ВЭФ 12 (Astrad VEF 12 Project) - including a transistor tutorial

[pwdrive]

At the risk of getting plonked, can we assume that the base current can be ignored?

Lawrence.

[Mark Hennessy]

That's one of the questions I was hoping that NM would ask. His questions are essential for us to help him - they will tell us where he's at.

NM, your feedback is essential - don't be shy!

Don't try to do it all at once - a path is formed by laying one stone at a time.

And we don't need excess precision. To the nearest 0.1 volt is fine. Understanding the process is more important than the maths.

[Terry]

I'm certainly behind the comments about ignoring physics and, in the main, text books.

As far as the holes in transistors are concerned, I've been happily ignoring those for over half a century!

Of course, I grew up firmly entrenched in the valve era. If there was ever a design using transistors in Practical Wireless during my schooldays, I can't remember it. Indeed, when they did start appearing a few years later, they promptly moved to a new magazine, Practical Electronics!

Thinking back to my schoolboy studies, I probable learnt the type of metal used for every valve electrode, the chemicals required to produce the emissive material for the cathode and so on. When you think about it, what I really was learning was how to build my own valves! Of course, I lacked the specialised equipment needed and the skills to operate them - not to mention a degree in glass blowing! - so, apart from any casual interest, there wasn't one single phrase in anything I read that was of any practical use whatsoever!

Then, of course, we moved onto transistors, growing 100% pure germanium and silicon, then poisoning it to create P and N types and so on, before finally putting them together in various ways to create PNP and NPN devices. It's strange, but the only physical aspect of transistor construction that had any practical application was never discussed in all these texts, causing much consternation to engineers when presented with battery operated transistor radios for repair that suffered from mains hum!

This, of course, was only a problem with glass encapsulated transistors if the paint got scratched allowing artificial light to fall on the junction - and was quickly resolved by the manufacturers filling the inside of the glass container with an opaque silicon grease.

Photo electric transistors were available, of course - at a high price - but these were presented as having been specially made to be photo sensitive when, in fact, there was just an ordinary transistor inside the clear glass encapsulation which had been modified to include a lens!

One thing that has not been mentioned here that is important to learn is which way up transistors work. No valve can conduct unless its anode is positive with respect to its cathode so there is no such thing as a valve that works 'upside down' (in an electronic, rather than physical sense).

Transistors, however, are different. An NPN transistor mimics a valve in that it conducts with it's collector positive with respect to its emitter whereas a PNP type conducts with it's collector negative with respect to its emitter. To anybody brought up in the valve era and attempting to grasp this new technology it was confusing as all the early transistors were PNP types and the battery positive terminal was usually connected to the chassis or earth line. Even the sight of all the electrolytics which appeared upside down compared to 'normal' added to the strangeness of the new circuits.

Analogue multi-meters don't 'do' negative, unlike their modern DMM cousins which just stick a negative sign in front of the number so you even had to reverse the test leads to take any measurements!

Still, we all survived the transition ...

[Nowhere-Man]

I had a busy night as ever. I don't always get much free time but I did just look at the diagram. I get a potential divider across the 47 K and 10 K of approximately 7.421052632 and 1.578947368 which should equal the source of 9 volts. The BC109 is NPN. However I will recheck it later to be sure.

[Nowhere-Man]

As it's NPN I can't really comment on the common connection and signal imput vs audio output as so far I chose to stick with PNP. PNP covers most 1960's radios so my idea was to stay with valves but just do the very basics on germanium.

[Mark Hennessy]

OK, that's an excellent start

First of all, let's not say "approximately", and then go on to give an answer with 9 numbers after the decimal point. We need to be realistic here: we are engineers who deal with real-world components and real-world measurements. The resistors might only be within 10% of their stated values, and a basic DVM might only be 1% accurate. I know from experience that some people are uncomfortable initially when adjusting to "engineering" - especially those who have done a lot pure maths or physics. However, they usually get the idea, and find that the world is still turning.

My approach is to always guess at what the answer should be, then reach for the calculator to give me the precision. The advantage of that is simple: it's easy to make a "typo" with the calculator. I spend most of my working life trying to convince graduates and school leavers to do this - after a life of learning by rote to pass exams, it doesn't come easily. However, it does make their lives better...

