[Iain Churches]

I have brought this discussion with Joyhn Byrns
over from another thread, as it has the makings
of what might prove to be aninteresting project.



"John Byrns" wrote in message
...
In article ,
"Iain Churches" wrote:


John wrote

The LF and HF ranges of your proposed +/- 3 dB equalizer are obvious
enough, but it is not so obvious what you are talking about for the MF
range?

Before the days of parametric EQ, many recording consoles had three
band equalisers on most channels. The mid band was often termed
"presence" (particularly on broadcast consoles) and was centred at 3kHz
(the BBC used 2.6kHz IIRC)


The BBC analogue meter is no longer in production, though I do know
a source from which they can probably be obtained NOS.

I did not mean to imply and actual BBC meter should be used, I assumed a
peak reading meter circuit designed specifically for this project, or a
clone of the BBC circuit if desired. I assume that analog meter
movements are still manufactured, or am I wrong about this?

The BBC type is pretty funky. Two movements on one meter face,
green needle pointer for left, red needle pointer for right. White
lettering on a black background. Few people outside broadcast
have seen them. They are always a talking point.

See:
http://www.kolumbus.fi/iain.churches/Pics/BBC_PPM.jpg

What do you mean by the "Peter Baxandall topology"? When I think of a
"Peter Baxandall topology" I always think of his feedback tone control
circuit, but today many people, at least here on this usenet group, seem
to consider the "Baxandall" tone control circuit to be the standard
passive tone control circuit.

It is a long time since I read the Baxandall articles. IIRC the first
circuit was a feedback tone control as you describe. The passive
version arrived soon afterwards. The circuit was later modified for
use with guitar amps in the USA, and known by the name of James.

I would like to see some evidence of who
actually developed the common passive tone control circuit, it is not
obvious to me that it was "Peter Baxandall" was responsible for it as
many today claim, although it is possible, I don't know and would like
to see some evidence to settle the issue once and for all.

Peter Baxandall was the author of the articles which AFAIK first
described this tone control in a British journal. All the textbooks I
have seen attribute it to him also.

Unfortunately I don't think the passive topology, whoever may have
designed it, easily accommodates the addition of a "MF" band.

Agreed. I was thinking of Baxandall for a two frequency
control LF and HF ( I would probably never use the mid control)
but if one needs mid, then there are plenty of other options.


I have been giving this idea quite a bit of thought. There are a lot of
alternatives, but I have picked one as a starting point, although it is
difficult to proceed very far without having a better idea what you want
the "MF" equalizer to do? Could you provide some input on the
specifications desired for the "MF" equalizer? Once I have had a chance
to think a bit about the requirements for the "MF" equalizer, I will
write up my thoughts and post them here.

As mentiond before, I think it should be fairy wide Q and centred on
3kHz Personally, I could live without it.

Let me do some listening tests with a parametric set up as a three band
(I have some Neve modules) and I will see if I can find a compromise.
I am sure there is no "one size fits all" solution.

Best regards
Iain

[Mike Gilmour]

"Iain Churches" wrote in message
. fi...
[clip]

The BBC analogue meter is no longer in production, though I do know
a source from which they can probably be obtained NOS.

I did not mean to imply and actual BBC meter should be used, I assumed a
peak reading meter circuit designed specifically for this project, or a
clone of the BBC circuit if desired. I assume that analog meter
movements are still manufactured, or am I wrong about this?

The BBC type is pretty funky. Two movements on one meter face,
green needle pointer for left, red needle pointer for right. White
lettering on a black background. Few people outside broadcast
have seen them. They are always a talking point.

See:
http://www.kolumbus.fi/iain.churches/Pics/BBC_PPM.jpg


I beg to differ Iain, where twin PPM needle meters are used the needles
colour coded Red and Green (A & B is BBC parlance) for Left / Right
metering, so it the opposite way around to hi-fi. Also you may see
White/Orange for Sum/Difference metering. i.e. For broadcast purposes Green
is Right and Red is Left, it certainly was so at the last FM station I was
presenting at. As far as I'm aware the BBC are exactly the same. You have to
think a little after working on hi-fi where red=right then working on a
broadcast desk where red = left. I was part of a team refurbishing an
on-air studio and I can certainly confirm the left channel balanced xlr's
were marked red.
I haven't tried googling but I'm certain the BBC will be the same.

Regards,

Mike

[Iain Churches]

"Mike Gilmour" wrote in message
...

I beg to differ Iain, where twin PPM needle meters are used the needles
colour coded Red and Green (A & B is BBC parlance) for Left / Right
metering, so it the opposite way around to hi-fi. Also you may see
White/Orange for Sum/Difference metering. i.e. For broadcast purposes
Green is Right and Red is Left, it certainly was so at the last FM station
I was presenting at. As far as I'm aware the BBC are exactly the same. You
have to think a little after working on hi-fi where red=right then working
on a broadcast desk where red = left. I was part of a team refurbishing
an on-air studio and I can certainly confirm the left channel balanced
xlr's were marked red.
I haven't tried googling but I'm certain the BBC will be the same.


Morning Mike,

Yes, I think you are right. I had forgotten that the BBC
convention was not the standard convention..

These meters were also widely used outside the BBC.
One saw them quite often on analogue studio consoles
where they had been fitted later to replace VU meters.
Often two of these PPMs were fitted to show send
levels to a pair of stereo mastering machines, and
sometimes the right-hand meter could be configured
as a return to show the output from
one or other of the two machines.

I wonder if these meters are still available. Any idea?
IIRC Canford Audio used to be the supplier.

Regards
Iain

[Iain Churches]

"Iain Churches" wrote in message
. fi...
I have brought this discussion with Joyhn Byrns
over from another thread, as it has the makings
of what might prove to be aninteresting project.


"John Byrns" wrote in message
...
In article ,
Or not, I held off on replying to see if anyone else jumped in to express
interest in the project, Patrick was the only one to comment.


It may be because he (Patrick) is the only one with
experience in building tone control circuits. For many
they are absolutely taboo. But there is no reason why
we can't discuss them, surely?


I have taken the time to start thinking about three different design
approaches for this equalizer. The first approach, and probably the
simplest is to use the Baxandall feedback tone control topology that
Patrick suggested. This seems like the simplest approach because it
doesn't require a gain makeup stage like the passive approach does. The
problem with the Baxandall feedback approach is that adding an MF
equalizer seems a bit of a kludge to me.

I seriously wonder about the need for an MF control. Without it
the FB Baxendall is fairly easy to implement.

A better feedback approach, which more easily accommodates an MF equalizer
would be to use a differential amplifier as the feedback stage. This is
the basis of my magnum opus design which also uses a gyrator to create a
virtual inductor for the MF equalizer. Nine triodes are required for each
channel with this design, 2 for an ALCF input buffer, 2 for the gyrator, 3
for the differential amplifier feedback equalizer stage, and 2 for an ALCF
output buffer.

So eighteen triodes for a stereo equaliser???

The third approach, which is my favorite, is a passive network with a gain
makeup stage, although I am not sure how to build a gain makeup stage with
such low gain.

First thoughts would be a mu follower at the front end.

I have an old Wireless World article by P.J. Baxandall which describes his
feedback tone control, I don't know if this was his original disclosure of
the circuit or not. I don't have any article by Baxandall that describes
the passive approach, if anyone has an early article by Baxandall on the
passive tone control circuit I would be interested in a copy.

