Electronic ignorance

Yes Neil but the basics are still the same what ever devices are used within reason. Some are just simpler to use than others. Once some one has a grip with other discrete devices the answers that are needed are in the data sheets on specific devices. The biggest problem with this area is that it does need some sums to be done. Electronics generally does even for power switching circuits if some one is worried about how hot it might get for instance. Linear use will generally be harder in as much as it usually needs more sums. IC can make that area a lot simpler.

Actually it's a good idea in some ways to go back even further and get to grips with buckets full of transistors in both linear and digital circuitry modes. The biggest problem with that is the number of thick hefty books that are about full of pretty useless ways of handling them. I linked to a light weight thin one that contains no junk. Schaum's isn't too bad and adds to that and there are several other on far more advanced aspects of electronics.

If some one wants to go retro a book called Electronic Designer's Handbook by Hemingway may be about 2nd hand from people like Abbe books. It covers transistor biasing pretty well but could be a bit clearer in places. Lots of "interesting" circuits in it. Hemingway used to be the head of electronics at BAC. It dates back to 1966. It lacks the important aspect to some extent - using transistors effectively means designing the circuits so that the actual parameters of the transistor don't matter too much. The passive elements have to be arranged to do that.

It's also possible to do virtual electronics. My favourite was electronics work bench. It's cute so some electronics people don't like it at all. Circuits are simulated and it's possible to connect virtual oscilloscopes and volt meters etc to it to actually see and measure what's going on.

It looks like there is still a free version about. Installation shown here

**LINK**

And downloaded from here

**LINK**

Might be wise to virus check it. It is an old version though.

John

-

I looked at the Ferranti document you linked to.

Section 1 dives straight into speed up capacitors for switching transistors - hardly a gentle introduction for beginners...

Neil

Electronics is not "obvious" in the way that mechanics are. That makes it harder to learn. You have to be able to do some maths to be able to figure out what it happening. Somehow you have to trust that your understanding, with the maths applied, really is what is happening.

If you get that bit wrong, then you won't figure out why it doesn't work without trial and error, which is going to be slow. Trial and error can sometimes be the answer, but without the maths you don't know how big the error was. The consequence is that you have no idea what to try next.

I've heard it called "stabbing in the dark". It is why electronics is difficult.

Mechanical engineering is not the same. There are things you have to know, but if the widget isn't long enough to reach, then you have to make another which is longer. If it keeps breaking, then it has to be thicker or made of a stronger material.

Mechanics and electronics are different beasts. Mechanics can be the art of making things to someone else's design. Hobby mechanics can simply be development of the skill and knowledge required to make something that looks nice and works well, from raw materials.

Electronics is different. By comparison there isn't much art in making a radio to someone else's design. It is enough for some people, but you can make the radio and it looks terrible, but works just as well as an immaculately constructed one.

The "art of electronics" is understanding and optimising what is happening. If you don't want to understand it, then there isn't much art to it.

I wouldn't recommend reading "The Art" from cover to cover. It's not that sort of book. The whole point of it, is that you get an electronics cookbook. If you want you can buy your own cookbook with circuits that you like, or want to explore. When you have built your circuit and it doesn't work, or you just don't understand why it does work, then "The Art", will explain why.

It probably won't answer your question directly, but if you use your brain, it will help you to realise the answer for yourself.

To those that don't like the book, I suspect it's because you've not tried hard enough.

Back again, a book I can recommend is Electronic Devices 4thEd by Thomas L Floyd it's published by prentice Hall and has 960 pages. The comprehensive book take you through basic power supplies to Ic's using test meters and using an oscilloscope, it contains very good coloured illustrations and it shows you typical meter reading or what to expect on the CRO. There are good illustrations of little circuits to make and test, with enough theory for a beginner without being to mathematical on the way. My copy must be 20 years old so you may get a copy via ebay or your local library .Looking in the contents it start with Pn junctions, diodes and their applications ,zenners, transistors, thryristors,triacs,opamps regulators, all very good..John

I know that it can be a difficult subject to get into. I've been there and done it. In my case that had to be transistors. The problem with those is that there were reams of books about full of all sorts of info based on hfe, HFE, ft, cbe and many other things. Big problem - these in some ways are irrelevant to the practical use of them other than things like if some one wants to design a high frequency amp it needs a high frequency transistor so high ft might be a good idea however gain/hfe comes into it as well. It's possible to do reams of analysis on this sort of things and plenty of books have been about that are full of it but the practical outcome is likely to differ so in many ways it's better to just get on with it and measure it. What is missing from all of these books on areas like this is that using a transistor where it's properties can influence the results is a bad idea so circuits are designed so that passive components set what it actually does not the transistor, that just has to be good enough. This is done by using feedback. The current that flows in a transistor in it's dormant state is also set by feedback but might be referred to as biasing. What makes op amps, comparators, schmit triggers, integrator and all sorts of things work - feedback some times positive and some times negative. All feedback is really is a set up where the output from something is connected back to it's input and that aspect dominates the behaviour of what ever it is. If some one reads up on feedback they may find it all described via black boxes. A black box is just something that has an input and an output. It doesn't matter what is actually in it.

