idea could actually ever work, and then analysis tools 
> to see if it's practical. That can be solving differential equations, 
> or simulating, or breadboarding. Whatever works. 
> 
> Having an EE degree, having all that theory pounded in to become 
> instinct, is almost mandatory, even if one uses the academic 
> techniques seldom or never. 
> 
> The valuable part is design, having the ideas for circuits and 
> products. There are lots of ways to delegate the grunt work.


Pulling up an 'applet' to do a calculation is not math. Knowing there's an applet available for the job is engineering.

On Saturday, December 9, 2023 at 11:25:51 AM UTC-5, John Larkin wrote:
> On Sat, 9 Dec 2023 07:13:48 -0800 (PST), Fred Bloggs 
> <bloggs.fred...@gmail.com> wrote:
> >On Friday, December 8, 2023 at 11:05:59?AM UTC-5, John Larkin wrote: 
> >> https://cosmosmagazine.com/science/mathematics/ripple-effects-of-maths-crisis-spread-to-engineering/ 
> >> 
> >> (The aussies call it 'maths') 
> > 
> >This may come as a surprise to you, but engineers were NEVER good at math.
> I have a couple that are. But I have more crazy ideas - that become 
> products - than they do. Is that a correlation somehow?
> >Hence all these charts, graphs, nomo's, table lookups, handbooks, standards, arithmeticization of transcendental math ( transforms) and whatever else it took to get them *numbers* in least time, if at all. Engineers used to be exceptionally good at arithmetic. They used to be ridiculously pathetic programmers totally lacking in analytical and organizational skills, and probably still are AFAIK. Then don't even let them near singularities. Heaviside's so-called analysis is mere symbolic arithmetic. Author of article is a case point, a complete idiot.
> Einstein almost invented a few things, like the laser, but didn't. 
> That's curious. 

Actually he did, he collaborated on the design of a refrigerator, and it was patented.

https://en.wikipedia.org/wiki/Einstein_refrigerator

> 
> Calculators erased the need to be good at arithmetic. Slide rules 
> didn't add or subtract or work to 9 places.

On Saturday, December 9, 2023 at 11:25:51 AM UTC-5, John Larkin wrote:
> On Sat, 9 Dec 2023 07:13:48 -0800 (PST), Fred Bloggs 
> <bloggs.fred...@gmail.com> wrote:
> >On Friday, December 8, 2023 at 11:05:59?AM UTC-5, John Larkin wrote: 
> >> https://cosmosmagazine.com/science/mathematics/ripple-effects-of-maths-crisis-spread-to-engineering/ 
> >> 
> >> (The aussies call it 'maths') 
> > 
> >This may come as a surprise to you, but engineers were NEVER good at math.
> I have a couple that are. But I have more crazy ideas - that become 
> products - than they do. Is that a correlation somehow?
> >Hence all these charts, graphs, nomo's, table lookups, handbooks, standards, arithmeticization of transcendental math ( transforms) and whatever else it took to get them *numbers* in least time, if at all. Engineers used to be exceptionally good at arithmetic. They used to be ridiculously pathetic programmers totally lacking in analytical and organizational skills, and probably still are AFAIK. Then don't even let them near singularities. Heaviside's so-called analysis is mere symbolic arithmetic. Author of article is a case point, a complete idiot.
> Einstein almost invented a few things, like the laser, but didn't. 
> That's curious. 
> 
> Calculators erased the need to be good at arithmetic. Slide rules 
> didn't add or subtract or work to 9 places.

Slide rule is a form of nomogram:

https://en.wikipedia.org/wiki/Slide_rule

Now how does an applications engineer need to know anything to 9-places? Digital doesn't count.

On Sat, 9 Dec 2023 09:07:05 -0800 (PST), Fred Bloggs
<bloggs.fredbloggs.fred@gmail.com> wrote:

>On Saturday, December 9, 2023 at 11:25:51?AM UTC-5, John Larkin wrote:
>> On Sat, 9 Dec 2023 07:13:48 -0800 (PST), Fred Bloggs 
>> <bloggs.fred...@gmail.com> wrote:
>> >On Friday, December 8, 2023 at 11:05:59?AM UTC-5, John Larkin wrote: 
>> >> https://cosmosmagazine.com/science/mathematics/ripple-effects-of-maths-crisis-spread-to-engineering/ 
>> >> 
>> >> (The aussies call it 'maths') 
>> > 
>> >This may come as a surprise to you, but engineers were NEVER good at math.
>> I have a couple that are. But I have more crazy ideas - that become 
>> products - than they do. Is that a correlation somehow?
>> >Hence all these charts, graphs, nomo's, table lookups, handbooks, standards, arithmeticization of transcendental math ( transforms) and whatever else it took to get them *numbers* in least time, if at all. Engineers used to be exceptionally good at arithmetic. They used to be ridiculously pathetic programmers totally lacking in analytical and organizational skills, and probably still are AFAIK. Then don't even let them near singularities. Heaviside's so-called analysis is mere symbolic arithmetic. Author of article is a case point, a complete idiot.
>> Einstein almost invented a few things, like the laser, but didn't. 
>> That's curious. 
>
>Actually he did, he collaborated on the design of a refrigerator, and it was patented.
>
>https://en.wikipedia.org/wiki/Einstein_refrigerator
>

"It has been suggested that most of the actual inventing was done by
Szil&#4294967295;rd". That aligns with my observation that professor types seldom
have original ideas.

I'd think that staring at equations would suggest possibilities, but
it rarely does. The positron is an interesting case.