I would like to see some evidence of who
actually developed the common passive tone control circuit, it is not
obvious to me that it was "Peter Baxandall" was responsible for it as
many today claim, although it is possible, I don't know and would like
to see some evidence to settle the issue once and for all.

Peter Baxandall was the author of the articles which AFAIK first
described this tone control in a British journal. All the textbooks I
have seen attribute it to him also.

I would be interested in seeing those articles and textbook references.
Baxandall's name has always been associated with the feedback tone control
circuit in articles and textbooks, but I never saw the passive circuit
attributed to him until audio groups appeared on the internet. It just
has too much of the feel of an Urban Legend to me without some serious
evidence to the contrary. Do you know what issue of Wireless World
Baxandall's article describing the standard passive tone control circuit
appeared in? I have a fair collection of Wireless World from the relevant
period and might be able to find it.

A friend of mine in the UK has a huge collection of early audio
publications, and will almost surely know about Peter Baxendall.
He sometimes lurks on RAT so it may well be that
he will make contact by e-mail.


As mentiond before, I think it should be fairy wide Q and centred on
3kHz Personally, I could live without it.

Broad makes it easier, without an MF range of some sort the design is
trivial

Let me do some listening tests with a parametric set up as a three band
(I have some Neve modules) and I will see if I can find a compromise.
I am sure there is no "one size fits all" solution.

Of course not, but with a limit of only 3, three position toggle switches
one must force a single size to fit all, even if some pushing and shoving
is required.

I did some listening last night to material ranging from Baroque (ensemble
plus voices) to small group jazz to Jethro Tull.. In almost every case, the
material sounded better with no EQ at all! But, if I had to choose three
bands, they would be 100Hz, 2.8kHz, and 10kHz (or possibly 8kHz
depending on the material)

Regards
Iain

[Iain Churches]

"John Byrns" wrote in message
...

My post was the result of a comment I made about a matching equalizer
in
a response to Iain's original preamp post. I didn't expect a positive
response from Iain, and was surprised at his marginally positive
response, although I did considered it might have been a joke on Iain's
part.

Now why should I joke about a thing like that, John? :-)

Yes, I did notice your interest, but there is no missing Patrick's point,
the mention of equalizers in this group always draws intense silence. I
suspect people believe they will be somehow tainted by even discussing
this forbidden item, much less actually using one in their system.

Perhaps you have contributed to this phobia of equalisers by
mentioning a design calling for eighteen triodes.

As regards the amount of lift and cut available in each
band, I personally believe that 6dB is excessive.

I agree that 6dB correction on any single element within
a mix is often called for, and I do it daily. But contemplate
what happens when you try to add even 3dB of mid to
a stereo mix just to add a little presence to say a vocal.
The flutes start to sound like clarinets, the clarinets
start to sound like oboes, and the oboes start to sound like........
Canada geese! A mid band control used across a stereo
track is the fastest way to destroy the timbre of any
recording.

Regards
Iain

[Mike Gilmour]

"Iain Churches" wrote in message
i...

"Mike Gilmour" wrote in message
...

I beg to differ Iain, where twin PPM needle meters are used the needles
colour coded Red and Green (A & B is BBC parlance) for Left / Right
metering, so it the opposite way around to hi-fi. Also you may see
White/Orange for Sum/Difference metering. i.e. For broadcast purposes
Green is Right and Red is Left, it certainly was so at the last FM
station I was presenting at. As far as I'm aware the BBC are exactly the
same. You have to think a little after working on hi-fi where red=right
then working on a broadcast desk where red = left. I was part of a team
refurbishing an on-air studio and I can certainly confirm the left
channel balanced xlr's were marked red.
I haven't tried googling but I'm certain the BBC will be the same.


Morning Mike,

Yes, I think you are right. I had forgotten that the BBC
convention was not the standard convention..

These meters were also widely used outside the BBC.
One saw them quite often on analogue studio consoles
where they had been fitted later to replace VU meters.
Often two of these PPMs were fitted to show send
levels to a pair of stereo mastering machines, and
sometimes the right-hand meter could be configured
as a return to show the output from
one or other of the two machines.

I wonder if these meters are still available. Any idea?
IIRC Canford Audio used to be the supplier.

Regards
Iain


Good Morning Iain,

Yes the meters are still available. They are manufactured by Sifam which is
part of Digitron Instruments.
See http://www.sifam.com/meters_home.lasso , they can still be obtained from
Canford but you may well have a closer distributor.
Also if you go to their home page http://www.sifam.com (save you fingerwork
:-) you'll also see they do a fine selection of small control knobs as well.

Regards,

Mike

[John Byrns]

In article ,
"Iain Churches" wrote:

"Iain Churches" wrote in message
. fi...
I have brought this discussion with Joyhn Byrns
over from another thread, as it has the makings
of what might prove to be aninteresting project.

"John Byrns" wrote in message
...
In article ,
Or not, I held off on replying to see if anyone else jumped in to express
interest in the project, Patrick was the only one to comment.

It may be because he (Patrick) is the only one with
experience in building tone control circuits. For many
they are absolutely taboo. But there is no reason why
we can't discuss them, surely?

I have taken the time to start thinking about three different design
approaches for this equalizer. The first approach, and probably the
simplest is to use the Baxandall feedback tone control topology that
Patrick suggested. This seems like the simplest approach because it
doesn't require a gain makeup stage like the passive approach does. The
problem with the Baxandall feedback approach is that adding an MF
equalizer seems a bit of a kludge to me.

I seriously wonder about the need for an MF control. Without it
the FB Baxendall is fairly easy to implement.

IIRC you were the one that first brought MF equalization into the
discussion so I am left to wonder why the turnabout? While I find MF
equalization useful I wouldn't have included it in a home brew project
of the sort I am talking about.

A better feedback approach, which more easily accommodates an MF equalizer
would be to use a differential amplifier as the feedback stage. This is
the basis of my magnum opus design which also uses a gyrator to create a
virtual inductor for the MF equalizer. Nine triodes are required for each
channel with this design, 2 for an ALCF input buffer, 2 for the gyrator, 3
for the differential amplifier feedback equalizer stage, and 2 for an ALCF
output buffer.

So eighteen triodes for a stereo equaliser???

Yes, even Patrick's Baxendall approach requires twelve triodes with
input and output buffers. The six additional triodes in my design are
to provide for the MF equalizer.

The third approach, which is my favorite, is a passive network with a gain
makeup stage, although I am not sure how to build a gain makeup stage with
such low gain.

First thoughts would be a mu follower at the front end.

I am leaning towards a passive network with an approximate loss of 6 dB,
how do you build a good linear gain stage with a voltage gain of only
2X? The problem is further complicated if the gain makeup stage does
double duty as the input buffer where it will have to provide twice the
voltage swing that is required of the output buffer. As I was typing
this paragraph I think I have found the germ of an idea to solve the
problem, I will comment further once I have had a chance to think it
through a little further.

I did some listening last night to material ranging from Baroque (ensemble
plus voices) to small group jazz to Jethro Tull.. In almost every case, the
material sounded better with no EQ at all! But, if I had to choose three
bands, they would be 100Hz, 2.8kHz, and 10kHz (or possibly 8kHz
depending on the material)

Thanks for that input, my current thoughts are leaning towards LF & HF
equalization only, switchable -6 dB, -4 dB, -2 dB, 0 dB, +2 dB, +4 dB,
+6 dB, with LF band switchable between 100 Hz and 200 Hz, and HF band
switchable between 4 kHz and 8 kHz.


