Re: Google Groups Soon To Be No More

[...]

If that is the company I think it is, their technical manuals, handbooks, applications notes and Technical Review set the highest standards. They were a pleasure to read and so informative, so can I take the opportunity of thanking you for your work.

When I first used a computer, I took the time to construct a lot of electronics symbols and put them in a library. They were designed on the grid system which my drawing programme used, so they aligned with each other and the lines joining them. I looked into the crossover problem and decided that a gap and staggered '+' junctions removed all possibility of misunderstandings, so that was the system I adopted.

The result is that quite complex diagrams are easy to draw and understand:

while to find what it was all about...

Yes - sometimes I find a prototype board in the junk pile and can't even remember making it or what it was supposed to be.

[...]

When resistors are a box with text and capacitors are a box with text and IC's are a box with text, the diagram becomes very difficult to read quickly.

The old Dutch diagrams used a squared zigzag line as a resistor (the symbol was later reserved for non-inductive resistors). I prefer the old Dutch/German symbol for electrolytic capacitors, as it is not easy to work out which way around the British Standard symbol is supposed to be.

The drive voltage is relative to the induced voltage on the transformer primaries (the yellow squares on the leftmost windings) so the TIP31A's act as comparators with the voltage across the DC resistance of the transformer winding in the emitters of the 2N3055's and ignore the signal voltages across the inductance of the windings. They turn on enough current to give a follower action, so each triple is effectively

*current* driving its output transformer winding and the triples can all be connected in parallel.

This arrangement uses the DC resistance of the output transformer winding as its current sensing, which avoids the losses of emitter resistors. It also means that clipping the drive voltage (the red LEDs marked "current clamp") automatically limits the peak current which the triples are required to deliver.

The outer feedback loop then controls the output voltage, so it looks like a low impedance source at the output terminals. Normally, having two transformers inside a feedback loop is a bad thing because of the phase shifts, but the quadrifilar 1:1 driver transformer with its capacitors could also be regarded as two centre-tapped chokes; any HF phase shifts are well outside the audio band. The feedback is therefore mainly limited by the phase shifts in the output transformer, which are minimised by interleaving.

The offsetting works very well but I still prefer a gap to make it quite obvious whan the wires aren't touching each other. That is more difficult to draw with a pencil but is easy with a library symbol. I also like to put my transistors and valves inside an enclosure to make them show up as anchor points when reading a large diagram (...and to keep the vacuum in).

[...].

I used some integrated amplifiers to drive individual loudspeaker units in a big monitor speaker system. The data sheet said they did not need Zobel networks on the loudspeaker connections, so I didn't provide any. After a while there was the most terrible distortion and bursts of oscillation. It turned out that the bandwidth of the amplifier was only good enough for stability if you ran it on the highest possible supply rails, I had de-rated it so the phase shifts had increased and it became unstable. The second printing of the data sheet showed a Zobel network was necessary.

{...]

But they needed a high voltage supply. This drove long overhead lines from a 24v 60AH battery. An inverter would have to be very well suppressed to avoid radiating interference from the output wiring. It was designed for the utmost eceonomy so that it would run all day unattended if only occasional speech was required.

The quiescent current of a few tens of milliamps was obtained by very careful thermal design in the output triples, all the TIP31A's, including the thermal reference, were mounted on an aluminium bar, well away from the power transistors and heatsinks. The sound quality was OK, although not up to the best hi-fi standards ...but that didn't matter for outdoor P.A. operation. It was far more important that it ran all day on one set of batteries, survived sustained overloads, shorts and open-circuits on the output terminals and, if someone connected the battery the wrong way around could be used again immediately afterwards without the need to replace fuses (or whole output stages)..

[...].

Was it just an experimental demonstrator or do you actually use it in your kitchen?

On the internet. There's

Hi, George,

Good to hear from you, man!

ES and astraweb are about it since aioe went casters-up in January.

Cheers

Phil Hobbs

Hi George, yes just register a free account at eternal September, I use Mozilla thunderbird to access it on Linux and windows PCs and NewsTapLite on smartphones. Works pretty well.

On 2023-12-24 18:10, John Larkin wrote:> On Sun, 24 Dec 2023 21:34:50

-0000 (UTC), Dan Purgert snipped-for-privacy@djph.net > wrote: >

Putting them on a website (even a private one) lets you make a permanent copy on archive.org, so that stuff doesn't get lost as easily. Does that work with Dropbox?

Cheers

Phil Hobbs

;)

It’s the public links we’re discussing, unless of course you feel that the traffic level here has fallen beneath notice.

Does archive.org save them if you ask them to? I do that routinely for links that I want to see preserved (mine and others’).

Cheers

Phil Hobbs

Jim T would have agreed heartily. ;)

Cheers

Phil Hobbs

None of that is viable. Something done by a willing individual is vulnerable; it has to be done by some organization that doesn't depend on an individual and thus has a chance to perdure.

And it has to be better than the existing Usenet.