Class B w/o crossover distortion (1975)

[hahfran]

The April 1975 issue of Wireless World had a few pages on a then novel topology Class B output without crossover distortion. I had such an about 50 watts amp done it works until today ( I use it for burn-in of loudspeakers by a 24 h driving at 16 hz ) . The circuit does not require adjustment of idling current or dc offset. It has excellent stability of parameters versus operating temperature. Now I ponder to have this using contemporary transistors as a power amp for active speakers. But the copy of the Wireless World is lost.
Electronics World doesn't keep it in the archives , either. Any idea how to get an April 1975 copy of Wireless World?

[Bigun]

That sounds interesting, I'd love to see it too !

p.s. It's not one of those Ian Hegglun articles on square law outputs ? - do you know the author ?

[gootee]

Do a search at google.com for:

"Class B" without "crossover distortion" April 1975 Wireless World

I haven't gone through all of the search results, yet. But it looks like there are some results that might have what you want.

For example, it looks like the PDF linked below at least mentions the article. The short description includes "A novel class B output, Wireless World, April 1975", but it's still downloading, here, and I haven't looked at it yet.

http://www.hi-fi.ro/fhifi/download.p...c2138bc9aa6729

Tom

[gootee]

And here's a Google Books result that mentions the same article:

High perfomance audio power ... - Google Books

If that's the right article ("A Novel Class B Output"), the google books result shows the author as "Visch, Nico.M".

Maybe a better search string for google.com would be:

"Class B" "crossover distortion" April 1975 "Wireless World"

Tom

[Key]

Is it this? http://220.232.128.43/audio/quad405/...J%20Walker.pdf

[hahfran]

Originally Posted by gootee

Do a search at google.com for:

"Class B" without "crossover distortion" April 1975 Wireless World

I haven't gone through all of the search results, yet. But it looks like there are some results that might have what you want.

For example, it looks like the PDF linked below at least mentions the article. The short description includes "A novel class B output, Wireless World, April 1975", but it's still downloading, here, and I haven't looked at it yet.

http://www.hi-fi.ro/fhifi/download.p...c2138bc9aa6729

Tom

Ok. That provides a hint.
Thanks.
I have recovered the schematics from the only document that made it through all those years, the PCB layout, however that has no component values except that the power BJTs are 2N3055. The circuit that prevents reverse-biasing is the current source T9 and the circuit T7,T8. Thus the NFB
hasn't to handle large voltage steps. I tested the crossover distortion with a dual channel scope in Channel I - Channel II mode ( output - inverted input, different sensitivity) which isn't accurate. However below 30 watts in 4 ohms, no crossover distortion is seen. That is not bad.
It could be interesting to combine this class B in a current dumping topology.

Here is the schematic

[AKSA]

This might help a bit, with values.

Hugh

[hahfran]

Ah..getting closer to the source. Or not...frankly I don't get it how it works, how it is class B, and what is the large 250 capacitor for? It is apparent why and how this topology provides an almost junction-temp independent setting.
This explains why it still works after more than 30 years. But not how it prevents crossover distortion...

[Wavebourn]

The schemo contains at least 2 errors: emitter of CCS shorted on emitter of output transistor, and 250uF Miller compensation cap instead of 250 pF one.

[AKSA]

Hahfran,

If you do the simulation you see that neither of the input transistors (with emitters joined) ever turns off. Thus, remaining biased on, there is a relatively simple step function (1.3V) between input and output at all points on the waveform.

Their collectors, one on the top triplet, the other on the bottom triplet, will then behave in a way which always maintains dc 1.3V between input and output, and this implies that their collector voltage will always control the triplets to ensure that the transfer function of the output triplets is neatly pasted over at the crossover point.

I'm not sure Anatoliy is correct; the 250uF is a bootstrap, not a miller cap (no gain).

Clever circuit, huh?

Hugh