Regards,

John Byrns

--
Surf my web pages at, http://fmamradios.com/

[John Byrns]

In article ,
"Iain Churches" wrote:

"John Byrns" wrote in message
...

Yes, I did notice your interest, but there is no missing Patrick's point,
the mention of equalizers in this group always draws intense silence. I
suspect people believe they will be somehow tainted by even discussing
this forbidden item, much less actually using one in their system.

Perhaps you have contributed to this phobia of equalisers by
mentioning a design calling for eighteen triodes.

An equalizer is not only a functional thing but needs be able to impress
visitors with its sheer visual excess. Also consider if you add
equivalent buffering to Patrick's design you will end up with twelve
triodes, and you won't have the MF equalizer which is what the other six
triodes provide.

As regards the amount of lift and cut available in each
band, I personally believe that 6dB is excessive.

I agree that 6dB correction on any single element within
a mix is often called for, and I do it daily. But contemplate
what happens when you try to add even 3dB of mid to
a stereo mix just to add a little presence to say a vocal.
The flutes start to sound like clarinets,

As a failed clarinet player I like it when the flutes sound like
clarinets.

the clarinets
start to sound like oboes, and the oboes start to sound like........
Canada geese! A mid band control used across a stereo
track is the fastest way to destroy the timbre of any
recording.

And since on any given day we have somewhere between a dozen and two
dozen Canada geese hanging about in the back garden I expect everything
to sound like Canada geese.


Regards,

John Byrns

--
Surf my web pages at, http://fmamradios.com/

[Iain Churches]

"John Byrns" wrote in message
...

An equalizer is not only a functional thing but needs be able to impress
visitors with its sheer visual excess. Also consider if you add
equivalent buffering to Patrick's design you will end up with twelve
triodes, and you won't have the MF equalizer which is what the other six
triodes provide.

Well. It'snot actually terribly functional, it seems to me. I have
perhaps half a dozen outboard EQ units, whuich we sometimes
take on location jobs. I haven't used one at home for years.
In fact I don't even use a preamp- CD player straight to the
1st grid via a stepped attentuator.

As regards the amount of lift and cut available in each
band, I personally believe that 6dB is excessive.

I agree that 6dB correction on any single element within
a mix is often called for, and I do it daily. But contemplate
what happens when you try to add even 3dB of mid to
a stereo mix just to add a little presence to say a vocal.
The flutes start to sound like clarinets,

As a failed clarinet player I like it when the flutes sound like
clarinets.

:-))


the clarinets
start to sound like oboes, and the oboes start to sound like........
Canada geese! A mid band control used across a stereo
track is the fastest way to destroy the timbre of any
recording.

And since on any given day we have somewhere between a dozen and two
dozen Canada geese hanging about in the back garden I expect everything
to sound like Canada geese.

OK. Here's a challenge for you.....Build an EQ than can turn a
Canada goose back into an oboe:-)

Best regards
Iain

[Iain Churches]

"Mike Gilmour" wrote in message
...


Good Morning Iain,

Yes the meters are still available. They are manufactured by Sifam which
is part of Digitron Instruments.
See http://www.sifam.com/meters_home.lasso , they can still be obtained
from Canford but you may well have a closer distributor.
Also if you go to their home page http://www.sifam.com (save you
fingerwork :-) you'll also see they do a fine selection of small control
knobs as well.


Hello Mike. Thanks for the Sifam link. I am glad to know these meters
(and lots of other good 'uns from the look of the Sifam site) are still
available.

I have just acquired a Radford LDO4 (low distortion oscillator)
something I have been looking for a very long time. It has an
excellent meter by Sifam, marked Radford.

http://www.kolumbus.fi/iain.churches...adfordLDO4.jpg

This unit came from the old CBS studios in Whitfield St London.

Best regards
Iain

[Mike Gilmour]

"Iain Churches" wrote in message
.fi...

"Mike Gilmour" wrote in message
...


Good Morning Iain,

Yes the meters are still available. They are manufactured by Sifam which
is part of Digitron Instruments.
See http://www.sifam.com/meters_home.lasso , they can still be obtained
from Canford but you may well have a closer distributor.
Also if you go to their home page http://www.sifam.com (save you
fingerwork :-) you'll also see they do a fine selection of small control
knobs as well.


Hello Mike. Thanks for the Sifam link. I am glad to know these meters
(and lots of other good 'uns from the look of the Sifam site) are still
available.

I have just acquired a Radford LDO4 (low distortion oscillator)
something I have been looking for a very long time. It has an
excellent meter by Sifam, marked Radford.

http://www.kolumbus.fi/iain.churches...adfordLDO4.jpg

This unit came from the old CBS studios in Whitfield St London.

Best regards
Iain




Excellent, yet more armourery for amplifier testing. The Sifam meters are
well built to survive years of H24 service. I note that Sifam have a
distributor in Finnland.

Best regards,

Mike

[Iain Churches]

"Mike Gilmour" wrote in message
...

"Iain Churches" wrote in message
.fi...

"Mike Gilmour" wrote in message
...


I have just acquired a Radford LDO4 (low distortion oscillator)
something I have been looking for a very long time. It has an
excellent meter by Sifam, marked Radford.

http://www.kolumbus.fi/iain.churches...adfordLDO4.jpg

This unit came from the old CBS studios in Whitfield St London.

Excellent, yet more armourery for amplifier testing.

Hi Mike

Bench oscillators seem vary considerably in their performance. Very
few are really "low distortion". Even the modern ones from RS
and Farnell see to be about 0.1% to 0.4%. Before I found the
Radford, the best I had was the Ferrograph RTS at 0.02%

It is interesting to compare their THD and also distortion profile.

Patrick gave an excellent tip to use a third octave filter
on the output of the generator. This cleaned up the waveform
considerably.

I read a Norman Crowhurst article where he recommended
that the distortion of the test generator should be an order
of magnitude below the lowest THD to be measured on
an amp under test.

The Radford is -100dB (0.001%) at 1V and so more than
meets the criterion.

Best regards
Iain

[Iain Churches]

"John Byrns" wrote in message
...
In article ,
"Iain Churches" wrote:


IIRC you were the one that first brought MF equalization into the
discussion so I am left to wonder why the turnabout? While I find MF
equalization useful I wouldn't have included it in a home brew project
of the sort I am talking about.

Not really a turnabout. This was one of the answers to the "what would
you like to see in a (fill in the blank) unit" which I asked of several
people.

I am leaning towards a passive network with an approximate loss of 6 dB,
how do you build a good linear gain stage with a voltage gain of only
2X? The problem is further complicated if the gain makeup stage does
double duty as the input buffer where it will have to provide twice the
voltage swing that is required of the output buffer. As I was typing
this paragraph I think I have found the germ of an idea to solve the
problem, I will comment further once I have had a chance to think it
through a little further.

I have a feeling that the passive Baxendall has a loss of 20dB.


I did some listening last night to material ranging from Baroque
(ensemble
plus voices) to small group jazz to Jethro Tull.. In almost every case,
the
material sounded better with no EQ at all! But, if I had to choose
three
bands, they would be 100Hz, 2.8kHz, and 10kHz (or possibly 8kHz
depending on the material)

Thanks for that input, my current thoughts are leaning towards LF & HF
equalization only, switchable -6 dB, -4 dB, -2 dB, 0 dB, +2 dB, +4 dB,
+6 dB, with LF band switchable between 100 Hz and 200 Hz, and HF band
switchable between 4 kHz and 8 kHz.