All of this doesn't help people to pick it up but just about every aspect of it involves variations on ohm's law. Kirchoffs variations on it can be very useful. This may include reactive components, inductors and capacitors. In a simple sense these are just resistor that vary according to the frequency that is being applied to them. They can also store energy. They can also cause something called phase shift which can be good or bad.

A much better pages on Kirchoff here. There are probable other good ones on the site as well.

**LINK**

A problem though. From a beginners point of view it's of no use what so ever. However as a for instance when some one looks at a circuit they can look at a node where say 2 resistors join and something else is connected they can say ahh the sum of the currents at that none is zero. Given that they can calculate what the actual voltage at that point will be and if say the thing connected to it was a transistor how much it could vary. That would allow them to pick resistors to obtain a current large enough to ensure it didn't vary too much. The could then put a resistor from the emitter to ground - that will set the current flowing through the transistor. A form of feedback that isn't immediately obvious. Now they have set the current they can add another resistor from the collector to the supply. They might chose a value that splits the available voltage from the emitter to the supply in half. They now have an amplifier. The gain will be roughly the collector resistance / the emitter resistance. Feedback again. If the gain of the transistor was infinite it would be exactly that. It wont be so the actual gain will be less. Op amps have a hell of a lot more gain so get a lot closer. Circuits with more transistors can also have a lot more gain so over all feed back can be used to define how it behaves. For things like audio amps feedback is used locally on each transistor and over all. The earlier active parts might need to be low noise so a low noise transistor would be used. It's data sheet is likely to give typical parameters at some current level - that current level is likely to be the one to use but even lower may be better - that sort of aspect is a bit of a black art. Designing for minimum distortion is too but not so bad as it relates to gain variations as the current varies. These days a low noise op amp it likely to be a much easier solution. They use feedback and summing at nodes too as do comparators, monostables etc. Positive feedback raises it's head in that are. The arrangement on the fictitious transistor was all negative - the feedback is inverted so output reduces input. Positive feedback increases it in some fashion. An oscillator is another example of that.

I've always wondered why some one hasn't written a book presenting the info in a similar fashion. Maybe it would be too thin and too simple to understand.

Arggg sorry about the typo's. No corrections this time.

John

-

Edited By Ajohnw on 31/07/2016 11:20:30

The reality is that you will get better advice from industry experts and / or suppliers than from a "one size fits all" book. So for power controllers, refer to TI-Unitrode, for FETs refer to IR, for IGBTs refer to Japanese manufacturers, for magnetics read EC Snelling etc etc. Conversely, The Art isn't much use for beginners either - makes it jack of all trades and master of none in my books.

Despite 40 plus years of hands-on electronics as both a professional and a hobbiest I'm still not a disciple of The Art. Is it something I should be trying to achieve? Would it make me a better person? Would I reach some kind of Nirvana?

I prefer to call it the 'multiple monkey' approach to circuit 'design'. If you have enough monkeys, each with a supply of components, who randomly select said components, then every now and again you'll end with a circuit that does something useful. It's surprisingly common, even in the world of professional electronics.

Andrew

Gosh this makes me feel old. I looked at the Horowitz book (managed to find a PDF) and wished I'd had that as my undergrad text, it's much easier to make sense of than the ones I had at the time. BUT the point is it's an undergrad textbook, which in my case would have been supported by probably two, if not a full three years of lectures and labs to create understanding. The first two years of that course were also supported by Physics (If I live to be a 100 I'll never understand Brillouin zones) and Mathematics (which by the end I decided was really Flute music and I didn't have a Flute...).

What I'm saying is, it's a rare bird who can read an electronics textbook and unaided then say "Ah, I get it!".

In the thread above it was mentioned that in mechnical engineering you can see whats going on and in electronic engineering you typically can't. This URL points at a free HTML simulator, that as easy as it comes to use (I know there are more technically and higher performance ones, but for me they are too difficult to get into).

All you need do is use this link

**LINK**

where I've created a simple diode rectifier circuit. You can add a 'scope view and watch how things change as you change things around. You can then add in a transistor and again, wander around and see what's going on. Best idea I can come up with to support the 'Art' book.

Steve

Edited By Steve Withnell on 31/07/2016 11:52:54

Edited By Steve Withnell on 31/07/2016 11:53:18

I couldn't possibly say!

