Maker Pro
Maker Pro

Why is electronics so complicated?

E

Eric Marlow

Jan 1, 1970
0
Here are a few points on the above subject I have been wondering
about, from a novice's "idealistic" perspective. No disrespect
intended.

1. Why are there so many different parts that essentially perform the
same function? Basic transistors, IC's switches, capacitors, etc. Mil
spec aside, seems true in too many cases, and appears to compound
every year.

Fewer more versatile/robust parts should cost less by reducing
diversity of production. Saves on landfill too. Everyone would
benefit. Why does the economics of parts supply resemble a a pumped-up
consumer marketplace?

2. Why not adopt more logical and uniform color and numerical coding
systems? Four differnet ways of specifying cap values, etc., etc. And
who decided to make 1% resistor markings less readable by printing
them on blue.

3. The IC revolution changed electronics. How about a next generation
of ready-made customizable modules for limited production ANALOG
projects? Similar to what is being done for micros.

4. How about a comprehensive electronics website? Design walkthroughs,
value calculators, recommendations for replacement of older parts with
newer, and so on. A good wiki project for the pros. Corporate
sponsorship?

Yes, I have already heard a few reasons regarding the above.

-Technology is changing too fast.

- Little attention given to update or discard of old standards and
knowledge bases.

- Electroincs and its applications are inherently diverse and complex.

- Everyone is used to it the way it is. Time and money invested. Alot
of this already exists, if you can find it. Don't be lazy.

Why am I not happy with these answers? They often don't apply in the
practical work I have encountered, at least so far. I suspect many
others of low to moderate ability feel the same.

Why undergo a process that could take far less time if better
presented in terms of uniformity, design rationalization, resources,
and skills targeting. Instead. it seems like the Rubik's Cube of
modern science. Everyone "rediscovers" the same thing over and over
again..

If these more user-friendly conditions were actively promoted, alot
more young people would be drawn to electronics, as a profession or
hobby. No, I'm not giving up though. Once you've got the bug, it is
fatal.

Eric Marlow
 
Here are a few points on the above subject I have been wondering
about, from a novice's "idealistic" perspective. No disrespect
intended.

1. Why are there so many different parts that essentially perform the
same function? Basic transistors, IC's switches, capacitors, etc. Mil
spec aside, seems true in too many cases, and appears to compound
every year.

They are not the same. Sometimes one can choose from a number of similar
parts, sometimes not. Sometimes one can save money by using a cheaper part.
Sometimes the better/more expensive part is needed. Those sorts of tradeoffs
are part of the job.
Fewer more versatile/robust parts should cost less by reducing
diversity of production. Saves on landfill too. Everyone would
benefit. Why does the economics of parts supply resemble a a pumped-up
consumer marketplace?

Central planning is a *BAD* idea. The economics of parts doesn't resemble the
consumer marketplace at all, except that prices plummet, and for similar
reasons (there *is* that competition between competing parts).
2. Why not adopt more logical and uniform color and numerical coding
systems? Four differnet ways of specifying cap values, etc., etc. And
who decided to make 1% resistor markings less readable by printing
them on blue.

No one reads color codes anymore. ;-) Most parts don't even have them. Too
small.
3. The IC revolution changed electronics. How about a next generation
of ready-made customizable modules for limited production ANALOG
projects? Similar to what is being done for micros.

It's been tried, often. It never works acceptably, perhaps because analog
electronics doesn't do well in technologies that are designed for processors.
The Cypress PSoC is one of the later attempts at it. We'll see how popular it
becomes.

http://www.cypress.com/?id=1353
4. How about a comprehensive electronics website? Design walkthroughs,
value calculators, recommendations for replacement of older parts with
newer, and so on. A good wiki project for the pros. Corporate
sponsorship?

People have tried. If you think you can do it, go ahead. There are sites
like opencores, for programmable logic. It's unreasonable, IMO, to expect it
to all coalesce under one roof. What's in it for corporations?

<snip>
 
J

Joerg

Jan 1, 1970
0
Eric said:
Here are a few points on the above subject I have been wondering
about, from a novice's "idealistic" perspective. No disrespect
intended.

1. Why are there so many different parts that essentially perform the
same function? Basic transistors, IC's switches, capacitors, etc. Mil
spec aside, seems true in too many cases, and appears to compound
every year.