Yes. That could be done nicely with rotary switches - but I still fancy my
toggle idea +3dB, 0dB, -3dB. Don't you think that 4-8kHz is a little low
for the HF. I would have thought 8-10kHz to put a little zizzle on the
cymbals.

I look forward to reading your further thoughts.

Meanwhile, did you look at the BBC ppm? What did you think of
the idea of the dual movement? It can also be switched M+S.

Best regards

Iain

[John Byrns]

In article ,
"Iain Churches" wrote:

"John Byrns" wrote in message
...

I am leaning towards a passive network with an approximate loss of 6 dB,
how do you build a good linear gain stage with a voltage gain of only
2X? The problem is further complicated if the gain makeup stage does
double duty as the input buffer where it will have to provide twice the
voltage swing that is required of the output buffer. As I was typing
this paragraph I think I have found the germ of an idea to solve the
problem, I will comment further once I have had a chance to think it
through a little further.

I have a feeling that the passive Baxendall has a loss of 20dB.

That would get down to what the definition of a "passive Baxendall" is?
However as per our previous discussions I have yet to see any evidence
that Baxendall developed the common passive tone control circuit. What
I call the common passive Hi-Fi tone control circuit does typically have
a loss of 20 dB as you say. The 20 dB loss matches well with the
standard "audio" taper pot to provide approximately flat response when
the two controls are centered. However the basic topology can have any
loss desired, basically the amount of maximum boost that can be obtained
is approximately the same as the flat loss of the passive network. A 6
dB loss provides for a less aggressive tone control, and can be
implemented with linear taper pots if you like to use pots. I prefer
using switches however because you can get a nice crisp shelving effect
at all settings.

I did some listening last night to material ranging from Baroque
(ensemble
plus voices) to small group jazz to Jethro Tull.. In almost every case,
the
material sounded better with no EQ at all! But, if I had to choose
three
bands, they would be 100Hz, 2.8kHz, and 10kHz (or possibly 8kHz
depending on the material)

Thanks for that input, my current thoughts are leaning towards LF & HF
equalization only, switchable -6 dB, -4 dB, -2 dB, 0 dB, +2 dB, +4 dB,
+6 dB, with LF band switchable between 100 Hz and 200 Hz, and HF band
switchable between 4 kHz and 8 kHz.

Yes. That could be done nicely with rotary switches - but I still fancy my
toggle idea +3dB, 0dB, -3dB. Don't you think that 4-8kHz is a little low
for the HF. I would have thought 8-10kHz to put a little zizzle on the
cymbals.

That's what the 8 kHz setting is for, or make it 5 kHz and 10 kHz if 8
kHz is too low for your taste. Maybe you missed the fact that this
proposal uses four switches, two switches to determine the amount of low
and high frequency boost/cut, and a second pair of switches to determine
the frequencies at which the boost/cut starts to become effective.

I look forward to reading your further thoughts.

Meanwhile, did you look at the BBC ppm? What did you think of
the idea of the dual movement? It can also be switched M+S.

Yes I like it, sort of a better way of doing the two sideways meters
thing with the tips of the pointers close together. It is sort of
physically large though, isn't it? Also it just occurred to me as I am
typing this, does the meter at the link also include the actual "PPM"
electronics or do you have to provide that yourself? I don't remember
seeing any specs that might relate to this such as power requirements
for the circuitry, which makes me wonder.


Regards,

John Byrns

--
Surf my web pages at, http://fmamradios.com/

[John Byrns]

In article ,
"Iain Churches" wrote:

"John Byrns" wrote in message
...
In article ,
"Iain Churches" wrote:

I did some listening last night to material ranging from Baroque
(ensemble plus voices) to small group jazz to Jethro Tull..
In almost every case, the material sounded better with no EQ at all!
But, if I had to choose three bands, they would be 100Hz, 2.8kHz,
and 10kHz (or possibly 8kHz depending on the material)

Thanks for that input, my current thoughts are leaning towards LF & HF
equalization only, switchable -6 dB, -4 dB, -2 dB, 0 dB, +2 dB, +4 dB,
+6 dB, with LF band switchable between 100 Hz and 200 Hz, and HF band
switchable between 4 kHz and 8 kHz.

Yes. That could be done nicely with rotary switches - but I still fancy my
toggle idea +3dB, 0dB, -3dB. Don't you think that 4-8kHz is a little low
for the HF. I would have thought 8-10kHz to put a little zizzle on the
cymbals.

I have a few further thoughts, but first I am curious why you want to
use toggles, is it a technical, or a styling issue? I went back to your
web pages to look at the picture of the 203A to see how toggles fit in
with its styling theme, but all I got was an error message "Haettua
sivua ei löydy!"

After some thought I have come up with a proposal that combines the best
features of both your equalizer concept and of mine. I will describe it
first in terms of rotary switches, although it could also be done with
toggles if the needed contact arrangements are available. I will
describe the idea in terms of the HF boost function, but the HF cut as
well as the LF functions would work similarly. The HF switch would have
three boost positions, the first would provide a 3 dB shelf starting at
10 kHz as you suggested. The second HF boost position would retain the
3 dB shelf, but would lower the frequency where the boost starts to 5
kHz. The third and final HF boost position would retain the 5 kHz
starting frequency, but increase the shelf to 6 dB. What do you think
of this idea?

A pair of three position toggles could also be used to implement a
similar scheme. One way would be to have the first toggle operate as
you propose and provide +3dB, 0dB, -3dB at 10 kHz, the second toggle
could provide +3dB, 0dB, -3dB at 5 kHz, and when both toggles were
activated a total boost of 6 dB would be provided, either by allowing
the two curves to simply sum together, or if toggles with suitable
contact arrangements are available a third optimized network could be
switched into the circuit. This arrangement with two three position
toggles would also allow other possibly useful combinations. For
example the 5 kHz toggle could be set to boost with the 10 kHz toggle
set to cut, which would provide a small bump in the 5 kHz to 10 kHz
region. I also think that the four toggles would give the equalizer box
a more interesting look than would only two toggles.

When you speak of frequencies with respect to shelving equalizers, what
are you speaking of? When I speak of a frequency in this regard I am
speaking of the frequency of the zero in the response of a boost, and
the pole in the response of a cut. I am wondering if that is also what
you mean, or if you have something else in mind?

With respect to circuit topology, as I said before I was leaning towards
a passive network, if for no other reason than that I long ago worked
out the design equations that allow picking the required component
values for any desired response. I decided that I would use a cathode
follower to provide a fixed low impedance drive to the network, and
after some thought I decided to use a triode with shunt feedback
following the equalizer network to make up for the loss of the network
as well as to provide a reasonably constant load on the network. After
doing a little work on the design of this stage I realized that I may
not be able to get it to work out exactly as I had hoped. As a result I
may decide to use a feedback style equalizer network similar to the
Baxendall, unfortunately I don't have the design equations for this in
my notebook, so my first task will be to work out the feedback equalizer
design equations, then proceed from that point.


Regards,

John Byrns

--
Surf my web pages at, http://fmamradios.com/

[robert casey]

Iain Churches wrote:

"John Byrns" wrote in message
...


My post was the result of a comment I made about a matching equalizer
in
a response to Iain's original preamp post. I didn't expect a positive
response from Iain, and was surprised at his marginally positive
response, although I did considered it might have been a joke on Iain's
part.


