[Tynan AgviÐr]

Mr. Graham, and Mr. Turner(and anyone else)

does adding a mic amp section for each channel make a project like this
much harder?

The pendulum MDP-1 is one of my favorite mic amps, but it has all that
stuff that I do not need nor want. I just want

Mic Gain( I use a lot of ribbon mics and some low sens so 60-70 gain is a
good number for me)
Panpot
phantom power on/off
attenuator (for really loud sources)

and thats pretty much it.

How does one achieve the 48v phantom standard using tubes, and are there
any tube types that are more suitable for audio applications?

(I just discovered how 48v came to be..Norwegian State Television in 1966
was using a 48 V DV powering system for most of its gear(lights, etc)...
They wanted to take advantage of this readily available store of voltage to
power their capacitor microphones, rather than use a separate power supply
for each one. Georg Neumann and Co capitalized on it and the rest is
history)

[Tynan AgviÐr]

You mean proper gain staging. Attenuation is generally not the
smartest idea. Especially when voltage gain is hard to come by with as
with tubes.

Graham

Please forgive me if I sound like a fool. I sound like one because I am one
with this stuff..but
Doesnt attenuation raise the SNR a good bit? I know this is the case with
some other gear I have(on my DAC for example, there is implemented an
attenuator(labeled as a volume control). When engaged at all, the audio
suffers. I had the manufacturer(mytek) disable it completely.

[Tynan AgviÐr]

Eeyore wrote in
:



Patrick Turner wrote:

Eeyore wrote:
Patrick Turner wrote:

Gee, all the guy wants is 6 triodes ( 3 x 6CG7 ) with bjt CCS
anode supplies, then all anodes with 47k to a common point, with
a cap to CF output buffer. Volume pots for each channel can be
before each 1/2 6CG7 triode.

It's very clear that you know nothing about the requirements of
pro-audio working.

If you put a level control BEFORE the gain stage your signal to
noise ratio will be truly appaling.

Not with line level signals and with low µ triodes.

The OP requires a MICROPHONE mixer.

Graham


Yessir. A couple of people have suggested line mixers, but I have no
interest in that.(defeats the whole purpose of the thing for me) Those
are readily available and many to choose from (from well respected
companies too). Ideally I would just have to bring(in addition to an
accessories bag with XLR cables, tape, etc).

1,) mics
2.) mixer

3.) recorder

4.) possibly HQ ADC (depending on which recorder I use...some of them
have ADCs that sound like hot ass..)


Thanks Mr. Turner and Graham(and Rudy and Ian) yall are great.

Please keep discussing, I love learning from you Old Guard guys.

[Eeyore]

"Tynan AgviÐr" wrote:

Mr. Graham, and Mr. Turner(and anyone else)

does adding a mic amp section for each channel make a project like this
much harder?

It's the critical section. That doesn't mean it's *hard* per se but mic pres
are an area where subjective opinions tend to carry weight over technical
specs.

Often, a 'technically inferior' mic pre may be judged to 'sound better'. Of
course this is all related to what the 'tube sound' is all about.



The pendulum MDP-1 is one of my favorite mic amps, but it has all that
stuff that I do not need nor want. I just want

Mic Gain( I use a lot of ribbon mics and some low sens so 60-70 gain is a
good number for me)

Fine. You ned to consider how you want that gain to be adjusted. A rotary
switch with 10dB steps would seem to be the most suitable method here I reckon.
You WILL less than 60/70dB gain most of the time.


Panpot phantom power on/off
attenuator (for really loud sources)

A pad is easy. However using a pad in place of adjustable gain leads to
significantly poorer signal-to-noise ratio. I typically offer a pad ONLY to
deal with very high level sources.


and thats pretty much it.

How does one achieve the 48v phantom standard using tubes, and are there
any tube types that are more suitable for audio applications?

You mean how do you provide a tube regulated 48V DC supply ?

I'd laugh at anyone suggesting such an idea. Tubes have NOTHING of value to
offer as DC regulation devices other than design problems, poor DC accuracy
performance and unreliability.

I already have a VERY low noise 48V phantom supply design using semiconductors.
If you believe that the choice of tubes for a DC supply would have ANY audible
effect, I'd rather not take you on as a client since I hate dealing with
idiots.

Graham

[Eeyore]

"Tynan AgviÐr" wrote:

You mean proper gain staging. Attenuation is generally not the
smartest idea. Especially when voltage gain is hard to come by with as
with tubes.


Please forgive me if I sound like a fool. I sound like one because I am one
with this stuff..but
Doesnt attenuation raise the SNR a good bit?

No. Attenuation used in place of gain staging ALWAYS ensures a POORER
signal-to-noise ratio. This is a very fundamental concept. It's easily provable
to anyone with the requisite technical understanding. It's simply down to
mathematics actually.


I know this is the case with some other gear I have(on my DAC for example,
there is implemented an
attenuator (labeled as a volume control). When engaged at all, the audio
suffers. I had the manufacturer(mytek) disable it completely.

Suffers in what way ? A properly implemented variable attenuator / volume
control should have NO effect whatever on sound quality. If it did, there was a
design defect.

Graham

[Tynan AgviÐr]

Eeyore wrote in
:


Suffers in what way ? A properly implemented variable attenuator /
volume control should have NO effect whatever on sound quality. If it
did, there was a design defect.

Graham





Yes Sir, Thats what I meant..(I meant to say that it raises the noise , but
said raised the SNR(somehow I equated raising the SNR with raising the
noise level ..duh), but yes, I understand that concept at least, though not
much more.)

and about attenuators..didnt you say that attenuation gives a poorer
performance? What are the ways to adjust ouput/gain instead of attenuating?
Please go easy on me, I am not technically minded at all. I am just a dumb
musician.