I rather like the 'Art of Electronics'. While it's not the be all and end all, it is one of the few books that sits between the blind faith cookbooks and the more academic books that aim to teach you something that you can be examined on. It deals with the real world imperfections and gotchyas that are rarely mentioned elsewhere. Having said that it's not a beginners book, and isn't intended to be so.

Andrew

This post is a very good example of why it is difficult for the likes of me to "get into " modern electronics. I've read a couple of the recommended books and chased the links, but still no wiser. After the first two or three replies I, for one ,am lost. People are trying to help I know, maybe it is just me. I should point out that I am quite good at maths and up to the introduction of solid -state stuff had no problem with electronics or power engineering. Maybe it is that "you can't see what is going on ", I could visualise what was happening in , say, a triode valve when the grid was powered up, but not in a solid-state chip ? The jargon does not help, but this can be learnt.

I was very lucky. My day ran a business renting out TVs and selling/repairing electrical goods. Hi mate* taught electronics for the CEGB and got me into the digital stuff (Dad was a valve man and still looks at transistors in a funny way...)

It really helps to have a mentor.

Neil

*He also only had half a thumb, due to an accident with a brass cannon. He said it should teach me a valuable lesson (it did and as a result I was always very careful when playing with pyrotechnics )

I found Solid state electronics particularly hard to get a handle on, but you have to treat it like the Romans would have.

Divide and conquer.

A discrete transistor is a special case of the solid state chip. It only has one transistor. Unlike a vacuum triode, you will be hard pressed to make a transistor at home. The vacuum triode was a British invention, and it is not surprising that it is better understood in Britain than the American transistor. Personally I think it is a sad reflection on our nation that if you can't make one in a shed then it can't be a British invention. It does seem to me that it is the case - I digress.

There are layers of thinking in a typical IC and you have to separate them and understand them individually.

At the lowest level you have the transistors and other passive circuit elements. One layer up you have the circuit techniques, things like current mirrors, differential amplifiers and logic cells. Two layers up is functional objects like whole memories, analogue signal conditioning, analogue to digital conversion and arithmetic logic unit functions. Typically these functions are glued together and sequenced with a special kind of functional object, the state machine. Where the state machine is externally programmable, the world of software dawns.

Together, these functions can be seen as modern wealth of VLSI architectures, video codecs, baseband digital tx/rx chipsets and so on.

Just try to deal with the one transistor first. There are different types. FET's aren't very different from thermionic valves. In a similar way that the grid potential intercepts charges in the vacuum triode, biasing the gate voltage of an FET varies the availability of charge carriers in the channel of the FET, by attraction/repulsion of an electric field.

Bipolar transistors are even easier to think about if you don't care about the accuracy of your model. If a semiconductor diode is a non return valve, then a transistor is a non return valve where the flow in the pilot controls the possibility for flow in the main valve. In both cases the most obvious effect is force needed to push the spring behind the valve and lift it from the seat.

Mathematically the way to predict transistor behaviour is through a model something like Ebers-Moll. Although it is a good model, actual transistors vary wildly. The model will never predict any particular transistor very well. By looking at the maths for the model you will see why quite quickly. One parameter is very small, but in an exponential relationship tiny changes dominate the behaviour.

Very few people have a need to model individual transistors. They naturally vary so much, there is no point. As has previously been described by others, you aggregate them and bring them under control in feedback loops. In the feedback loops passive components, which can be made accurately, tame and linearise the transistor parasitic effects.

Single transistors are good for power handling, but they usually have poor gain. Small transistors consume power in their own right, but can be much faster. You can always get enough gain by cascading amplifier stages. The only way to make higher frequency circuits is to use smaller transistors. Very small transistors are hard to package individually. High frequency circuits usually are packaged into VLSI chips with other parts.

High frequency discrete transistors can be made, but they are usually made in more exotic materials like Gallium Arsenide, where higher electron mobilities can be exploited.

New fields include superconducting technologies where electromagnetism can be used to control conductivity. Indeed, in some cases new types of sensor are possible using exotic physical techniques (Yttrium barium copper oxide for SQUIDS and high temp superconductors, Camium Telluride for x-ray detectors, and Indium Phosphide for millimeter wave.)

If you need high power and high frequency, it might just be that vacuum tube technology is what you actually need. In the end the Tevatron and LHC are just enormous vacuum tubes. I'm pretty sure they are still used fairly widely in high power applications like broadcast transmitters, industrial RF heaters and nuclear research.

Not only is it the case that no-one knows it all, no-one can think about all the things they know at once.

It's just divide and conquer, and that is all.

To be honest, for people who don't want to understand all the fiddly bits, these days it's possible to do almost anything you want just by hooking modules together.

