Question for the experienced folks

When you guys first look at a schematic, how do you analyse it?

I mean how do you break it down. I sort of try to trace the path of
electrons which is fine for small circuits. But sometimes in a medium
sized schematic, I get overwhelmed trying to figure out where the
electrons are going and what's going on.

Where or what do you guys first look at when you look at a schematic.
How do you guys figure out what's going on. Are you guys "chasing"
the path of electrons as I do?

Is this something that comes only through experience or are there some
tips I could use.

Another thing is I am used to thinking of current as flowing from the
negative to the positive. That is, I envision the electrons going up
from the 'ground' or negative to the positive terminal of the battery.
Conventional current has it the opposite way. Which way do you guys
use when you look at schematics?

Thanks

 

I dunno about everybody else, but I concentrate on the
signals. Electrons just carry charge, and how much goes
where, when, is what usually matters.

The ends; load first to see what it does, then inputs
(signal, power, and ground) to see what it has to work with.

Nope. Break it into blocks (amplifiers/whatever or logic
gates as appropriate) and see what happens to the signals
between input and output.

Learn to see the blocks at a glance. Some draftsmen seem
to do their best to make it difficult to identify a given
block by routing lines all over the place to make another
block more obvious.

I'm old-fashioned (electron flow), but only about
power-supply-related stuff and toobs. It really doesn't
matter as long as you stay consistent.

Mark L. Fergerson

 

I think I generally try to see where the _signal_ is going. For an
overall view, I don't really care where the electrons (or mythical
positive charge carriers) are going unless I have to get into the gory
details of the operation of an individual stage.

I think I tend mostly to use conventional (positive) current.
Negative (electron) current was invented as an aid to teach
technicians how vacuum tubes work, since you can't really explain the
inner workings of a vacuum tube using conventional current, but it
isn't needed with semiconductors (it isn't really needed with tube
circuits either, unless you are trying to explain the inner workings
of the tube (particularly cathode ray tubes!!)).

 

I think some (or a lot) of what you are experiencing is because you
are knew to it. You seem to be analyzing each little piece, because
it is all new, but once you've become more familiar with it all you won't
be bothering.

Don't ask me how to get to the point, because it's been so long, and I
suspect it just grew organically. But you basically look at a schematic
and picked out the familiar. There's the input, there's the output, there's
the power bus. Oh, I recognize that IC, it does this..., so that gives
a clue about what the circuit is doing. That transistor over there is
a low power type, while that one is a power transistor. You know these
things because you've looked them up in the past, and the details stick
with you. You've understood what little parts of the circuit are doing
because you've looked at them in depth in the past, but likely in
the form a a simple schematic that only covers that one stage.

Often you will not need to analyze the whole thing, because much of
it will be familiar. But if none of it is familiar, then a large schematic
is of course very daunting.

And of course, a lot of us grew up with magazine articles (and books
for that matter) where there was a description, sometimes detailed sometimes
not, of what was going on. In the internet age, I suspect that may no longer
be the case, with many people extracting schematics from such articles
and simply posted the schematic alone.

Michael

 

Divide and conquer..
first identify where the power comes from..
Isolate that systema and remember without proper power and biasing you are
dead..
second find the ouput and first input..
then try to imagine how a signal gets thru or how it is generated...
the rest is support and glitter...
you might spend some time studying basic modules such as amplifiers,
oscilllators etc..
good luck
hank wd5jfr

 

The easiest way is probably to take up a 3 year university course
and study EE properly. Then, when you leave you will still not
know how to look at a large schematic, but at least you will have
the underlying knowledge, and you will
be able to start gaining real experience. Otherwise you may find
you are spending your time wondering how things work, without
realising the approach to solve these puzzles.

The proper way is the easiest way in the long run.

 

I'm not an expert. However, I've been getting much better at this in the
last few months.

What I've found is that most circuits are built out of 'standard'
subcircuits. After analyzing some of these, you get to know these standard
subcircuits, and at that point, you begin to see the circuit at a higher
level of abstraction.