Competition. Every company and every country wants to have their own
little 40V NPN transistor :)

Fewer more versatile/robust parts should cost less by reducing
diversity of production. Saves on landfill too. Everyone would
benefit. Why does the economics of parts supply resemble a a pumped-up
consumer marketplace?

No, a good design engineer specs in various part numbers from different
manufacturers. The same goes for many other products. Tomorrow I'll have
to change the oil on our cars. I can buy a Fram oil filter, or a Bosch
filter, or a ... in the end they all filter the same oil on the same cars.

Or, why are there three different brand gas stations at many
intersections, all gas and prices being more or less the same?

2. Why not adopt more logical and uniform color and numerical coding
systems? Four differnet ways of specifying cap values, etc., etc. And
who decided to make 1% resistor markings less readable by printing
them on blue.

The topper is dark brown lettering on black ICs :)

3. The IC revolution changed electronics. How about a next generation
of ready-made customizable modules for limited production ANALOG
projects? Similar to what is being done for micros.

There have been many attempts but prices were too high, us analog guys
didn't bite, and the products fizzled in the marketplace. Cypress PSoCs
come closest to what you desire, they have configurable analog blocks in
there.

4. How about a comprehensive electronics website? Design walkthroughs,
value calculators, recommendations for replacement of older parts with
newer, and so on. A good wiki project for the pros. Corporate
sponsorship?

There's tons of them. Books like Art of Electronics (must have), site
like Discovercircuits that are suited for beginners, ham radio sites,
huge application note collections and so on.

Yes, I have already heard a few reasons regarding the above.

-Technology is changing too fast.

Not really. I still use CD4000 logic and 2N3904, like I did 30 years ago.

- Little attention given to update or discard of old standards and
knowledge bases.

- Electroincs and its applications are inherently diverse and complex.

That becomes easier with practice. A lot easier.

- Everyone is used to it the way it is. Time and money invested. Alot
of this already exists, if you can find it. Don't be lazy.

:)


Why am I not happy with these answers? They often don't apply in the
practical work I have encountered, at least so far. I suspect many
others of low to moderate ability feel the same.

Why undergo a process that could take far less time if better
presented in terms of uniformity, design rationalization, resources,
and skills targeting. Instead. it seems like the Rubik's Cube of
modern science. Everyone "rediscovers" the same thing over and over
again..

That's why we have Usenet. If you get stuck and have really tried, ask.

If these more user-friendly conditions were actively promoted, alot
more young people would be drawn to electronics, as a profession or
hobby. No, I'm not giving up though. Once you've got the bug, it is
fatal.

T'was the same with me. Best is to pick a few favorite parts and stick
with those a while, then slowly expand the selection.
 
E

Eric Marlow

Jan 1, 1970
0
So, Soviet IR imaging equipment was tremendously bulkier than its
Western competition (because of the need for lots of off-the-shelf
lenses instead of a few custom made ones), and it never worked as good.

But keeping the shelves stocked was easy.

The other side of the coin is that this phenomenom encourages
resourcefulness and ingenuity at the design level The Russians were,
and still are, world leaders in certain areas. Here is just one
example:

http://www.articlesextra.com/supercavitation-torpedoes.htm
And once you do impose standardization, how can you innovate? Over half
of the innovation that happens in the IC world is dreaming up new
specifications and designing to them -- how can you dream up new
specifications if the specs are dictated from On High?

Commercial innovation is not altruistic, but motivated by returns.
This is ultimately supported by the end user - that's us..

The shift of manufacturing to Asia has no doubt also contributed to
the unmanagable diversity of electronic parts. It would be fair to ask
though whether cost cutting, planned obsolscence, or marketing ploys
are true "innovations". I still prefer a phonograph over an MP3
player.
I have no answer to that -- but your 1% resistor markings aren't a
problem any more, because the current standard 1% resistor is a 0603 or
0402 surface-mount part with no markings at all!

Ah, I think I see a pattern emerging. Illegible markings were the
industry's way of priming us for no markings at all. Very clever then.
And the lead in the solder impairs our ability to think independently.
It's been tried. I've been a pro for over 20 years now, and I've seen
this done at least three times. Each one has been the coolest thing in
the world according to its marketing folks , and a big resounding flop
in the real world.

A notable example of successful modules in the analog world is audio
power amps. Function generator chips are close. But if you want an
"all purpose" DC to 100MHz linear power amp with adjustable offsets
and impedance get out the big bucks.
Part of the problem is that what you want would be hugely expensive.
It's not a matter of "don't be lazy" -- it's a matter of "don't bother
weeding the fallow fields when the harvest needs to be brought in".