For example, why buy a lithium charger chip when you can get ten prefabricated modules with the same chip and fitted connectors for the price of two bare ICs?

Neil

I think a good book for a beginner is Electronics Pocket Book by E A Parr.

ISBN 0-408-00481-9

http://www.sciencedirect.com/science/book/9780408004817

The link will let you look at the first page of each chapter to see if it is what you need. 

Martin 

Edited By Martin Connelly on 31/07/2016 15:59:23

Neil- That last post is a good example of my problem. I assume you mean a chip to build a charger for a li. battery, then mention pre-fab modules and ICs, you are assuming we know what they are. I do luckily. Problem is if I want to build said charger I would have no idea what to look for. I suppose I look up a circuit on the net and then find somebody selling an XYZS 2255 or whatever and then solder it to a couple of wires ? On different tack- I was doing a minor repair on my van and testing with a MCVM discovered that the negative (earth) wire is coloured red. Is this a modern idea to help sell diodes ?

Hi Rod

I have the latest "Learning the Art of Electronics 2016" which is a practical lab book. From the flyleaf: It proceeds at a rapid pace but requires no prior knowledge of electronics.

My best electronic book buy ever.

It does not require the companion "Art of Electronics third edition", as it contains the necessary theory.

HTH

Dave

Edited By Dave Attwood on 31/07/2016 16:12:49

Thanks for that recommendation, Dave

**LINK**

http://www.cambridge.org/gb/academic/subjects/physics/electronics-physicists/learning-art-electronics-hands-lab-course?locale=en_GB&query=

MichaelG.

Gordon W, That seems bonkers. What make of van is it and is it original wiring? I think we should all be warned if battery RED is now Negative. Seems to go against all we have ever learned about car batteries.
Its about the same as domestic wiring. They seem to change the colours every 10 years - In April, domestic mains live will probably be brown with red spots.
BobH

Something needs to click Gordon and then it's easier to understand what you might happen to be reading. I didn't receive any help at all really from books initially and then a magazine decided to describe a transistor as a current controlled variable resistor. The control current goes into the base and the current in the collector will be some multiple of this. Beta and hfe. The emitter current will be this plus the base current. An odd sort of variable resistor because unless it blows up it doesn't care how many volts it has across it or the current that flows through it. Looked at this way this link becomes more understandable. Not a bad one actually as it mentions the problems.

**LINK**

I don't think it mentions that Vbe is always around 0.6 volts on silicone transistors just like a diode. It varies in actual values and also has a temperature variation of around 2 millivolts per degree C. It shows npn transistors. pnp have the arrow drawn the other way. It indicates which way current flows. +ve to -ve.

My son has a crap physics teacher and was having problems with simple use of capacitors and ohm law. A decent analogy is a bucket or water with a hole some where down it's side. The flow rate of the water depends on the depth of water above the hole and it's size. The depth is pressure - that's what volts are in electronic terms - the pressure that is available to cause some current to flow. Bigger the hole the lower the resistance is. Put another hole above this one and if it's the same size the flow will be reduced because the pressure is lower. When the bucket is full and no holes it has a certain capacity for providing energy. The volume sets for how long when a hole is added and the position relative to it's height and it's size sets the flow rate. As the water level drops so does the pressure there to provide the flow so that drops as well - exactly what happens when a resistor is put across as charged capacitor. If a hose pipe is used to fill the bucket it will have some flow rate so will fill at an even rate - just like charging a capacitor with a current. If it was filled by another bucket with a hole in the filling rate would change over time. Just like charging a capacitor with a resister. They don't do inductance but in a simple sense an inductor is the entire opposite of a capacitor. eg put a voltage across it and the current will rise at a linear rate and melt if something doesn't limit the current. A capacitor would in theory charge instantaneously as nothing is limiting the current. An inductor also store energy. Formulae for both will be about.

I found a pdf of the art of electronics v3. Not bad. It is all there in respect to transistors so is likely to be in all cases. I feel it's a bit jumbled though. The mag article I mentioned went through biasing slowly and easily in simple terms using their rather daft model.

People will have another problem. Why bother. Say some one wanted to design a steam engine and didn't want to just use some one else's design. They would probably spend some time looking around on the subject. They might spend hours and hours doing this and then design one. It might not work so more thought and looking around and try it again. Later they might do another based on the results of that one to improve it and so on. Compare that to wanting to understand electronics. No driving force really unless some one wants to do something specific with it and they could maybe buy a Velleman kit or follow some one else's design out of a book or a magazine.

Another similar thing want to buy and use a lathe - what do you want to make.

John

-

Edited By Ajohnw on 31/07/2016 17:05:44

So would my "Practical Wireless Circuits" by F. J. Camm be helpful?