There are exceptions to this approach, but its a pretty good starting point.

For example, there was a circuit posted on a.b.s.e. recently as a 'teaching
exercise' by Jim Thompson. Unless you recognize the common collector
amplifiers and the current mirror, you have no place to start. Trying to
understand this by analyzing the transistors with ebers-moll is pretty
difficult. However, by looking at the different amplifiers, you can get some
sense of what its doing.

I think a basic text on different subcircuits, like the first 3 or 4
chapters of "The Art of Electronics", for example, gives you the ability to
recognize these standard transistor and opamp subcircuits.

Regards,
Bob Monsen

 

OT question: What is a.b.s.e? Sounds like ng I need to check out...

Thanks,

Dave

 

ha........uahhhhh Well that means you need to know what compnents do and
certain configurations are noticable if you ever had any electronics
theory, it takes years to learn from an aspect of disecting a schematic
diagram.................i wish i could help, but their are far too many
areas to cover with you. There are logical approaches however you still
have understand your instruments, and waveforms etc........

Nells

 

(snip)

I break the circuit down into functions that I can briefly describe
and mentally try to keep the functions I have figured out separate
from what I am still working on. It helps if some conventions have
been followed in drawing the schematic, and if they have not been
followed, it sometimes helps if I redraw sections that are giving me
problems. My favorite conventions are that positive supply voltages
are above more negative supply voltages, so I can picture the positive
convention current sort of falling through the circuits. The second
convention is that forward signals pass through the circuit from left
to right, and feedback signals from outputs back to inputs passes
right to left. Of course, there are exceptions that do not easily
fall into these layout standards.

The other big one is to know the action of each of the components
cold, treating each like a black box. For instance, you should be
able to predict how a transistor will react to any sort of bias,
regardless which leads appear to be inputs and which appear to be
outputs. Mathematical rigor is not nearly as useful as a general idea
about what happens to one lead if you know something about what is
going on at the other leads.

I start where ever the concept of a signal is most obvious. For
amplifiers, this might be the input or the output. For oscillators,
the output is usually the obvious signal, and I work my way back into
the circuit, assuming it is producing the expected signal, and figure
out how that outcome i being produced. Microprocessors often do not
make a lot of sense unless you have some idea of the program. with
various pins being able to be analog inputs, digital inputs, digital
outputs, and some other functions. A lot can be deduced by what is
connected to the pins, but when someone has been especially efficient
at minimizing hardware, it can be a puzzle without looking at the
code.

 

Thanks all for your insight. The common theme seems to be it comes
with experience and knowledge of circuits commonly used as building
blocks.

 

If your really lucky you can find a detailed circuit description to go along
with the schematic.

This is what I loved about magazines like Popular Electronics, Nuts &
Volts... etc.

 

If you can identify a main component, you can search the net for a data
sheet which
can give you detailed info and sample circuits.

 

This is exactly correct. These are what Architects and Programmers
call "Design Patterns". Very little about any given circuit are
actually terribly unique.

1. Signals *usually* flow left to right. Power (positive to negative)
*usually* flows top to bottom.

2. High level signal processing is often done the same way in every
circuit of a certain type (e.g. radio receivers, op amps, etc. have a
predictably sequence of stages)

3. There are a finite number of ways to "do" amplifiers, oscillators,
etc. so the transistor level implementation is recognizable which
helps to identify high level stages. E.g. for amplifiers you have CC,
CE, CB and variant on those like cascade, diff amp and neut pairs.
Learn these rules of thumb.

4. Component-level elements (resistors, capacitors, transistors,
diodes) have particular characteristics that have highly predictable
uses. Capacitors always suggest blocking DC and passing AC with the
intent to preserve or eliminate one or the other. Transistor emitters
are always low impedance and are often used in places where having low
impedance has some advantage; the converse for collectors. More rules
of thumb.

