[Radio Wrangler]

One friend is a radio amateur who's built a number of different HF transceivers by picking bits of circuits from all over the place. Building a multiband transceiver is quite an achievement, and he has built a whole series. Some got published, some he put out as kits when he started a small firm with a friend. However, he avoided PLLs like the plague having had problems with them and no successes. His radios used a bank of crystal oscillators (one per band) mixed with the VFO and filtered, as the local oscillator. He consequently ran ito difficulties of getting people with sparse test equipment to align bandpass filters.

I suggested the PLL approach as a lot easier to adjust and could be set up with just a voltmeter. He described his experiences. He'd built PLLs with oscillators from one source, dividers from another, PSD and loop amps from others. He hadn't realised how bespoke the timeconstants and gain of the loop amp needed to be.

So the next time we met (almost 500 miles each way) I took him through the poles and zeroes business and showed him a recipe approach to PLL loop design.We built one. He made a board for a full all-HF-band PLL summing loop and it worked first time. We had a bit of fun trying to get it to misbehave, and proved to ourselves that it not only worked, it was robust.

I went back North, and he wrote a construction article for Radcom. It was seen as a bit heavy for PW. It won him the RSGB Talbot-Wortley trophy for the best technical article that year. A successful circuit and a gong isn't bad, but I think the biggest return for his effort was entry to a further level in the design game. and the slaying of a dragon.

Thevenin...

This site is sprinkled with warnings about high voltage power supplies and their dangers. But what about high currents? Remember those car batteries with big lead connections on the top joining the cells? People have been severely injured by metal watch straps touching across the four volts between adjacent connections. In the worst cases hands have had to be amputated because the damage was so severe. The current welded the strap in place and the rest is gruesome.

Voltage and current are sort of mirror images of each other. Not the same, but with many similarities. Most of our designs see the signal as a voltage which gets processed. It could equally well have been a current. Good understanding of a circuit sees the signal as a voltage in some places and a current in others.

Thevenin's theorem is one of a pair. A mathematician would see them as duals. There is an exact symmetry between the pair. The other theorem is Norton's. With our bias towards thinking of voltage first, Norton's seems much weirder.

Back to dangerous power supplies. Normal flyback TV EHT is seen as dangerous, but mains derived EHT is seen as VERY dangerous. THe difference being that the latter can easily supply fatal levels of current before the voltage falls much.

However, you can easily make much higher voltages stroking the cat on a dry day, or combing your hair. These processes are routinely survived, provided the cat is the domesticated variety. What saves you is the very large source impedance involved. It causes the voltage to plummet when only a tiny current is taken.

Another manifestation of source impedance is in the good old dry battery. As they aged and the active materials inside were used up, the zinc was still zincy and the carbon was still just as carbony, so their off-load voltage stayed up until things had really gone, but before the off-load voltage fell, the effective series resistance went up. Anything that wanted a bit of current met a falling voltage.

Thevenin's theorem says that any real world voltage source can be modelled by an ideal voltage source, along with a series resistance. AS FAR AS THE THINGS AFTER IT CAN SEE.
What Thevenin doesn't do is model what's going on inside the supply. So a Thevenin equivalent circuit will look like the real thing as far as voltage off load and voltage droop as you load it are concerned. But don't think about the efficiency of the supply or the power dropped in the equivalent series resistor - these may not be representative at all.

This is 'Black Box Theory' and we aren't talking PYE record players. Thevenin's perfect voltage generator and resistor live in a black box. What goes on inside is opaque and needs to be to stop the non-realistic power dissipation misleading the viewer in some cases.

So all Thevenin is, is a way of formally describing how some voltage source droops when you take current.

Norton does exactly the same thing, but his black box contains a perfect current source shunted by a resistor. It looks weirder, but it's convenient sometimes.

David