It is already expensive. Not to suppliers but to individuals and small
businesses at ground level. How many old-timers have said, "here is a
newcomer asking THATquestion again". Multiply that by thousands every
month. Time is more than money. It is life, enthusiasm, and
intellectual resources that could be put to more productive use
elsewhere. There is the "harvest" you mentioned earlier in your post.

Eric Marlow
 
W

WhySoSerious?

Jan 1, 1970
0
And in the hardware world, washers can be of any material (nylon,
brass, steel, galvanized, stainless, aluminum...)
and any exterior dimension, internal hole size or even shape,
and some are even wedge-formed or spherical. All perform
the same function, essentially. Just, not the same function, exactly.


Somehow, I fear that your explanation is not descriptive enough *for
him*, in an all too descriptive way.

If one wants to make a meter needle move and indicate a certain level
of a certain event to him, one must make a circuit that converts between
the signals of the source to the signals that the meter needs to provide
to observer with a meaningful indication of the desired effect, event,
level or such.

There is no "hook up da meter and it goes" "off the shelf" solution.

That is true in nearly ALL cases. The circuit and meter that measures a
bullet speed, and the meter that measures your body temperature are two
entirely different measurement devices, and require two entirely
different circuits in front of the actual "meter" to drive it.

We just happen to call them both "a meter".

So, your (the op) assertion that the "parts" are "all essentially the
same" is ludicrous, at best. It was ludicrous back at the beginning of
the previous century, and it is more so at the beginning of this century.
 
T

TheGlimmerMan

Jan 1, 1970
0
its a free market, people keep thinking they can do bettter.



not with trs, where you get a mix of device specs on each wafer. Its a
case of how can we specify our tr models to maximise customer desire
and total income per wafer. And naturally, opinions are ever varied.



it doesnt



I've wondered about that too. The downside is that the parts wont
match end user's desired specs well, and a lot of the component
modules sold dont get used, which tends to up the price. Nonetheless
for some largeish systems it could prove a very versatile chip,
effectively a tiny analogue computer.



massive project. Who would pay for it and why?



not really a problem. Date the presentations, and present them in an
order that takes into account how often they're read, which will
follow the relevance of teh technology at the time.



hugely so. To do what you suggest in just one area would be
prohibitively expensve, and you've not offered a viable business
model.




If you've got enough nouse to do electronic engineering you can teach
yourslf, there's no real need for what you're proposing afaik.


NT

If this guy has EVER used a cell phone, he needs someone to tell him
that the "circuit" in that phone would take up a basketball court sized
building to house it id it were back in the 50s when Dick Tracy comics
were around.

We make them smaller and smaller so we can put more inside each element
and make the gadgets we make with those elements work better, use less
power, and take up less space.

The "parts' in now way shape or form "perform essentially the same
function", or we could get online with our clock radio.
 
E

Eric Marlow

Jan 1, 1970
0
So, your (the op) assertion that the "parts" are "all essentially the
same" is ludicrous, at best. It was ludicrous back at the beginning of
the previous century, and it is more so at the beginning of this century.

Who implied "all" parts needed to be the same? Obviously not possible.


Granted, with specialization, this ideal becomes less relevant. But
what about the rest of the field? The foundations, so to speak, that
set the tone for this discipline.

I am referring to more "basic" applications, and there are far too
many to mention, that could benefit from a periodic consolidation of
both design and componentry.

IMHO there is a tipping point beyond which choice and redundancy costs
disproportionately more. If you don't believe it, take the misses into
a store that sells specialty bathroom fittings :)

Elsewhere, this type of cost may be better hidden, beneath a veneer of
complacency.

Eric Marlow
 
J

Jasen Betts

Jan 1, 1970
0
Here are a few points on the above subject I have been wondering
about, from a novice's "idealistic" perspective. No disrespect
intended.

1. Why are there so many different parts that essentially perform the
same function? Basic transistors, IC's switches, capacitors, etc. Mil
spec aside, seems true in too many cases, and appears to compound
every year.
competition.

Fewer more versatile/robust parts should cost less by reducing
diversity of production. Saves on landfill too. Everyone would
benefit.

Da comrade!
Why does the economics of parts supply resemble a a pumped-up
consumer marketplace?
competition.