On Sat, 09 Dec 2023 08:25:01 -0800, John Larkin <jl@997PotHill.com>
wrote:

>On Sat, 9 Dec 2023 07:13:48 -0800 (PST), Fred Bloggs
><bloggs.fredbloggs.fred@gmail.com> wrote:
>
>>On Friday, December 8, 2023 at 11:05:59?AM UTC-5, John Larkin wrote:
>>> https://cosmosmagazine.com/science/mathematics/ripple-effects-of-maths-crisis-spread-to-engineering/ 
>>> 
>>> (The aussies call it 'maths')
>>
>>This may come as a surprise to you, but engineers were NEVER good at math. 
>
>I have a couple that are. But I have more crazy ideas - that become
>products - than they do. Is that a correlation somehow?
>
>>Hence all these charts, graphs, nomo's, table lookups, handbooks, standards, arithmeticization of transcendental math ( transforms) and whatever else it took to get them *numbers* in least time, if at all. Engineers used to be exceptionally good at arithmetic. They used to be ridiculously pathetic programmers totally lacking in analytical and organizational skills, and probably still are AFAIK. Then don't even let them near singularities. Heaviside's so-called analysis is mere symbolic arithmetic. Author of article is a case point, a complete idiot.
>
>Einstein almost invented a few things, like the laser, but didn't.
>That's curious.
>
>Calculators erased the need to be good at arithmetic. Slide rules
>didn't add or subtract or work to 9 places.

Maybe not. But they *were* a coveted status symbol. If someone strode
into the office with a slide rule hooked to their belt, like a big,
swinging dick, you *knew* immediately he was an engineer. Lesser
minions were simply in awe. If you want to make an entrance - I mean a
*real* entrance - clutching a calculator simply won't cut it.

John Larkin <jl@997PotHill.com> wrote:
> On Sat, 9 Dec 2023 07:13:48 -0800 (PST), Fred Bloggs
> <bloggs.fredbloggs.fred@gmail.com> wrote:
> 
>> On Friday, December 8, 2023 at 11:05:59?AM UTC-5, John Larkin wrote:
>>> https://cosmosmagazine.com/science/mathematics/ripple-effects-of-maths-crisis-spread-to-engineering/
>>> 
>>> 
>>> (The aussies call it 'maths')
>> 
>> This may come as a surprise to you, but engineers were NEVER good at math. 
> 
> I have a couple that are. But I have more crazy ideas - that become
> products - than they do. Is that a correlation somehow?
> 
>> Hence all these charts, graphs, nomo's, table lookups, handbooks,
>> standards, arithmeticization of transcendental math ( transforms) and
>> whatever else it took to get them *numbers* in least time, if at all.
>> Engineers used to be exceptionally good at arithmetic. They used to be
>> ridiculously pathetic programmers totally lacking in analytical and
>> organizational skills, and probably still are AFAIK. Then don't even let
>> them near singularities. Heaviside's so-called analysis is mere symbolic
>> arithmetic. Author of article is a case point, a complete idiot.
> 
> Einstein almost invented a few things, like the laser, but didn't.
> That's curious.
> 
> Calculators erased the need to be good at arithmetic. Slide rules
> didn't add or subtract or work to 9 places.
> 
> 

You guys are just calling sour grapes. ;)

Mathematical notation is a technology of thought. It enables even mildly
skilled users to make correct inferences of a complexity far beyond the
reach of ordinary rhetorical thinking. 

The problem is mostly how it&rsquo;s taught at the lower levels.  Specifically,
the junior-high  notion of &ldquo;simplification&rdquo; taught us to collect all terms
with the same x dependence, with no notion of their size, origin, or
significance. 

Applied to design problems, that leads to monolithic formulas with big
complicated expressions for the polynomial coefficients, which naturally
give zero insight. 

Keeping the various contributions separate makes it much easier to see
which ones are important and how to improve things. 

Cheers 

Phil Hobbs 
(Who has no wish to go back to engineering, pyramid style.)


-- 
Dr Philip C D Hobbs  Principal Consultant  ElectroOptical Innovations LLC /
Hobbs ElectroOptics  Optics, Electro-optics, Photonics, Analog Electronics 

On Saturday, December 9, 2023 at 7:13:54 AM UTC-8, Fred Bloggs wrote:
> On Friday, December 8, 2023 at 11:05:59 AM UTC-5, John Larkin wrote: 
> > https://cosmosmagazine.com/science/mathematics/ripple-effects-of-maths-crisis-spread-to-engineering/ 
> > 
> > (The aussies call it 'maths')
> This may come as a surprise to you, but engineers were NEVER good at math. Hence all these charts, graphs, nomo's, table lookups, handbooks, standards, arithmeticization of transcendental math ( transforms) and whatever else it took to get them *numbers* in least time, if at all.

Oh, the 'engineers' of medieval times were war machine builders; those of two centuries ago were
steam engine tenders.   A modern engineer could be... almost any kind of worker.   

The math of Maxwell's equations opens up insights not otherwise available: you can't have a
current-sense resistor without associated series inductance, because (insert knowledge of a field
construct called Poynting's vector), and you can sense magnets with a Hall device because (insert
knowledge of a field effect and its interaction with P-type and N-type doped materials).

To be an inventor, one needs a good imagination and a lot of junk - Thomas Edison

Lots of mathematics is part of the 'good imagination', because it guides around blockages.   It's part of
the 'lot of junk' because... well, really, I can go years without thinking about  Poynting's vector; that's
just a dusty reference amid the others on my mental bookshelf.   Like other unused materials, it
is waiting for use, though: don't we all like to admire the goods in a junkyard?

As for charts, nomographs, and tables "to get them numbers", that's also to encourage visualization,
and pure mathematics benefits greatly from visualization; one wouldn't study Bessel functions without
a copy of Jahnke and Emde for the pictures...

<https://archive.org/details/in.ernet.dli.2015.212842>