a dangerous positive-feedback loop

The innovation is currently down to standards, such as USB, which are protected by patent.

The examples you give are over a century old?

Negative feedback predates electronics, or even electrification.

I'd expect that people were making water flow feedback loops thousands of years ago.

USB is awful, and hardly universal. A typical Intel hairball.

We recently invented a really radical time-locked-loop. Wish I could reveal it here.

mandag den 25. juni 2018 kl. 17.36.25 UTC+2 skrev John Larkin:

ow

h?

to

se's

y?

EE

ke

what is so awful about USB?

The connectors. Only the big B is reliable and obviously polarized. Everything else takes three tries to mate, and a few hundred tries to break.

Some driver implementations are terrible, like ones where a device only works on the connector that was used at install time. You've got to remember.

Port assignments are awful, at least in Windows.

And the "universal" bus should allow multiple masters and communication between any nodes. It's Intel-centric.

Several decennia must pass before an invention can be recognized as revolutionary. You want more recent ones? Computers small enough to carry in a pocket, the web, precision navigation for everyone, high-resolution satellite pictures of any place on earth, near ubiquitous connectivity, vast stores of near limitless information, virtual money, virtual reality, I could go on. This era is full of miracles we so easily take for granted, but which are becoming really life-changing and there will be more.

Sometimes a minor technological advance can turn into an important enabler. Electric cars and flying gadgets were not usefully possible before the advent of super magnets. LED lighting was of limited use until really efficient blue LEDs were invented. DNA editing was a haphazard affair before the invention of CRISPR. We haven't seen anything yet on that last one. Prepare for some major upheavals.

Jeroen Belleman

Yes, it was the abstruse (to me) higher mathematics demanded by the EE syllabus that resulted in me keeping it strictly hobby-only.

Have you ever used Type-C? It's pretty robust and there are no polarity issues. 'A' is pretty robust but not obviously polarized. yeah, Mini-B and Micro-B are pretty horrible.

I don't know how you can say that USB is Intel-centric. It is, well, universal.

an

There probably aren't fewer "real electronics" guys around. John Larkin pro bably means "real electronics guys like John Larkin" which is less of a pro blem.

If yoy can do it in an FPGA, it isn't an analog algorithm. Certain transfor ms that were formerly done with analog circuits to get them to run fast eno ugh to be useful can now be run fast enough in the digital domain in a prop erly configured FPGA (often with quite a lot of parallel execution).

Electronics with real devices in real laboratories is a lot more expensive to teach than Python programming and some forms of sociology - serious soci ology involves expensive data collection and at least one fraud

involved faking the data - which was a lot cheaper than collecting it. The guy's students eventually woke up to what he was doing and blew the whistle on him.

I suspect the limited numbers doing proper electronic engineering has more to do with the cost of running the courses than it does with difficulty lev el, which the universities may be pushing up to limit the numbers of studen ts they have to bother to instruct.

can

's

heir

n in

ns'

g has always been re-using known circuits & methods.

John Larkin seems to have an exaggerated idea of the level of innovation in volved in his products. He has claimed that it takes on average some two we eks of design time (not necessarily his) for him to have a new product to s ell, which does suggests that his new products aren't all that different f rom their predecessors.

rote:

angerous precivilised world, human minds ever on the lookout for a possible threat. It's well known that the fear system is very overactive.

ing

age of people that largely sit outside main social groups, ie dropouts. Soc ial fears are all ancient survival strategies.

tv, mostly social programs. And they socialise more. To get good at electro nics you have to choose to spend a good bit of time away from people develo ping the skills. Electronic engineering is hence primarily an introvert's g ame, and often a dropout's game. Same goes for invention, to get there take s a lot of time & skill development, and that runs counter to the average p erson that thinks only of socialising in their free time.

Fewer people acquire a complete EE education than get exposed to it. I didn 't get mine from formal instruction, but rapidly found out that trying to g et junior engineers to do stuff like Laplace transforms for me was a waste of time. They would lend me their lecture notes, which were usually compreh ensive enough to let me do what needed to be done.

As far as I can see, academic training is useful in letting you know that p articular techniques exist, and that particular textbooks exist when you ca n look up what you need, but what gets taught is the bare miniumum, suffici ent to solve toy problems. Whenever I've wanted to do something practical I 've had to go to the library to find at least one decent textbook, and to t he literature to find some kind of practical example of something that work ed - the literature tends to be more informative about the things that can go wrong that you have to pay particular attention to.

Even a hobbyist ought to understand what a Fourier transform is (and does).

Mind you, I did two years of undergraduate pure math without running into the concept, but Theory of Computation Part 1 (which was a first year subject, though I took it as a graduate student) did deal with it.

USB should not exist. Everything should be Ethernet+WiFi+5G, one gigantic worldwide network for everything, including voice.

My front-door camera, my printer in the basement, all my phones, all my hard drives, a weather station in Alaska, everything should be seamlessly connected. But don't let Intel or Microsoft have anything to do with security.

The problem there is that academics want to behave academically, namely fill pages with provable equations but avoid intuition and practical applications. I was already a serious hobbyist, and had done some professional design, before I went to college. Profs would do their academic stuff and it would be revelations to me, because I already had the examples. The rest of the guys just kept their heads down and copied equations.

Signals and Systems and control theory are incredibly valuable to designing electronics, but as much for the visual, intuitive appreciation of the concepts as for pages of equations.

The control theory class I had was incredibly abstract. Absolutely no concrete example given at all. I don't know how much of that is down to the prof (I don't think he was very good though). I had already taken it upon myself to solve AC circuits, and recognized the equations, what poles and zeroes are doing. I don't think anyone else in the class had a clue what was going on.

Tim

What was funny was when they got close to talking about nonlinear systems in their closed-form world. Wrist injuries from hand waving.

It's sad that those other guys missed, usually forgot, the stuff that we learned.

Engineering education was very practical and hands-on prior to WWII. The MIT RadLab and similar projects, and then Lawrence's invention of Big Physics, incurred a lot of physics envy amongst engineering educators, so they amped up the abstraction. It became a point of pride to not be practical.

I sometimes see the opposite effect in EE students: fiddle with Spice but not understand what's going on.

There is also the issue of culture. British text books on technical subjects are typically very cold, academic works consisting of mostly impenetrable slabs of equations page after page after page after page. American text books, by contrast make the subject much more accessible and interesting. Equations in EE are unavoidable, of course, but the US authors have the gift of sprinkling them sparingly with plenty of plain English commentary and very welcome light-hearted and sometimes humorous touch. So you can learn EE from both US and English books of course, but the US texts are *far* less intimidating. The British texts approach remind me of the mechanistic, charmless character 'Gradgrind' in Dickens' Hard Times.

Engineering is an industrial art, like welding, and as such all the text books are manuals.

Oh, good grief! It should be 5G, even though 5G doesn't exist? Grow up!

Good luck with any of that.