Transistors

[Nowhere-Man]

People frequently disagree in electronics and at the highest level.
I simplify semiconductors a bit. First I picture the PNP junctions. Each P layer will have a slightly negative potential hill. Clearly the electrons will fill a few holes only at the junction, from N layer to the P layer. So bias will either widen the negavive voltage at the P layer or dismantle it. For emitter we need to dismantle the PD so a + bias will suffice. For collector we enhance the existing - field so need battery -. That creates a big resistance. As to valves I think the speed is 0.001 of a megacycle from cathode to anode but would need to look it up. Once it strays into physics it goes a bit beyond vintage radio although it's all educational.

[Craig]

For a good and non-physics guide to transistor operation have a look at https://www.st-andrews.ac.uk/~jcgl/Scots...p/comp.htm and click on bipolar transistor at the bottom. Fermi is not mentioned ;-)

[Mark Hennessy]

It's a shame that I'm away for a few days with only a phone - otherwise I'd be joining in. In short, I usually cite Douglas Self's views on this, and he in turn cites some heavyweight sources... The transistor is voltage operated, and the base current is an unfortunate side-effect. Sometimes it's appropriate to use the current-operation model; sometimes the voltage-operational model is best. The latter is the true MO, however. That takes some swallowing, granted

[pwdrive]

NM wrote: "As to valves I think the speed is 0.001 of a megacycle from cathode to anode but would need to look it up."

One nanosecond (typical):

https://books.google.co.uk/books?id=41vU...nd&f=false

Lawrence.

[Nowhere-Man]

As I said before, I am bogged down making my fibreglass boat fit to live in so I try to squeeze in radio theory at night. Later on, my goal wll be to fix that Asrad radio I was working on. I found the schematic was all a bit unfamiliar so far. I'm used to triode heptode mixers and double diodes as per 1950's. So, during my "out-of-term", I hope to fill in the gaps, as it were.
The boat work all boils down to pumps, pipes, breakers and woodwork. One good thing is the transistor sets won't be blocked by steel walls.

[Pat Pending]

In the first instance thermionic devices were developed from observations, for example the plate that was placed inside an early electric lamps to stop the glass bulb from becoming blackened by emissions from the filaments, was eventually developed into the Fleming diode. The principles of the technology was understood before the applications were developed.

Transistors were envisioned, and their potential advantages were the driving force in the race for their early development, it's a little like building a shed then drawing the plans afterwards. So the transistor was based on some broad notions about the physical properties of semiconductors. If the team at Bell hadn't "got lucky" the French scientists who were developing their own version would have, or someone else.

The point of this ramble is that because of the way that the transistor came into being, the early texts that explained how the device worked, are for the most part hogwash, and only really serve to overcomplicate and confuse.

No self respecting electronics designer would slavishly use the vagaries of transistor gain parameters or the flawed notion of the current model to design their electronic creations. It's a good thing that once exam time is over, most of those academic theories on the solid state, that are printed as facts, but have little value In the real world, can be forgotten.

[Nowhere-Man]

For the most part I agree with you. It is evolution. Recently I was stunned to discover the IF in radio came about due to regenerative tuning in homes. Hank Hill would be adjusting his feedback but he'd cause oscillation. Lo and behold, Dale next door would suddenly get the same program but on next doors frequency. So, it was clear frequencies could be changed. The 455 KHz came by trial and error. Only after came the texts to explain how it worked.
Once transistors allowed for PCB work, I guess that spelled the end of engineered radios that could be serviced. You could keep refining the boards and just factory solder a whole range. Like painting by numbers. At some point for me the magic was lost. I totally dislike modern transistors. I prefer the large germanium ones with coloured leads. I lve those sixties car radios too with the strung dials.

[Terry]

Don't forget that 455kHz is the US IF. In Europe it should be 470kHz (originally 465kHz).

MW radios primarily designed for the US market have always suffered from a problem in the UK when listening to transmissions on 909kHz - Radio 5 Live but originally the BBC Home Service on 908kHz - because of a beat between the incoming signal and the second harmonic of the IF.

[Murphyv310]

Correct me if I'm wrong but I would doubt that IF frequencies were worked out by trial and error. Early superhet radios had lower IF frequencies often around 120-170 kHz, this was done as the valves of the time performed better at lower frequencies. Later on with a higher frequency IF the valves had improved and it was also simpler and cheaper to wind the higher frequency coils. Also don't forget that higher quality and communication sets will have two frequency changers and dual IF, not to be confused with AM & FM sets.

[ppppenguin]

High IF gives better image rejection. Low IF gives more gain with older valves/transistors. It's a tradeoff. Then you choose your IF to minimise spurious effects such as 2nd harmonic beating with wanted stations.

Professional receivers often convert to a first IF above the highest frequency they can receive. Gets rid of image and other spurious problems. Also allows continous coverage of frequencies which wouldn't be possible if the IF coincided with a wanted frequency. The upconverter is sometimes followed by a 2nd conversion to a lower IF.