Develop a "feel" for electronics

[Brumby]

The side affect of a voltage drop is always heat of some sort, is that correct?

Not strictly correct.

A loss of energy will almost always result in generating heat - but as long as the energy remains within the system, heat does not necessarily get created.

When current flows through a resistor, there is no means for the energy to be kept within the system (the resistor) and it will heat up.  However, capacitors can store energy in an electric field and inductors can store energy in a magnetic field, so heat isn't an immediate issue.

The real world is a little more complex and the lines might seem to blur a bit when you start working with actual components.

[vk6zgo]

I am determined to learn electronics from the ground up and to be completely familiar with all the concepts.

It understand all the maths but it still doesn't give me a real "feel" for the concepts.

One writer I came across talked about thinking of the capacitors in a circuit as actually rising and falling (in space!) as a way of intuitively understanding what's occurring.

This is the kind of intuitive feel tha tI want to have.

Does anyone have suggestions as to books/authors/resources that could help here please?

Cheers.

Go back to square one!

Get to know your way around resistive circuits first.
Learn Ohm's Law,---think about it as a real thing,not as just some "Maths".
Try to visualise what is happening.
Learn Kirchoff's Law---try to analyse circuits using it.


Capacitors:-

Capacitors almost always need  to be explained using models which are not necessarily correct,but do to get the idea over.
This is how I learnt:-

  WARNING! WARNING!-refers to Electron Flow--EEs may want to turn away,now.

Initially,a capacitor has no charge but it does have a few free electrons on its plates.

When it is connected to a battery,free electrons at the positive connected plate are attracted to the positive batt terminal.
Current flows until that plate is at the same potential as the positive batt terminal.
It now appears positive with respect to the other plate.

This sets up an electric field between the plates.

Free electrons in the negative plate are attracted by the positive plate's potential,but cannot flow through the dielectric,& mass at the plate surface as close as they can get to the source of attraction,robbing the rest of the plate of  free electrons.

This looks like a positive charge to the batt negative terminal & current flows until the capacitor negative plate & the batt terminal are at the same potential .

The capacitor is now "charged".

AC current through capacitors is simply a series of charge/discharge cycles.


This description is better with a blackboard & diagrams,plus it is simplified to help people get a working understanding of how the things work.

It doesn't happen in nice steps like in the above but simultaneously.
Also,"free electrons" is a bit of a simplification for "electrons which may be readily dislodged from their associated atoms".
The "Electric Field" in the dielectric is shown in various ways in different diagrams.



You can chase these up later,but right now,this model makes a lot more sense than hydraulics & diaphragms.

[miguelvp]

The floating capacitors above Chicago just dumped a lot of rain and some lightning

[free_electron]

One writer I came across talked about thinking of the capacitors in a circuit as actually rising and falling (in space!) as a way of intuitively understanding what's occurring.

 Analogies can be very useful, however everyone's head is wired differently so you have to find what works for you. For capacitance I use the analogy in physics of a spring, a thing that can charge or discharge energy.

a capacitor is a big tube blocked in the middle by a balloon. pump wate rint he tube and it will squeeze the balloon flat. release and the pressure int he balloon will expand forcing the water back out. apply too much pressure and the baloon bursts. just like a real capacitor : the dielectric shorts out.

[dom0]

However, I suspect that I have skipped or misunderstood some things. For example, a BJT current mirror: I can see how the current setting side works, I can see how the varying load side works but how does the first side transmit the "information" to the second side? Is it the small increase in base voltage - this varies very little from the 0.6-0.7V as both BJTs (npn) have grounded emitters. I'm definitely missing something!

It could be to do with this current-to-voltage and voltage-to-current converter idea. To me, this sounds like converting apples to oranges!

Almost all amplifier circuits and all translinear circuits (like a current mirror) cannot be understood when thinking of bipolar transistors as current-controlled devices. To understand them as voltage-controlled devices (with a necessary, though wildly variable, base current) is not only necessary for understanding many circuits, but is also physically way more accurate (consider a switching application as a simple example).

One of the biggest mistakes in text books and other introductory texts to transistors is that they tend to state something like "Well, guys, a bipolar transistor has three terminals, base, emitter, collector, and the base current controls the collector current. Well, okay, that's a simplification, we'll see a better model on slide 1814".

--

Re. specifically the current mirror. Q1, the input transistor, Q2, the output transistors. Q1 base-collector shorted, hence Q1 develops a Ube exponentially related to the input current (=> Shockley's equation). The emitters of Q1 and Q2 are tied together, as well as the bases are tied together. Hence the Ube of Q2 is identical to the Ube of Q1, hence the emitter currents are identical.

(This description disregards all errors and is only meant for a first understanding of the circuit)

[MrSlack]

Very true.

ARRL Expermental Methods in RF Design has a good chapter on transistor and amplifier models that isn't bad. I learned how it all works from the predecessor of that book: Solid State Design for the Radio Amateur. Even Art of Electronics beats around the bush a bit too much on this front.

That book and blowing up lots of rather expensive gold leaded Motorola 2n2222's was how I worked it out. It's the same way I learned to ride a bike: do it by riding around in circles on on concrete. I was too scared to fall off due to the consequences

[PerranOak]

Thanks again all.

Part of my problem is this voltage to current conversion.

I've become confused by the BJT being a "current amplifier" or being a "voltage controlled device", dom0. I suppose it's easy to think of the base current being amplified by beta to give the emitter current. As base current increases so does emitter current as the BJT turns on more. So, to think of it as a voltage device means that the base voltage (Vb or Vbe?) must increase to turn it on more. BUT they always quote Vbe as 0.7V-ish (an attribute of the BJT itself) so this can't increase therefore it must mean Vb, yes? BUT Vb increase relative to what? 