Now why should I joke about a thing like that, John? :-)


Yes, I did notice your interest, but there is no missing Patrick's point,
the mention of equalizers in this group always draws intense silence. I
suspect people believe they will be somehow tainted by even discussing
this forbidden item, much less actually using one in their system.


Perhaps you have contributed to this phobia of equalisers by
mentioning a design calling for eighteen triodes.


For signal level audio handling, FETs or regular transistors could be
used, although that would be heresy to mention any sand in this
newsgroup. :-) Better than op-amps! Though some op-amps can be
improved by forcing their outputs into "single ended mode" by sourcing
or drawing an ma or so of current with a resistor to a supply rail like
+ or - 20V. To avoid crossover distortions that the large amounts of
feedback work hard to but don't completely eliminate. Is it 18 tubes
for a stereo equalizer? That would usually mean 9 twin triode tubes.

[Patrick Turner]

robert casey wrote:

Iain Churches wrote:

"John Byrns" wrote in message
...


My post was the result of a comment I made about a matching equalizer
in
a response to Iain's original preamp post. I didn't expect a positive
response from Iain, and was surprised at his marginally positive
response, although I did considered it might have been a joke on Iain's
part.


Now why should I joke about a thing like that, John? :-)


Yes, I did notice your interest, but there is no missing Patrick's point,
the mention of equalizers in this group always draws intense silence. I
suspect people believe they will be somehow tainted by even discussing
this forbidden item, much less actually using one in their system.


Perhaps you have contributed to this phobia of equalisers by
mentioning a design calling for eighteen triodes.


For signal level audio handling, FETs or regular transistors could be
used, although that would be heresy to mention any sand in this
newsgroup. :-) Better than op-amps! Though some op-amps can be
improved by forcing their outputs into "single ended mode" by sourcing
or drawing an ma or so of current with a resistor to a supply rail like
+ or - 20V. To avoid crossover distortions that the large amounts of
feedback work hard to but don't completely eliminate. Is it 18 tubes
for a stereo equalizer? That would usually mean 9 twin triode tubes.

Class A opamp chips are around with utterly blameless measurements
compared to a typical old 741.

Discrete circuitry with individual bjts or j-fets requires huge amounts
of NFB to get the same low THD/IMD as triodes without NFB.

I have never comcluded that discrete SS parts sound better than opamps.
The old days of 1972 where many makers did still use discretes are long
gone
because opamps became cheap and better measuring than anything you could
do with discretes.
Typically, a discrete gain stage would have two transistors in a
darlington pair,
and another pair as a follower output, or the two pairs used
as Ziclai pair. Lots of local current and voltage FB is involved.
Biasing is often awkward,
and the design time is high, so opamps are the choice for many
engineers.

I still prefer triodes and discretes, and SE circuitry.
But the distortion is rarely ever lower than 0.01%, and
opamps get 0.001%

Patrick Turner.

[Iain Churches]

"John Byrns" wrote in message
...

I have a few further thoughts, but first I am curious why you want to
use toggles, is it a technical, or a styling issue?

A little of each. I don't think anyone needs, or indeed should be
offered 20dB of EQ. And for style reasons, three knobs spread
across a 200mm front panel looks decidedly odd. (I am taking
the word of my discriminating panel of evaluators here)

I went back to your
web pages to look at the picture of the 203A to see how toggles fit in
with its styling theme, but all I got was an error message "Haettua
sivua ei löydy!"´

You should have got a message in three languages, Finnish,
English and Swedish:

Haettua sivua ei löydy!
Cannot find the page!
Den sökta sidan kan ej hittas!


I took the page down because the unit is to be professionally
photographed later this week.

After some thought I have come up with a proposal that combines the best
features of both your equalizer concept and of mine. I will describe it
first in terms of rotary switches, although it could also be done with
toggles if the needed contact arrangements are available. I will
describe the idea in terms of the HF boost function, but the HF cut as
well as the LF functions would work similarly. The HF switch would have
three boost positions, the first would provide a 3 dB shelf starting at
10 kHz as you suggested. The second HF boost position would retain the
3 dB shelf, but would lower the frequency where the boost starts to 5
kHz. The third and final HF boost position would retain the 5 kHz
starting frequency, but increase the shelf to 6 dB. What do you think
of this idea?

So that would be:
3dB shelf at 10kHz (that's good)
3dB shelf at 5kHz (that quite good)
6dB shelf at 5kHz (that pretty horrific IMO)

I had an e-mail from a German recording engineer earlier today
with comments on this thread. He said " Do you think that such
people buy an oil-painting and ask for a brush and some paints
so they can "correct" the colours in a Holbein?"

He makes a good point!


A pair of three position toggles could also be used to implement a
similar scheme. One way would be to have the first toggle operate as
you propose and provide +3dB, 0dB, -3dB at 10 kHz, the second toggle
could provide +3dB, 0dB, -3dB at 5 kHz, and when both toggles were
activated a total boost of 6 dB would be provided, either by allowing
the two curves to simply sum together, or if toggles with suitable
contact arrangements are available a third optimized network could be
switched into the circuit. This arrangement with two three position
toggles would also allow other possibly useful combinations. For
example the 5 kHz toggle could be set to boost with the 10 kHz toggle
set to cut, which would provide a small bump in the 5 kHz to 10 kHz
region. I also think that the four toggles would give the equalizer box
a more interesting look than would only two toggles.

This idea is much more interesting, and also much more versatile.
It sounds like something Peter Walker might have thought of.

When you speak of frequencies with respect to shelving equalizers, what
are you speaking of? When I speak of a frequency in this regard I am
speaking of the frequency of the zero in the response of a boost, and
the pole in the response of a cut. I am wondering if that is also what
you mean, or if you have something else in mind?

That's how I visualise it also. Is there any other interpretation?

With respect to circuit topology, as I said before I was leaning towards
a passive network, if for no other reason than that I long ago worked
out the design equations that allow picking the required component
values for any desired response. I decided that I would use a cathode
follower to provide a fixed low impedance drive to the network, and
after some thought I decided to use a triode with shunt feedback
following the equalizer network to make up for the loss of the network
as well as to provide a reasonably constant load on the network. After
doing a little work on the design of this stage I realized that I may
not be able to get it to work out exactly as I had hoped. As a result I
may decide to use a feedback style equalizer network similar to the
Baxendall, unfortunately I don't have the design equations for this in
my notebook, so my first task will be to work out the feedback equalizer
design equations, then proceed from that point.

I am intending to dig out some circuits of both the Baxendall feedback
and passive types when I can. At the moment, my time is taken up by
this 24 track editing job, and also by planning and progress chasing
the pre-production of the preamp project.

Best regards
Iain

[John Byrns]

In article ,
"Iain Churches" wrote:

"John Byrns" wrote in message
...

I have a few further thoughts, but first I am curious why you want to
use toggles, is it a technical, or a styling issue?

A little of each. I don't think anyone needs, or indeed should be
offered 20dB of EQ. And for style reasons, three knobs spread
across a 200mm front panel looks decidedly odd. (I am taking
the word of my discriminating panel of evaluators here)

It's not clear to me how the rotary switch vs. toggle switch question
relates how much EQ is offered, either approach could offer a maximum of
20 dB, 3 dB, or another number of your choice? As regards the style
issue, since the MF equalizer seems to have been unanimously dropped,
there would only be two knobs in any case. None the less I'm not sure I
agree with your "discriminating panel of evaluators" with regard to the
appearance of three knobs on a panel. No matter what the number of
knobs is the spacing between knobs as well as the spacing between the
knobs and the edges of the panel is exceedingly critical if the panel is
not to look decidedly odd, but if done right any reasonable number of
knobs can look good.