5. Thevenin and Norton theorems often seem ridiculously simplistic and
useless to new engineers but they are worth their weight in gold. If
you can assume linearity, the entire universe can be reduced to a
single ideal source and impedance - sort of the EE equivalent to
Archimedes' lever; he needed a fulcrum, we need linearity. Collapsing
stages in you mind to a Thevenin or Norton can often help "seeing" the
circuit - consider the transistor terminal impedance rules of thumb
above in a multistage amplifier.

5. Train youself to be able to write loop and mesh equations "on
inspection". This means being able to look at something like a
transistor output circuit and write Vcc-IcRc-Vce-(Ic+Ib)Re=0 without
labeling currents or polarities on a piece of paper - just write out
the equations as you trace the loop or mesh with you figure. If you
can do this, you won't get bogged down with analysis mechanics. Sort
of like when you can finally play piano without "thinking" about which
keys to hit.

6. Remember that the human brain has difficulties with intuition of
anything that isn't linear. The math goes from conventiently simply
to virtually impossible when you go linear to nonlinear. Even master
analog circuit designers do nonlinear transistor design with the
simplest models in linear form. That's the whole bit about Q-points,
linearization and biasing. SPICE is for when you need to know what
the nonlinear effects *actually* are. Up to that point, between you,
me and the mouse in my pocket, we'll just assume the entire world is
linear.

 

I feel Mr. Black's pain. I, too, look at schematics and think "Eeesh" as I
try to trace each possible path of the electons/holes. I have the same
problem as he does of not recognizing "chunks" of the whole picture. I've
learned a little more by reading Nuts and Volts (same name as their web
site), but they quickly get over my head... funny, same is happening with
this news group... "Basics" to me is different from what most of you folks
are posting. Am I on the wrong news group? It would be cool if the
benevolent of this group would occasionally post a flash-card sized "chunk"
of a circuit. This would be good for the Intermediates out there - "they"
say the best way to learn a subject is to try to teach it. With time,
novices like I will have a nice cache of flash cards, and hopefully some
"chunk" vision like the big boys. Thanks, and good luck Mr. Black.

 

But haven't you found, like me, that there are many sources of that
sort of 'compartmentalised explanation' on web sites? And posts to
this and sci.electronics.design and alt.binaries.schematics.electronic
are sometimes accompanied by schematics with explanations.

For example, my own posts (whether in enquiring or helping mode), are
often complemented by schematics posted to my electronics web page at
http://dspace.dial.pipex.com/terrypin/indexpersonal.html. Or to
alt.binaries.schematics.electronic. Of the 300 or so files in that
list (which happens to be fairly up to date at the moment), quite a
few are schematics. And a fair proportion of those are accompanied by
notes and simulations or actual output waveforms showing how I think
they work. Maybe that resource could help you? At present the files
are not categorised, but most filenames are reasonably meaningful, so
selective browsing could be possible. Alternatively, specify a topic
and I'll suggest any potentially relevant web files.

Also, posted circuits are often analysed in depth by the real experts
here. A shining example in this news group is John Popelish. A
structured collection of his thorough and lucid explanations over
recent years, with associated schematics, could constitute an
electronics tutorial course that would be hard to improve on.

 

Yes, these groups do help, but the pure quantity is hard to sift. Also, I
have no idea who are the experts and who are the posers. I read a little
(books) and then surf a little, in hopes I a light bulb turns on. I still
find it hard to understand why a Basics news group is far beyond basic.
However; that I must accept.

 

Have you posted any questions about electronics basics? I don't
remember seeing you here. I don't think there is any question about
electronics that is too basic to ask, here.

 

I started out as a beginner by tracing electrons. After some years of
schooling I abandoned this, since that's not how circuits actually work.
In the education business, tracing electrons is called "the sequential
fallacy." Just try and figure out the brightness of several light bulbs
and switches connected in series/parallel networks and you'll rapidly find
that tracing the electrons doesn't tell you much. When the circuit breaks
into several "Y" branches, how can you know the value of current in each
branch?

Circuits instead work by Ohm's law; by Voltage Nodes and Current Loops.
If you know the voltage between the ends of a component, then you can
figure out the current inside that component. And if you know the current
inside a component, then you can figure out the voltage-drop across that
component. After lots of practice, thinking in terms of Ohm's Law becomes
automatic, and you start to be able to "see" the operation of circuits.