So here, I'd say that if the 47k was actually 50k, then I'd have a sixth of 9V across the 10k resistor (1.5V). But as the upper resistor is a little lower, that voltage will be a little higher in practice.

Let's take your results and record them with our "engineering" hats on: approximately, we have 7.4V and 1.6V

You've discovered that there is 1.6V across the 10k resistor. Another way of putting this: the base voltage is 1.6V with respect to 0V (or ground). That's the first step done.

We're left with the emitter voltage and the collector voltage. I'll give you this one for free: we can't do the collector just yet. So we must do the emitter first.

Now, if you've read up a bit on transistors already, you might know that it's actually really easy to work out the emitter voltage, given that we now know the base voltage. The sum is literally one you can do in your head. However, you might not have gleaned the vital bit of information yet. If not, just ask

Finally, just an aside: please, please don't worry about NPN vs PNP transistors. They work in exactly the same way. The difference is the polarity of the battery (and any polarised components like electrolytic capacitors). And as for silicon vs germanium, again, they work in exactly the same way. Yes, there are some detail differences - the most obvious one has a bearing on the above question so I won't say more (but that's perhaps a useful hint?) - but let's come back to that later.

So really, transistors are easier than valves. I think so, at least. Ignoring JFETs and MOS-FETs (which you won't find in a 1960s radio) there's only 1 basic type to learn. Compare that to valves, where you have triodes, tetrodes, pentode, et al.

Yes, the BC109 is NPN. You probably know that the direction of the arrow tells you that, as will the datasheet. I picked it because it's used by Hacker in their 1960s radios (it's one of my favourites )

[Nowhere-Man]

I am far from being a maths exponent or physics student. Having said that, the science aspect does interest me a lot. I aren't yet sure to what degree I want to be an engineer. Probably I'll be happy with a reasonable level of practical application. I don't agree with you that transistors are easier because the valves are just free electrons. No other atoms involved like antimony and boron. I'll read your postings and ask questions over time. One thing I'd appreciate is do you think I can test the goodness of a transistor without removal? Bias maybe? I was getting only 5 volts on a collector so do you think for an AF transistor that's suspect?
I'll address any other points later if I get free time.

[Mark Hennessy]

The inner working of transistors and valves for this are best left to one side for now. We're focussing on the practical application, so you can understand the basic circuits that use transistors, and that's what I mean when I say that transistors seem easier than valves to me.

Testing transistors?

When you understand roughly how they work in a circuit, then yes, you can "test" them in circuit. But not with a tester. Rather, by examining the behaviour of the circuit and deciding if the transistor is doing the right thing. To comment further on your specific question, I'd want to know what the supply voltage was. Ideally, a copy of the circuit would be best. But soon, you'll be able to judge for yourself.

In the meantime, you usually need to remove the transistor. If you haven't already got one, the Chinese component testers do a pretty good job. This is the one I bought: http://www.ebay.co.uk/itm/381351034073 - but note that this one requires assembly. Plenty of others ready-built at the same price point. Note that they can only do so much, so occasionally they catch you out.

To help you move to the next stage of analysing this circuit, I'll tell you that the emitter voltage is lower than the base voltage. By how much? Well, this is a number that is pretty much fixed - it's different for germanium and silicon, but pretty consistent for each type. It is a number that you encounter whenever you have a semiconductor PN junction. Remember (or discover) that number, and subtract it from 1.6V. It's that easy!

You're closer than you think

[Nowhere-Man]

Well, the so-called potential hill is 0.3 so that must be surpassed. However I have a magic number 0.98 that gives the "current transfer ratio" Subtract it from 1 to get 0.02 (ratio digit) and then divide that same digit "into" 0.98 to get 49 which would be the current gain (only in grounded emitter circuits.
I hope I'm not being too boring. I am ordering some transistors I may need to do repairs so the above useless information just keeps me occupied. The VEF circuit I posted further back but it's a bit blurred.

[Mark Hennessy]

0.3 would be the right number for germanium. However, this is silicon. So try again

I'm not sure about the "magic number" sentence - I won't think about it any more right now, but perhaps we'll come back to it. I'm trying to keep you focussed on the problem - nothing else.

Let me give you a bit more: if that transistor was germanium (perhaps an AC127), can you see how the 0.3V would result in 1.3V at the emitter?