I went back to your
web pages to look at the picture of the 203A to see how toggles fit in
with its styling theme, but all I got was an error message "Haettua
sivua ei löydy!"´

You should have got a message in three languages, Finnish,
English and Swedish:

Haettua sivua ei löydy!
Cannot find the page!
Den sökta sidan kan ej hittas!

I took the page down because the unit is to be professionally
photographed later this week.

I thought the photographs were already professionally done.

After some thought I have come up with a proposal that combines the best
features of both your equalizer concept and of mine. I will describe it
first in terms of rotary switches, although it could also be done with
toggles if the needed contact arrangements are available. I will
describe the idea in terms of the HF boost function, but the HF cut as
well as the LF functions would work similarly. The HF switch would have
three boost positions, the first would provide a 3 dB shelf starting at
10 kHz as you suggested. The second HF boost position would retain the
3 dB shelf, but would lower the frequency where the boost starts to 5
kHz. The third and final HF boost position would retain the 5 kHz
starting frequency, but increase the shelf to 6 dB. What do you think
of this idea?

So that would be:
3dB shelf at 10kHz (that's good)
3dB shelf at 5kHz (that quite good)
6dB shelf at 5kHz (that pretty horrific IMO)

Another idea for multiple toggle switches would be to restrict the
maximum boost/cut to 3 dB in all cases and allow for three different
frequencies where the boost/cut becomes effective. This would also
simplify implementation if the third frequency which naturally follows
from the first two was acceptable. For example, if the HF equalizer had
a 10 kHz toggle and a 5 kHz toggle, if both were activated the frequency
would naturally become 3.3 kHz, but still limited to a 3 dB boost/cut.
So we would have:

3dB shelf at 10kHz (that's good)
3dB shelf at 5kHz (that quite good)
3dB shelf at 3.3kHz (that ? )

I had an e-mail from a German recording engineer earlier today
with comments on this thread. He said " Do you think that such
people buy an oil-painting and ask for a brush and some paints
so they can "correct" the colours in a Holbein?"

That's exactly as I would expect, it's the party line from those in the
recording industry. I have known some in the art/painting business and
while they wouldn't "ask for a brush and some paints so they could
"correct" the colours in a Holbein?", when lighting a painting I have
seen them take considerable pains with the number and positioning of
lights, as well as the colour temperature of the lights, and there are
no messy paints to get on your frock, this is not much different than
audio tone controls in concept.

He makes a good point!

That depends on your perspective.

A pair of three position toggles could also be used to implement a
similar scheme. One way would be to have the first toggle operate as
you propose and provide +3dB, 0dB, -3dB at 10 kHz, the second toggle
could provide +3dB, 0dB, -3dB at 5 kHz, and when both toggles were
activated a total boost of 6 dB would be provided, either by allowing
the two curves to simply sum together, or if toggles with suitable
contact arrangements are available a third optimized network could be
switched into the circuit. This arrangement with two three position
toggles would also allow other possibly useful combinations. For
example the 5 kHz toggle could be set to boost with the 10 kHz toggle
set to cut, which would provide a small bump in the 5 kHz to 10 kHz
region. I also think that the four toggles would give the equalizer box
a more interesting look than would only two toggles.

This idea is much more interesting, and also much more versatile.
It sounds like something Peter Walker might have thought of.

See the proposal above for a way of using two toggles to select between
three different frequencies with a maximum boost/cut of 3 dB, and still
allowing the combination of boost on one switch with cut on the other.


Regards,

John Byrns

--
Surf my web pages at, http://fmamradios.com/

[Iain Churches]

"John Byrns" wrote in message
...
In article ,
"Iain Churches" wrote:

"John Byrns" wrote in message
...

I have a few further thoughts, but first I am curious why you want to
use toggles, is it a technical, or a styling issue?

A little of each. I don't think anyone needs, or indeed should be
offered 20dB of EQ. And for style reasons, three knobs spread
across a 200mm front panel looks decidedly odd. (I am taking
the word of my discriminating panel of evaluators here)

It's not clear to me how the rotary switch vs. toggle switch question
relates how much EQ is offered, either approach could offer a maximum of
20 dB, 3 dB, or another number of your choice? As regards the style
issue, since the MF equalizer seems to have been unanimously dropped,
there would only be two knobs in any case.

A rotary switch could be up to 24 positions, but a toggle can have only
three: centre off, up= +x dB, down= -xdB. So a single step of 3dB
makes a lot more sense than a single step of say 6dB.


None the less I'm not sure I
agree with your "discriminating panel of evaluators" with regard to the
appearance of three knobs on a panel. No matter what the number of
knobs is the spacing between knobs as well as the spacing between the
knobs and the edges of the panel is exceedingly critical if the panel is
not to look decidedly odd, but if done right any reasonable number of
knobs can look good.

It's a simple exercise to mock up a panel (better than a drawing) in the
various combinations. Some look OK, others look decidedly odd, others
are pleasing.

If you are considering building something for use by others and not just
for yourself, I feel it is essential to ask for opinions and also look very
carefully at the solutions adopted by others.

I went back to your
web pages to look at the picture of the 203A to see how toggles fit in
with its styling theme, but all I got was an error message "Haettua
sivua ei löydy!"´

I thought the photographs were already professionally done.

Thanks:-) The unit was standing on a graduated background draped
across my bench and taped to two mic stands:-)

So that would be:
3dB shelf at 10kHz (that's good)
3dB shelf at 5kHz (that quite good)
6dB shelf at 5kHz (that pretty horrific IMO)

Another idea for multiple toggle switches would be to restrict the
maximum boost/cut to 3 dB in all cases and allow for three different
frequencies where the boost/cut becomes effective. This would also
simplify implementation if the third frequency which naturally follows
from the first two was acceptable. For example, if the HF equalizer had
a 10 kHz toggle and a 5 kHz toggle, if both were activated the frequency
would naturally become 3.3 kHz, but still limited to a 3 dB boost/cut.
So we would have:

3dB shelf at 10kHz (that's good)
3dB shelf at 5kHz (that quite good)
3dB shelf at 3.3kHz (that ? )

3dB shelf at 3.3kHz is pretty close to the BBC centre frequency for
"presence" I like the idea.

I had an e-mail from a German recording engineer earlier today
with comments on this thread. He said " Do you think that such
people buy an oil-painting and ask for a brush and some paints
so they can "correct" the colours in a Holbein?"

That's exactly as I would expect, it's the party line from those in the
recording industry. I have known some in the art/painting business and
while they wouldn't "ask for a brush and some paints so they could
"correct" the colours in a Holbein?", when lighting a painting I have
seen them take considerable pains with the number and positioning of
lights, as well as the colour temperature of the lights, and there are
no messy paints to get on your frock, this is not much different than
audio tone controls in concept.

To me, the audio equivalent of a correctly lit painting against a suitable
background is an amp/speaker combination in a room treated for
optimum performance. But then again, I can see applications where
a limited range equaliser might be of use. Many pop CDs are
excessively over-bright, and -3dB at 10kHz would go a long way to
correcting the additional EQ probably added at the mastering stage.

Best regards
Iain Churches

[John Byrns]

In article ,
"Iain Churches" wrote:

"John Byrns" wrote in message
...
In article ,
"Iain Churches" wrote:

"John Byrns" wrote in message
...