On the other hand, to figure out the general function of parts of a large
schematic, you simply need some experience in recognizing particular kinds
of simple circuits. For example, you might look at one handful of
transistors and see that it's a voltage regulator, so it must be and part
of the power supply on that board, while another bunch of transistors
forms a square wave oscillator, while another is an AC power amplifier.
You can draw squares around different parts of the schematic to separate
it into functional blocks.

Unfortunately beginners will have a terrible time doing this because they
lack experience in recognizing the standard kinds of circuits everyone
uses. Schematics with lots of ICs helps, since the ICs usually act as
functional blocks in the first place.

Build lots of simple circuits such as "common emitter power amp" or
"Hartley Sineway Oscillator" or "threshold detector." Play with LM555
oscillators and Op amp chips and Voltage Regulators. That way you get to
know them intimately, and can recognize them by their chip numbers.
Complicated schematics are made of building blocks, and once you learn to
recognize the building blocks instantly by eye, you can break the
complexity down into a "block diagram" in your mind.

Since electrons flow in some components, while positive charges flow
within others (batteries, gas discharge, electroplating, "proton
conductors" in fuel cells, etc.,) it's not correct to assume that
"electric current" means "electron flow." To be accurate, electric
current is ANY type of flowing charges. But that makes things hard to
deal with.

Also, knowing the polarity of the flowing charges is only critical if
you're under the sway of the Sequential Fallacy and performing "electron
tracing" I mention above. Think instead in terms of circuits being like
drive belts, with all circuit loops containing a belt inside the wires,
then the polarity of charges isn't so important. Circuits aren't like
hollow tubes with bullets flying through. They're more like bicycle
chains. Push a belt along, and the entire belt starts flowing.

See:

Which way does the electricity really flow?
http://amasci.com/amateur/elecdir.html

Franklin and positive electrons: he wasn't wrong
http://amasci.com/miscon/eleca.html#frkel


The advanced professionals (engineers and physicists) use a well known
method of simplifying this situation. They ignore what actually happens
inside the components and instead assume that all charges always have the
same polarity: an "unsigned" polarity, the positive one. It's mostly a
math convention: if a particular numeral doesn't have a polarity sign, we
assume that it's positive. Also, ammeters and resistors (Ohm's law) deal
in positive current, so if we avoid thinking in terms of flowing
negatives, then we avoid having to perform a "double negative" in our
minds each time we want to visualize things clearly.

And then, if the ACTUAL polarity of the flowing charges is important for
some reason, we can abandon the flow of "conventional" positive charges
and think in terms of several kinds of flowing charge. For example, in
liquid conductors there are two types of charge flows, postive charges and
negative charges flowing past each other in opposite directions. And in
transistors there are four kinds of charge flow, two kinds in the n-doped
parts and two other kinds in the p-doped parts.

You probably know that there's an older standard for charge, the negative
one. This standard evolved during the age of vacuum tubes when we mostly
had to visualize electron clouds in tubes and in wires. This made it easy
to convince ourselves that all currents really were flows of negatives,
and any components which used positive charge flow could be swept under
the carpet. Doing this makes it very easy to understand vacuum tubes, but
harder to understand the innards of semiconductor operation, or the
strange things that go on in the electrolyte between the plates in a
battery. As long as our circuits aren't dominated by vacuum tubes,
there's no big reason to declare the electron flow to be the "convention"
for charge flow.

Also see:

Typical beginners' mistakes in understanding electricity
http://amasci.com/miscon/elect.html

Collected electricity articles
http://amasci.com/ele-edu.html


(((((((((((((((((( ( ( ( ( (O) ) ) ) ) )))))))))))))))))))
William J. Beaty SCIENCE HOBBYIST website
http://amasci.com
EE/programmer/sci-exhibits amateur science, hobby projects, sci fair
Seattle, WA 206-789-0775 unusual phenomena, tesla coils, weird sci