Transistors

[Mark Hennessy]

Personally, I think it's crucial to actually build some circuits and see how they behave in reality, but perhaps that's easier said than done sometimes...

It was a bit late last night to flesh this out, but here's what I had in mind:

If you don't already have one, pick up one of these: http://www.ebay.co.uk/itm/Solderless-Pro...2425853343

Then, something like this: http://www.ebay.co.uk/itm/600PCS-30-Valu...3009077609

And: http://www.ebay.co.uk/itm/200-Pcs-630V-0...3055146455

And: http://www.ebay.co.uk/itm/120Pcs-12Value...3011051073

Finally: http://www.ebay.co.uk/itm/18-Values-180p...3011069161

All those are examples rather than specific recommendations. Shop around, pick the best deals. But ultimately, you've got a complete starter kit for hardly any money - a fraction of what you'd pay from Farnell/CPC/RS et al. The quality is absolutely fine for learning.

With those, plus a power supply (just a PP3 battery, frankly) and a multimeter is all you need to get going. You can throw together the circuits we've discussed and see them work for yourself. You can alter component values and get a feel for what is going on. Seriously, you need to put the books to one side and start doing this now. You'll start to realise the harsh reality of component tolerances and why so much of what you bring here is academic in reality.

Remember the golden rule: DC conditions first, then add signal.

Signal generators and oscilloscopes are not essential at first. But you will need to pick up something sooner or later, and these need not be expensive or bulky either. See this thread (all the way through to the end) for an LCD 'scope kit for very little money indeed: http://golbornevintageradio.co.uk/forum/...p?tid=4823 - though "real" 'scopes can be found for free, and if you can accommodate their bulk, I'd always recommend that. Plenty of kits for signal sources exist from the usual suppliers, or you could build your own. Some basic test gear is the foundation of an electronics hobby, and building it yourself is the best sort of learning experience you can get.

Please, make this your focus now. I appreciate that you have a lot of other stuff going on right now, but rather than aimlessly reading crusty old text books in your spare time, research and plan setting up your mini-workshop. You will get lots of help and advice from us on here about that, especially as we will see that you're finally setting off in the right direction at long last. Indeed, you might even get offers of free or very cheap equipment! I certainly have stuff I can give away if I'm convinced it's going to a good home (which I'm not yet, but I'll be delighted to change my mind).

Actually doing stuff is a really good way to move something from your short-term memory into your long-term memory. More than once you've said how quickly and easily stuff gets forgotten, and this is why. It's not enough to read something, even repeatedly. I find it really hard to remember things unless I've linked them with experiences. Why do you think we have practical exercises in courses? When I teach electronics, at least half the time (often more) is practical time. Short bits of theory, practical, consolidation, next bit of theory to build on previous, practical, consolidation, next bit of theory, etc, etc.

Having made this post, I reserve the right to edit/delete posts that continue to cite the old books, or are otherwise off-topic

[Nowhere-Man]

"I appreciate that you have a lot of other stuff going on right now, but rather than aimlessly reading crusty old text books in your spare time, research and plan setting up your mini-workshop. "

You could say my boat is effectively finished. It is also vintage and rescued. The other boat will therefore be up for sale but I've yet to determine where I will store the radios. Probably I'll build a shed for that. Since March most days were occupied by manual work and radio work relegated to nocturnal periods. Or just when walking the dog( in my head). So for storage of tools and my radio gear a shed is now on the agenda and one of my boats will be on brokerage.
After this I have 2 projects to finish - 2 valve radios to repair and that VEF. It's a race against Winter. I tend to cease the boat work when the weather turns cold and nasty. So practical radio work is scheduled for maybe October. Sale of one of my boats could also smooth things over.

[Nowhere-Man]

A few points: I have one or two stubborn quirks with regard to radio engineering. For one, I dislike modern boards. I once built a rectifier and started with the idea of using breadboard but disliked it. I reverted to point soldering and so far the 50's radio still works fine. The other day I dismantled a modern radio with a view to salvage a few parts and was very unimpressed by the board. To put it one way, one of my receivers is dated 1947 and, after a service, it still pulls in lots of signals. Far more robust than any modern circuit.
I agree about the need to build stuff and to obtain signal generator and so forth. However I never use Ebay so it might be a good idea to visit a vintage radio rally. By the way, has anyone heard of a "spotter"? These were crystal oscillators that generated a 100 kilocycle harmonic and were supposed to zero beat a 10 MHz signal, and thereafter produce a sound spotter at every 100 kilocycles. This is the closest I got to a signal generator.
Last point. I don't mind using silicon transistors but dislike the fiddly, frail specimens used in modern appliances. I did quite like those Soviet Top Hat affairs as the size was relatively chunky. I love GE transistors but why are they no longer produced?

[ppppenguin]

I've not heard them called "spotters" but this kind of spot frequency generator was commonplace before the days of synthesised tuning. They provided a grid of accuarate frequencies to calibrate other equipment.

Why not Ge transistors? Basically because they're pretty hopeless. They are leaky, highly temperature sensitive, useless at high temperatures and generally don't work very well. Silicon wins hands down on everything except saturation voltage and possibly performance at extremely low temperatures. The latter is only of interest below perhaps -50C. Saturation voltage issues have been dealt with by using MOSFETS which can have extremely low on-resistance and no saturation voltage effects.

What if makers had put lots of development into Ge? I don't know, but Ge remains a relatively rare and expensive element while Si is abundant and cheap.

There are a few specialised and exotic Ge devices still made. They use a mixture of Ge and Si in the same device.

There's nothing fiddly or frail about Si devices such as BC548 in TO92 plastic packages. Dirt cheap, work well in loads of circuits.

As soon as you get much beyond very simple circuits, such as those found in domestic radios and TVs, hand wiring is a nightmare. Agreed some PCBs are badly designed or cheaply made but there's no sensible alternative once you get to a certain complexity.

[Refugee]

The other point about GE transistors is that they were all made individually and not in wafers or sheets of up to 100 at a time like silicon transistors are.

[ppppenguin]

I'm sure that if they'd put in the developmnet effort they would have found some mass production process for Ge transistors. Something akin to the planar process that revolutionised Si transistor production and paved the way for ICs.

[Nick]

I'm with Mark on this. Get some bits and play with them. A couple of cheap dmms help too - avos are not ideal as they present significant loading.

[ppppenguin]

And if you destroy a few BC548 class transistors in experiments who cares. They cost a few pennies each. Even from RS you can get 100 for a fiver or so.

[Craig]

Transistors are as cheap as chips (ha ha) now. Looking at PW in 1970 a typical transistor (like a silicon BC109 or a germanium like an OC71) were about five bob each - which is about £3.80 each today. I'm pretty sure that is why the Quad 33 preamp, released in 1968, only had six (BC109) transistors per channel.

Compare that to Jeffrey's BC548 at 4p each in 200 off (so £8 for 200). An effective price reduction of 100 fold as compared with 1970.

[Terry]

If you don't like breadboards, how about Veroboard - stripboard with copper strips and holes punched at 0.1" intervals, now available from a number of manufacturers.



The strips can be cut to provide breaks using a spot face cutter - essentially  a drill bit mounted in a handle so you just twist it to remove the copper. You can design your own layouts or there is even a free CAD program available these days.

For circuits using transistors and discrete components, the electronic schematic can be transferred to the board quite easily with the actual components in the same position as the schematic.

When I used it a lot, I found the most economical way was to buy it in long strips and cut it up as required.

Try entering Veroboard into a search engine - select images and you will find lots of circuit layouts and actual built up boards and ranging from simple to extremely complex, so you can see that it is a very flexible tool to have at your disposal.