I have a few further thoughts, but first I am curious why you want to
use toggles, is it a technical, or a styling issue?

A little of each. I don't think anyone needs, or indeed should be
offered 20dB of EQ. And for style reasons, three knobs spread
across a 200mm front panel looks decidedly odd. (I am taking
the word of my discriminating panel of evaluators here)

It's not clear to me how the rotary switch vs. toggle switch question
relates how much EQ is offered, either approach could offer a maximum of
20 dB, 3 dB, or another number of your choice? As regards the style
issue, since the MF equalizer seems to have been unanimously dropped,
there would only be two knobs in any case.

A rotary switch could be up to 24 positions, but a toggle can have only
three: centre off, up= +x dB, down= -xdB. So a single step of 3dB
makes a lot more sense than a single step of say 6dB.

Yes, but there is no regulation that says a rotary switch can't be built
with only three positions like a toggle switch. The only reasons I can
see for using a toggle switch are styling reasons, or possibly for
mechanical reasons. The rotary switch has the edge when complex contact
configurations are required.

None the less I'm not sure I
agree with your "discriminating panel of evaluators" with regard to the
appearance of three knobs on a panel. No matter what the number of
knobs is the spacing between knobs as well as the spacing between the
knobs and the edges of the panel is exceedingly critical if the panel is
not to look decidedly odd, but if done right any reasonable number of
knobs can look good.

It's a simple exercise to mock up a panel (better than a drawing) in the
various combinations. Some look OK, others look decidedly odd, others
are pleasing.

I believe there are rules that can be applied to determine what
combinations are likely to look good and which aren't.

If you are considering building something for use by others and not just
for yourself, I feel it is essential to ask for opinions and also look very
carefully at the solutions adopted by others.


Another idea for multiple toggle switches would be to restrict the
maximum boost/cut to 3 dB in all cases and allow for three different
frequencies where the boost/cut becomes effective. This would also
simplify implementation if the third frequency which naturally follows
from the first two was acceptable. For example, if the HF equalizer had
a 10 kHz toggle and a 5 kHz toggle, if both were activated the frequency
would naturally become 3.3 kHz, but still limited to a 3 dB boost/cut.
So we would have:

3dB shelf at 10kHz (that's good)
3dB shelf at 5kHz (that quite good)
3dB shelf at 3.3kHz (that ? )

3dB shelf at 3.3kHz is pretty close to the BBC centre frequency for
"presence" I like the idea.

After some thought I have realized that the "Baxendall" style feedback
approach is the only way to go when using the two toggle approach to
combine a HF boost and cut to create a hump or dip in the response.
This is because when using the feedback approach, the boost and cut
actions are completely independent of each other, with no interaction.
The passive network has serious interaction problems when one switch is
set to boost and the other to cut.


Regards,

John Byrns

--
Surf my web pages at, http://fmamradios.com/

[Patrick Turner]

John Byrns wrote:

In article ,
"Iain Churches" wrote:

"John Byrns" wrote in message
...
In article ,
"Iain Churches" wrote:

"John Byrns" wrote in message
...

I have a few further thoughts, but first I am curious why you want to
use toggles, is it a technical, or a styling issue?

A little of each. I don't think anyone needs, or indeed should be
offered 20dB of EQ. And for style reasons, three knobs spread
across a 200mm front panel looks decidedly odd. (I am taking
the word of my discriminating panel of evaluators here)

It's not clear to me how the rotary switch vs. toggle switch question
relates how much EQ is offered, either approach could offer a maximum of
20 dB, 3 dB, or another number of your choice? As regards the style
issue, since the MF equalizer seems to have been unanimously dropped,
there would only be two knobs in any case.

A rotary switch could be up to 24 positions, but a toggle can have only
three: centre off, up= +x dB, down= -xdB. So a single step of 3dB
makes a lot more sense than a single step of say 6dB.

Yes, but there is no regulation that says a rotary switch can't be built
with only three positions like a toggle switch. The only reasons I can
see for using a toggle switch are styling reasons, or possibly for
mechanical reasons. The rotary switch has the edge when complex contact
configurations are required.

Farnell and others offer perfectly fine rotary wafer switches
which can have up to 12 wafers fitted to the same shaft,
and with anu number of stop positions from 2 to 12.
So tone settings with 5 positions each side of a 0.0dB position
can be done, so that the 0.0 position has zero reactive elements
in the shunt NFB path and the tone control stage may be used as an anode
follower.
You can't do this with pots.

But a line stage may be made with say gain tube, then pot or attenuator,
then anode follower as the buffer after the
gain attenurator.
This can all be done with only two triodes, say 12AU7/6SN7/6CG7.
But the better gain block is the µ-follower topology, so for deletable
gain stage
and the tone/buffer stage, 2 twin triodes are required.
The latter would be best for a professional grade amp because when the
input signal is fed direct to the gain pot,
the following stage with shunt FB needs a low Z source to give the
wanted attenuations.
The low Ra medium µ triodes do have low enough Ra to power into a shunt
NFB network.
But not a 12AX7.

So there is as usual a lot to think about and get right whichever
approach you take
so you have low N&D, and a simple reliabe circuit which sounds glorious.


BTW, another way to achieve logarithmic volume attenuation
but with a simple LINEAR pot or LINEAR switched attenuator is to have an
anode follower stage with its shunt NFB arranged so the linear pot is
the R2 of the shunt NFB network, and the wiper goes to the grid,
and if the pot is say 100k, then about 10k is used as R1 to the input,
and
100k taken to 0V to give the bias for the whole resistance network.
This means that a 12AU7 gain tube will have gain = about 5 max,
but only about 0.75 at the 12 o'clock position,
and maximum cut of about 40dB, so although the level never goes to no
signal at all,
the mimimum is low enough.

This means that the NFB present when a CD player signal is present means
that
the volume stage has near unity gain, and incredibly low THD,
exact channel balance.
least NFB occurs when the volume is high when the input signal is very
low.
Crown used this set up in their preamps.
Noise in the pot is reduced by the NFB.

Those amoung you wanting to try this little used but
effective way to control gain will find it works very well.

Patrick Turner.



None the less I'm not sure I
agree with your "discriminating panel of evaluators" with regard to the
appearance of three knobs on a panel. No matter what the number of
knobs is the spacing between knobs as well as the spacing between the
knobs and the edges of the panel is exceedingly critical if the panel is
not to look decidedly odd, but if done right any reasonable number of
knobs can look good.

It's a simple exercise to mock up a panel (better than a drawing) in the
various combinations. Some look OK, others look decidedly odd, others
are pleasing.

I believe there are rules that can be applied to determine what
combinations are likely to look good and which aren't.

If you are considering building something for use by others and not just
for yourself, I feel it is essential to ask for opinions and also look very
carefully at the solutions adopted by others.


Another idea for multiple toggle switches would be to restrict the
maximum boost/cut to 3 dB in all cases and allow for three different
frequencies where the boost/cut becomes effective. This would also
simplify implementation if the third frequency which naturally follows
from the first two was acceptable. For example, if the HF equalizer had
a 10 kHz toggle and a 5 kHz toggle, if both were activated the frequency
would naturally become 3.3 kHz, but still limited to a 3 dB boost/cut.
So we would have:

3dB shelf at 10kHz (that's good)
3dB shelf at 5kHz (that quite good)
3dB shelf at 3.3kHz (that ? )

3dB shelf at 3.3kHz is pretty close to the BBC centre frequency for
"presence" I like the idea.