And
who decided to make 1% resistor markings less readable by printing
them on blue.

blue is for metal film. 1% is the brown band spaced differently at one
end.
3. The IC revolution changed electronics. How about a next generation
of ready-made customizable modules for limited production ANALOG
projects? Similar to what is being done for micros.

op-amps, power amp modules, such things exist.
 
J

Jasen Betts

Jan 1, 1970
0
But why do they want to change the pin configurations? I found a NPN
transistor F370 or somesuch that has the collector and emitter
reversed. They kept the base in the middle, but switched the other
two. It works just fine if you turn it 180 degrees.

Perhaps that's for when you get the silkscreen wrong :)
 
E

Eric Marlow

Jan 1, 1970
0
There were fast and slow op amps with low input bias currents, fast
and slow op amps with low offset voltages, a variety of of op amps
with relatively low - and well specified - input noise figures, and a
bunch of relatively fast op amps, some designed for regular voltge
feedback, and some designed for "current feedback".

Do you ever get the impression the manufacturers are not producing
more versatile or tailorable packages just so they can move more
stock?

Eric marlow.
 
Do you ever get the impression the manufacturers are not producing
more versatile or tailorable packages just so they can move more
stock?

What are you talking about? Tailorable packages? You want to move pins
about? Isn't that what you want; manufacturers forcing you into a small set
of parts so they can sell more stock?

I'm having problems with CPLD or low-end FPGA packages. They don't produce
packages that suit our applications. They've decided that they'd rather not
support small customers who can't afford the latest production techniques
(blind vias, via-in-pad, etc.). I understand it and will design around those
manufacturers, but they will get told why I'm not using their parts.

OpAmps and the like, come in all sorts of packages. The same devices
generally come in singles, duals, and quads. We generally use duals for
wiring/layout/cost reasons but lately I've gone to a few quads.
 
Who implied "all" parts needed to be the same? Obviously not possible.


Granted, with specialization, this ideal becomes less relevant. But
what about the rest of the field? The foundations, so to speak, that
set the tone for this discipline.

I am referring to more "basic" applications, and there are far too
many to mention, that could benefit from a periodic consolidation of
both design and componentry.

Why? When a 1000 ohm resistor costs a tenth of a penny, who cares if there
are a hundred different $.001 resistors? Consolodating that to one resistor
isn't going to save anything, quite likely the opposite (because they aren't
the same).
IMHO there is a tipping point beyond which choice and redundancy costs
disproportionately more. If you don't believe it, take the misses into
a store that sells specialty bathroom fittings :)

Redundancy? If you think all bathroom fittings are the same, *you* haven't
asked the misses.
Elsewhere, this type of cost may be better hidden, beneath a veneer of
complacency.

What cost. You've identified *none*, or any savings that could be had. Given
the cost of electronics over the past fifty years, you're barking up the wrong
tree!
 
Compared to software, electronics is elegant, simple, obvious and easy
to learn.

"Why undergo a process that could take far less time if better
presented in terms of uniformity, design rationalization, resources,
and skills targeting. Instead. it seems like the Rubik's Cube of
modern science. Everyone "rediscovers" the same thing over and over
again.. "

Seriously, learn some programming and I think you'll come back to
electronics real quick!

I think modern electronics is the worst thing that happened to
software...

No, modern programmers were the worst thing that's happened to software. 'C'
is the worst thing that's happened to modern programmers.
With the computers we have now, there is no consequence to
bad software, it all runs the same.

Badly.

There is no consequence to bad software because customers expect bad software.
 
J

Joerg

Jan 1, 1970
0
Jasen said:
Perhaps that's for when you get the silkscreen wrong :)

Seriously, in the SMT world there are "R" versions. Back in the late
80's we called them the "Oh s..t!" versions that could save the day.
 
Seriously, in the SMT world there are "R" versions. Back in the late
80's we called them the "Oh s..t!" versions that could save the day.

The ones (SOT23 - SOT23R) I was looking at recently (Diodes LDOs) had pin-2
fixed and swapped pins 1 and 3. To fix an "aw s..t", I would have expected
pin-3 fixed and 1 and 2 swapped. I just figured that they were trying to
match the pinout of some other part(s).
 
J

Joerg

Jan 1, 1970
0
Eric said:
The other side of the coin is that this phenomenom encourages
resourcefulness and ingenuity at the design level The Russians were,
and still are, world leaders in certain areas. Here is just one
example:

http://www.articlesextra.com/supercavitation-torpedoes.htm

They often build one prototype or maybe a few for the military. In the
consumer world it's all bleak and dry, no leadership there. Sometimes
their people can't even put enough food onto the table. My definition of
"world leader" is slightly different.

