The Zobel correction - UK Vintage Radio Repair and Restoration Discussion Forum

Cruisin Marine · 26th Mar 2021, 12:46 am

Zobel networks, used on OB feeds etc.
Most interesting, the seasonal ground temperature changes are worth noting
https://en.wikipedia.org/wiki/Zobel_...ormer_roll-off

Radio Wrangler · 26th Mar 2021, 7:56 am

That's a rather good wikipedia article and gives a good overview of Zobel's amplitude equaliser work. If you want to see an article on practical implementation with design curves and formulae for component values, Look up Harry Kimball's article in, I think it was the joutnal of the society of motion picture and television engineers, 1938. This paper is reprinted in Giacoletto's reference book (successor to Landee, Davis and Albrecht).

This sort of equaliser is not just used for flattening the response of cables, they also sometimes get used to correct the flatness of things like filters and transducers.

There is a separate class of equaliser used for correction of phase responses, sometimes called all-pass networks.

But the equalisers were only part of Zobel's work. He was also deeply involved in early filter theory, which is in a separate wikipedia article. He invented the image-impedance trick which is used in a lot of network synthesis which turned into what were called m-derived filter designs. If you look at books on filter design before 1967, you'll see it was mostly m-derived and constant-k which were the buzz-words. 1967 was when Zverev was published, and he collected together network analysis and synthesis techniques and moved it on into what he called 'Modern filter synthesis' So if you look in books and articles on filters (Electric Wave Filters, to give their full name) then you'll see that Anatol Zverev's monster book is the number one reference.

Another luminary in the filter landscape was Seymour B Cohn.

But Zobel covered so much of network theory that it's faintly hilarious when people speak of "The Zobel network" as if there was just one specific type. Cohn has suffered a similar fate.

David

Vintage_RC · 26th Mar 2021, 12:31 pm

Before the days of PC's and filter design software I used Anatol Zverev's monster book regularly, I still have a copy of the book on my bookshelf although it is a long time since I opened it.

Radio Wrangler · 26th Mar 2021, 3:55 pm

Most filter design software just works on the most basic mainstream filter topologies.

A firm I worked for recently hired a small contracting firm to design a proof-of-concept board for a receiver. They used off the web filter calculators and just put in the centre frequencies and bandwidths, and out came component values. The software must have been really basic and didn't know of the need to go to coupled resonator realisations sometimes. So it came back with ludicrous L and C values. Infinitesimal capacitors working with inductors large enough that the self resonant frequency was going to ruin things. They didn't know better and laid out their board.

Filters use some rather imperfect components and are very sensitive to strays. The filter they generated looked OK on Spce, but had a couple of nodes needing zero stray C to ground. 0.05pf added to the model there, ruined them.

On the whole, you have to keep an eye on filter design.

Zverev and a few minutes with a pocket calculator will do nicely. if you know the pitfalls.

David

Cruisin Marine · 26th Mar 2021, 4:25 pm

I still use my Chris Bowick book RF Circuit Design and a calc when designing RF filters, never fails me.

Back to that article...
I love the way the standards back in the day were so high, with people in the control room, adjusting the EQ for perfection, or as close to they could get it.
In recent years, Talksport has a loud buzz on their carrier for at least ages without doing anything about it (Shoreham site), despite it being reported.
How can you be that neglectful?

Jez1234 · 5th Apr 2021, 8:46 pm

The Zverev technique for tuning bandpass filters is a form of practical witchcraft! That's genius.

Radio Wrangler · 5th Apr 2021, 10:41 pm

Some people still haven't worked out why some of my filters have places for shorting links

Any sufficiently advanced technology is indistinguishable from magic. (and vice-versa)

David

Jez1234 · 6th Apr 2021, 12:31 am

Seeing it work, "just like it says on the tin", on the SA screen, on a filter I'd designed for a 2m receiver (output of LO multiplier chain) was one of the most joyous and grin inducing moments in decades of electronics for me I must admit.