Another thing is that I imagine a sinusoidal signal at the base of a BJT causing a larger signal at the collector. However, this only seems to occur because the BJT changes its resistance with the input - input rises, resistance lowers - so, as the resistance of the BJT falls it is like a voltage divider with less of a "share" of the Vcc and so the output signal falls. Is this right?

I did burn a 2N2222 and blew the fuse in my DMM - I learned that current mirrors don't work so well with saturated BJTs! 

[tggzzz]

I've become confused by the BJT being a "current amplifier" or being a "voltage controlled device", dom0. I suppose it's easy to think of the base current being amplified by beta to give the emitter current. As base current increases so does emitter current as the BJT turns on more. So, to think of it as a voltage device means that the base voltage (Vb or Vbe?) must increase to turn it on more. BUT they always quote Vbe as 0.7V-ish (an attribute of the BJT itself) so this can't increase therefore it must mean Vb, yes? BUT Vb increase relative to what? 

You need to be a little more precise in your thinking; good quality textbooks would be a starting point.

For this topic you need to find the equations relating current and voltage in semiconductor junctions (look for the exponential), the relationship between I_C I_B and I_E, and the difference between "large signal" and "small signal" models.

Good explanations take a lot of time and work. I certainly don't have time to regurgitate bad explanations!

[dom0]

The thing is that Ube <=> Ie ~ Ic is an exponential relationship, so it takes comparatively little change in Ube for a large change in Ie, approximately ~20-25 mV per Ie doubling. Take a look at Shockley's equation.

Ube (for the same Ie / Ic) will also have a spread of about 10-50 mV between transistors of the same type and manufacture, and a little more between types.

So while Ube in an absolute sense is a poorly controlled parameter the relative change of emitter current to base-emitter voltage is well defined over many decades of emitter current.

This also explains the need for proper (stable over different transistors, time and temperature) biasing of amplifiers ; using a fixed voltage bias or even a fixed base current bias works well for quick simulations and the like, but never works in practice.

Ube <=> Ub: Ube / Vbe means the base-emitter voltage. Ub / Vb sometimes means the same, but sometimes means the base-to-ground voltage ; which one depends on the author, I've seen both.

The figure of Ube ~0.65 or 0.7 V is often used in amplifier design for calculating bias points and the like, because at typical currents (couple hundred µA to a few mA) it is close enough, and the figure doesn't need to be accurate - biasing should (see above) not be very sensitive to the exact value of Ube.

[JacquesBBB]

To get a feel for electronics, I  see two good (among many others ) possibilities

- The water analog, as stated above. You can get many examples in "Practical Electronics for Inventors" by Scherz and Monk.

- Play with the Everycircuit simulator
http://everycircuit.com/

https://youtu.be/vcXZylSj9DI

[orolo]

About current mirrors and learning. My experience is that gaining intuition is related to useful paradigms: the hydrodynamic analogy is just a metaphor, it will never fit well enough. A useful paradigm in this case is the translinear principle: https://en.wikipedia.org/wiki/Translinear_circuit . A current mirror is the simplest translinear circuit there is, and translinear loops elegantly capture not only the equal currents in mirrors (multiple mirrors too), but also in unequal mirrors like the ones you will find in voltage references and other ICs. My favorite reference on translinear circuits are the Translinear Lores: http://cas.ee.ic.ac.uk/people/dario/files/E416/translinear-lores.pdf . The translinear paradigm intuitively translates current sharing to diode loops in a circuit. That's an useful insight. You grasp the global paradigm, then reduce the particular cases piecemeal.

All the mystery in the current mirror is just a simple application of Kirchof's Voltage Law, the exponential law in diodes, and the astute observation that exponentials turn products into additions. And to assume that the transistors are pretty well matched. With that unifying principle, you can explain some pretty hairy circuits with great ease.

The fact that Vbe in an active transistor is almost always about 0.7V is also a consequence of the exponential law. If the base current Ib = Is*exp(Vbe/Vt), then Vbe = Vt * ln(Ib / Is), so you need to square the input current into a transistor (referred to the leakage current Is, which is well into the nanoamp range) to double a voltage. It is clear that voltage won't change much once it reaches its active value. You would need transistor-burning currents. In other words, Vbe is logarithmic in Ib, and logarithms rise slooow past unity. And unity is in the nanoamp scale for a Si transistor.

Getting to know the subtleties of Vbe is key to understand how transistors work. The matter of Vbe temperature dependence is another fundamental concept. To get an intuitive grasp, I found Bob Pease's articles on "What's all this Vbe stuff, anyhow?" (parts 1 and 2) absolutely essential: http://www.ti.com/ww/en/bobpease/assets/www-national-com_rap.pdf , page 113. Pease turns the complex Vbe-temp dependence into something intuitive with a graph and some common sense (not to mention geniality and a lot of hands-on experience). There is also a great article on transistor matching in page 111. And a ton of useful info, the document is a gold mine.

I did burn a 2N2222 and blew the fuse in my DMM - I learned that current mirrors don't work so well with saturated BJTs! 

Hmm, burning. Anyway, you probably dug up already: http://www-mtl.mit.edu/~jldawson/documents/jldrec13.pdf  Adding emitter degeneration adds some feedback and 'matches' the transistors, preventing runaway, at the cost of some compliance. The article also gives an elegant, standalone explanation of the mirror from first principles.