And it certainly isn't related to common part numbering. At least when
the iron curtain was still up the countries behind it were literally
begging to get their hand onto western parts. But they needed Dollars or
Deutschmarks for that which were hard to come by for them. Although,
they had a way of milking western relatives of their citizens for those.

Commercial innovation is not altruistic, but motivated by returns.
This is ultimately supported by the end user - that's us..

The shift of manufacturing to Asia has no doubt also contributed to
the unmanagable diversity of electronic parts. It would be fair to ask
though whether cost cutting, planned obsolscence, or marketing ploys
are true "innovations". I still prefer a phonograph over an MP3
player.


Ah, I think I see a pattern emerging. Illegible markings were the
industry's way of priming us for no markings at all. Very clever then.
And the lead in the solder impairs our ability to think independently.

It's cheaper. Burning a brownish lettering into a black plastic chip
costs less than silk screen printing.

A notable example of successful modules in the analog world is audio
power amps. Function generator chips are close. But if you want an
"all purpose" DC to 100MHz linear power amp with adjustable offsets
and impedance get out the big bucks.

Real men roll their own :)

It is already expensive. Not to suppliers but to individuals and small
businesses at ground level. How many old-timers have said, "here is a
newcomer asking THATquestion again". Multiply that by thousands every
month. Time is more than money. It is life, enthusiasm, and
intellectual resources that could be put to more productive use
elsewhere. There is the "harvest" you mentioned earlier in your post.

Time is money, but we also have an obligation to pass knowledge on to
the next generation and not take it with us when we die. This
automatically means there will always be newcomer questions, even when
you yourself have reached the ripe old age of 70 or so.
 
On a freezing cold day
"[email protected]"
slipped on the ice, and cried

You are a complete clueless idiot.
C rocks.
Just finished an other program in it.

I rest my case.
You cannot even do modern electronics without C.

What a lie.
BASIC sucks sucks sucks sucks sucks sucks.

Wow, that's an impressive statement.
C is universal it *shines*.

You can try to shine a turd, but it's still a turd.
 
I wonder.

Yes, it's a wonder you can breathe.
Well, what other languages than BASIC do you actually program in?

I don't program in BASIC, dickhead.
It is a fact.

What a stupid shit!
Lack of clout and now go into the abstract?

No need to go into the abstract. Turd are physical entities. You can't
polish C, so I suppose that would be an abstract idea.
C is about as simple as programing gets.

....and you show how simple one can get.
Unlike asm with zillions of different instructions for as many different processors,
C knows only a few statements.
With those few statements you can do ANYTHING (Turing).
It allows you to make your own statements, commands if you will, so is 100% extensible.
And it also allows 100% exact typing.
What more could you wish than somebody else do the thinking for you :)

All completely irrelevant.
There is a C compiler for practically every processor.

When is the last time you wrote a C program?
If ever.

I have, but now I don't admit to knowing how to spell C. Ew, even the thought
makes me feel dirty.
 
Mmm.... I think you'll find that in both software and electronics, things can
get as complex as you'd like them to be. It's not at all "elegant, simple,
and obvious" how you build, e.g., a 1GHz op-amp with low noise, input
capacitance, and offset voltages, nor is it entirely simple and obvious how
you might implement something like a hardware-based MP3 decoder once the real
world throws on the constrain of it having to cost $0.25 in quantity.

I do think software tends to be a bit more forgiving in that fixing flaws or
adding features often has no "per unit" cost... whereas with hardware it
usually means board spins or blue wires.

That only means that you *will* have to fix flaws in the software, which costs
a *lot* more than people are willing to admit.
 
True... although today pretty much all CPUs come with bugs (errata) as well,
right? Intel's FDiv bug was incredibly expensive, even though realistically
it didn't affect that many people at all.

Sure, all CPUs come with bugs but as you note *very* few are ever seen. The
FDIV bug was an example of the worst. OTOH, all consumer software is riddled
with bugs. I suppose it's all cheap to "fix", because it never is.
Even op-amps that jump to the opposite rail of what they "should" is arguably
a bit of a bug.

I don't consider unexpected behavior outside of a part's spec to be a bug in
the part at all. It is *certainly* a bug in your circuit.
But software takes the cake insofar as bugs adding costs, IMO.

Yep. I get yelled at for bugs. Programmers get time to make more.
 
Top