26th Mar 2021, 12:46 am	#1
Cruisin Marine Heptode Join Date: Jul 2018 Location: Worthing, West Sussex, UK. Posts: 992	The Zobel correction Zobel networks, used on OB feeds etc. Most interesting, the seasonal ground temperature changes are worth noting https://en.wikipedia.org/wiki/Zobel_...ormer_roll-off

26th Mar 2021, 7:56 am	#2
Radio Wrangler Moderator Join Date: Mar 2012 Location: Fife, Scotland, UK. Posts: 22,868	Re: The Zobel correction That's a rather good wikipedia article and gives a good overview of Zobel's amplitude equaliser work. If you want to see an article on practical implementation with design curves and formulae for component values, Look up Harry Kimball's article in, I think it was the joutnal of the society of motion picture and television engineers, 1938. This paper is reprinted in Giacoletto's reference book (successor to Landee, Davis and Albrecht). This sort of equaliser is not just used for flattening the response of cables, they also sometimes get used to correct the flatness of things like filters and transducers. There is a separate class of equaliser used for correction of phase responses, sometimes called all-pass networks. But the equalisers were only part of Zobel's work. He was also deeply involved in early filter theory, which is in a separate wikipedia article. He invented the image-impedance trick which is used in a lot of network synthesis which turned into what were called m-derived filter designs. If you look at books on filter design before 1967, you'll see it was mostly m-derived and constant-k which were the buzz-words. 1967 was when Zverev was published, and he collected together network analysis and synthesis techniques and moved it on into what he called 'Modern filter synthesis' So if you look in books and articles on filters (Electric Wave Filters, to give their full name) then you'll see that Anatol Zverev's monster book is the number one reference. Another luminary in the filter landscape was Seymour B Cohn. But Zobel covered so much of network theory that it's faintly hilarious when people speak of "The Zobel network" as if there was just one specific type. Cohn has suffered a similar fate. David __________________ Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done

26th Mar 2021, 12:31 pm	#3
Vintage_RC Heptode Join Date: Sep 2018 Location: Twickenham, London, UK. Posts: 538	Re: The Zobel correction Before the days of PC's and filter design software I used Anatol Zverev's monster book regularly, I still have a copy of the book on my bookshelf although it is a long time since I opened it. __________________ Alan G6PUB, BVWS

26th Mar 2021, 3:55 pm	#4
Radio Wrangler Moderator Join Date: Mar 2012 Location: Fife, Scotland, UK. Posts: 22,868	Re: The Zobel correction Most filter design software just works on the most basic mainstream filter topologies. A firm I worked for recently hired a small contracting firm to design a proof-of-concept board for a receiver. They used off the web filter calculators and just put in the centre frequencies and bandwidths, and out came component values. The software must have been really basic and didn't know of the need to go to coupled resonator realisations sometimes. So it came back with ludicrous L and C values. Infinitesimal capacitors working with inductors large enough that the self resonant frequency was going to ruin things. They didn't know better and laid out their board. Filters use some rather imperfect components and are very sensitive to strays. The filter they generated looked OK on Spce, but had a couple of nodes needing zero stray C to ground. 0.05pf added to the model there, ruined them. On the whole, you have to keep an eye on filter design. Zverev and a few minutes with a pocket calculator will do nicely. if you know the pitfalls. David __________________ Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done

26th Mar 2021, 4:25 pm	#5
Cruisin Marine Heptode Join Date: Jul 2018 Location: Worthing, West Sussex, UK. Posts: 992	Re: The Zobel correction I still use my Chris Bowick book RF Circuit Design and a calc when designing RF filters, never fails me. Back to that article... I love the way the standards back in the day were so high, with people in the control room, adjusting the EQ for perfection, or as close to they could get it. In recent years, Talksport has a loud buzz on their carrier for at least ages without doing anything about it (Shoreham site), despite it being reported. How can you be that neglectful?

5th Apr 2021, 8:46 pm	#6
Jez1234 Banned Join Date: Mar 2021 Location: Morpeth, Northumberland, UK. Posts: 936	Re: The Zobel correction The Zverev technique for tuning bandpass filters is a form of practical witchcraft! That's genius.

5th Apr 2021, 10:41 pm	#7
Radio Wrangler Moderator Join Date: Mar 2012 Location: Fife, Scotland, UK. Posts: 22,868	Re: The Zobel correction Some people still haven't worked out why some of my filters have places for shorting links Any sufficiently advanced technology is indistinguishable from magic. (and vice-versa) David __________________ Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done

6th Apr 2021, 12:31 am	#8
Jez1234 Banned Join Date: Mar 2021 Location: Morpeth, Northumberland, UK. Posts: 936	Re: The Zobel correction Seeing it work, "just like it says on the tin", on the SA screen, on a filter I'd designed for a 2m receiver (output of LO multiplier chain) was one of the most joyous and grin inducing moments in decades of electronics for me I must admit.