After some thought I have realized that the "Baxendall" style feedback
approach is the only way to go when using the two toggle approach to
combine a HF boost and cut to create a hump or dip in the response.
This is because when using the feedback approach, the boost and cut
actions are completely independent of each other, with no interaction.
The passive network has serious interaction problems when one switch is
set to boost and the other to cut.

Regards,

John Byrns

--
Surf my web pages at, http://fmamradios.com/

[John Byrns]

In article ,
Patrick Turner wrote:

John Byrns wrote:

In article ,
"Iain Churches" wrote:

"John Byrns" wrote in message
...
In article ,
"Iain Churches" wrote:

"John Byrns" wrote in message
...

I have a few further thoughts, but first I am curious why you want
to
use toggles, is it a technical, or a styling issue?

A little of each. I don't think anyone needs, or indeed should be
offered 20dB of EQ. And for style reasons, three knobs spread
across a 200mm front panel looks decidedly odd. (I am taking
the word of my discriminating panel of evaluators here)

It's not clear to me how the rotary switch vs. toggle switch question
relates how much EQ is offered, either approach could offer a maximum
of
20 dB, 3 dB, or another number of your choice? As regards the style
issue, since the MF equalizer seems to have been unanimously dropped,
there would only be two knobs in any case.

A rotary switch could be up to 24 positions, but a toggle can have only
three: centre off, up= +x dB, down= -xdB. So a single step of 3dB
makes a lot more sense than a single step of say 6dB.

Yes, but there is no regulation that says a rotary switch can't be built
with only three positions like a toggle switch. The only reasons I can
see for using a toggle switch are styling reasons, or possibly for
mechanical reasons. The rotary switch has the edge when complex contact
configurations are required.

Farnell and others offer perfectly fine rotary wafer switches
which can have up to 12 wafers fitted to the same shaft,
and with anu number of stop positions from 2 to 12.
So tone settings with 5 positions each side of a 0.0dB position
can be done, so that the 0.0 position has zero reactive elements
in the shunt NFB path and the tone control stage may be used as an anode
follower.
You can't do this with pots.

Precisely, and pots also don't give you exactly the response curve you
might desire, except at their extreme settings.

But a line stage may be made with say gain tube, then pot or attenuator,
then anode follower as the buffer after the
gain attenurator.

That is true and is generally how I prefer to do it. However the
passive approach doesn't allow switching in a HF boost and an HF cut at
different frequencies without serious interaction between the
characteristics of the two curves, the pole and the zero affect one
another in the passive approach. The feedback approach effectively
isolates the pole and the zero from each other eliminating interaction.

This can all be done with only two triodes, say 12AU7/6SN7/6CG7.
But the better gain block is the µ-follower topology, so for deletable
gain stage
and the tone/buffer stage, 2 twin triodes are required.
The latter would be best for a professional grade amp because when the
input signal is fed direct to the gain pot,
the following stage with shunt FB needs a low Z source to give the
wanted attenuations.
The low Ra medium µ triodes do have low enough Ra to power into a shunt
NFB network.
But not a 12AX7.

So there is as usual a lot to think about and get right whichever
approach you take
so you have low N&D, and a simple reliabe circuit which sounds glorious.

In Ian's case if the equalizer is designed to follow his preamp, the
ALCF at the output of the preamp would eliminate any need for an input
buffer in the equalizer. If the equalizer is to be used with any random
preamp or source, then an input buffer stage would be required to assure
the required low impedance drive to the feedback equalizer stage.

I don't agree that there is a lot left to think about, it sounds to me
like we have worked out the design completely, and there is only one
potential question left. That question is whether it is acceptable to
take the output of the equalizer direct from the mu-follower feedback
equalizer stage, or if an ALCF should be inserted between the
mu-follower and the output jack?

What other questions remain to be resolved?


Regards,

John Byrns

--
Surf my web pages at, http://fmamradios.com/

[Patrick Turner]

John Byrns wrote:

In article ,
Patrick Turner wrote:

John Byrns wrote:

In article ,
"Iain Churches" wrote:

"John Byrns" wrote in message
...
In article ,
"Iain Churches" wrote:

"John Byrns" wrote in message
...

I have a few further thoughts, but first I am curious why you want
to
use toggles, is it a technical, or a styling issue?

A little of each. I don't think anyone needs, or indeed should be
offered 20dB of EQ. And for style reasons, three knobs spread
across a 200mm front panel looks decidedly odd. (I am taking
the word of my discriminating panel of evaluators here)

It's not clear to me how the rotary switch vs. toggle switch question
relates how much EQ is offered, either approach could offer a maximum
of
20 dB, 3 dB, or another number of your choice? As regards the style
issue, since the MF equalizer seems to have been unanimously dropped,
there would only be two knobs in any case.

A rotary switch could be up to 24 positions, but a toggle can have only
three: centre off, up= +x dB, down= -xdB. So a single step of 3dB
makes a lot more sense than a single step of say 6dB.

Yes, but there is no regulation that says a rotary switch can't be built
with only three positions like a toggle switch. The only reasons I can
see for using a toggle switch are styling reasons, or possibly for
mechanical reasons. The rotary switch has the edge when complex contact
configurations are required.

Farnell and others offer perfectly fine rotary wafer switches
which can have up to 12 wafers fitted to the same shaft,
and with anu number of stop positions from 2 to 12.
So tone settings with 5 positions each side of a 0.0dB position
can be done, so that the 0.0 position has zero reactive elements
in the shunt NFB path and the tone control stage may be used as an anode
follower.
You can't do this with pots.

Precisely, and pots also don't give you exactly the response curve you
might desire, except at their extreme settings.

But a line stage may be made with say gain tube, then pot or attenuator,
then anode follower as the buffer after the
gain attenurator.

That is true and is generally how I prefer to do it. However the
passive approach doesn't allow switching in a HF boost and an HF cut at
different frequencies without serious interaction between the
characteristics of the two curves, the pole and the zero affect one
another in the passive approach. The feedback approach effectively
isolates the pole and the zero from each other eliminating interaction.

This can all be done with only two triodes, say 12AU7/6SN7/6CG7.
But the better gain block is the µ-follower topology, so for deletable
gain stage
and the tone/buffer stage, 2 twin triodes are required.
The latter would be best for a professional grade amp because when the
input signal is fed direct to the gain pot,
the following stage with shunt FB needs a low Z source to give the
wanted attenuations.
The low Ra medium µ triodes do have low enough Ra to power into a shunt
NFB network.
But not a 12AX7.

So there is as usual a lot to think about and get right whichever
approach you take
so you have low N&D, and a simple reliabe circuit which sounds glorious.

In Ian's case if the equalizer is designed to follow his preamp, the
ALCF at the output of the preamp would eliminate any need for an input
buffer in the equalizer. If the equalizer is to be used with any random
preamp or source, then an input buffer stage would be required to assure
the required low impedance drive to the feedback equalizer stage.

I don't agree that there is a lot left to think about, it sounds to me
like we have worked out the design completely, and there is only one
potential question left. That question is whether it is acceptable to
take the output of the equalizer direct from the mu-follower feedback
equalizer stage, or if an ALCF should be inserted between the
mu-follower and the output jack?

What other questions remain to be resolved?

Matching the preamp to the proposed use.

Equiping it with enough features for the professional.

Every bit of gear needs a mission statement.

Patrick Turner.

Regards,

John Byrns

--
Surf my web pages at, http://fmamradios.com/