[Tynan AgviÐr]

It's the critical section. That doesn't mean it's *hard* per se but
mic pres are an area where subjective opinions tend to carry weight
over technical specs.


so when I work with a designer, how to I go about conveying my ideas of
what sounds good, component wise? I mean, how does one go about selecting
transformers, tubes, etc?(The builder should be able to select items based
on what I tell them about my sound preferences, yes/?) I know of the
Lundahl that I mentioned because my electronaut m63 uses it(it had a Jensen
before that, and I did not like the sound as much).



Often, a 'technically inferior' mic pre may be judged to 'sound
better'. Of course this is all related to what the 'tube sound' is all
about.


My requirements are that 1.) No detail is lost...no rolled off highs, lows,
etc(and yes, I know Ribbon Mics are quite famous for this, but they work
very well for certaain applications)2.) that no details are lost, and that
the signal sounds "sweeter" than how it did before..bigger, Euphonic,
smooth..nothing taken away, just everything sweetened..

Fine. You ned to consider how you want that gain to be adjusted. A
rotary switch with 10dB steps would seem to be the most suitable
method here I reckon. You WILL less than 60/70dB gain most of the
time.

Yes Sir.
I like sliding faders better for "feel" but the rotaries are able to yield
better sonic results, yes? (because of the ability to use stepped
attenuation)

A pad is easy. However using a pad in place of adjustable gain leads
to significantly poorer signal-to-noise ratio. I typically offer a pad
ONLY to deal with very high level sources.


yes, but sometimes a pad is necessary with the mics that I have. For
example , I have a couple pairs of josephson c617 bodies with Gefell MK221
and MK202 capsules..those things are so sensitive that with any moderate
levels at all, it is out of control without a pad. . Just impossible. Just
as in mics, I like having the PAD as a safety device.(I am thinking about a
limiter too)



You mean how do you provide a tube regulated 48V DC supply ?


Yes Sir. Sorry for my poor wording. Ill get better, I promise.
I wasnt suggesting anything, I am just curious as ive seen t


I already have a VERY low noise 48V phantom supply design using
semiconductors. If you believe that the choice of tubes for a DC
supply would have ANY audible effect, I'd rather not take you on as a
client since I hate dealing with idiots.

Graham

Well, I am an idiot when it comes to this stuff, so I guess that rules you
out to work with me. I only know what sounds I like. How to get there
electroncally..hell if I know. I always rely on specialists for that.
So no...I wasnt suggesting anything. I was asking a question....If there
comes a time when I want to make a statement, I will do so in a manner that
leaves no doubts as to what I was doing. . If I had my own ideas about this
stuff I wouldnt be looking for help. I can tolerate "idiots" as they can
learn, provided they want to. Inflammatory and disrespectful assholes are
the only people I refuse to work with(one big reason I dont sing nearly as
much as I used to)..Opera and Classical Music--full of Arrogant Assholes.
Can not stand that type of person.

thanks for your continued help, Sir. It is much appreciated

[Eeyore]

"Tynan AgviÐr" wrote:

Eeyore wrote in

Suffers in what way ? A properly implemented variable attenuator /
volume control should have NO effect whatever on sound quality. If it
did, there was a design defect.


Yes Sir, Thats what I meant..(I meant to say that it raises the noise , but
said raised the SNR(somehow I equated raising the SNR with raising the
noise level ..duh), but yes, I understand that concept at least, though not
much more.)

Raising the SNR (more dBs of SNR) is the same thing as lowering the noise, so
you had that one the wrong way round.

If using a volume control adversely affect the SNR by a serious amount then
it's probably been placed at the wrong point in the signal path.


and about attenuators..didnt you say that attenuation gives a poorer
performance?

Yes it does since you're 'losing' signal, yet the following circuit's noise is
unchanged. However sometimes the signal you're attenuating is SO large that
this is not really an issue any more. I would use an attenuator ONLY to allow a
mic input to cope with very high signal levels as encountered with close miking
for example so that the mic doesn't overload the input stage. Using an
attenuator to adjust overall gain through a circuit block will ALWAYS degrade
performance, better to use active gain controls.


What are the ways to adjust ouput/gain instead of attenuating?

What's called 'gain staging' and/or variable feedback networks. If the negative
feedback of a stage is adjustable, then so is its gain without adversely
affecting its SNR performance the way an attenuator does.


Please go easy on me, I am not technically minded at all. I am just a dumb
musician.

Understood.

Graham

[Ian Thompson-Bell]

Tynan AgviÐr wrote:
Mr. Graham, and Mr. Turner(and anyone else)

does adding a mic amp section for each channel make a project like this
much harder?

The pendulum MDP-1 is one of my favorite mic amps, but it has all that
stuff that I do not need nor want. I just want

Mic Gain( I use a lot of ribbon mics and some low sens so 60-70 gain is a
good number for me)
Panpot
phantom power on/off
attenuator (for really loud sources)

and thats pretty much it.


Phase reverse??

How does one achieve the 48v phantom standard using tubes, and are there
any tube types that are more suitable for audio applications?


With tubes only you basically need an input transformer and as you know
there are plenty of good ones out there. Phantom power is then easily
applied to the primary of the transformer - it's just another power supply.

As to tube types more suitable to audio applications there is a strong
potential for another flame war here. That said, the ECC81/82/83 series
are popular triode choices along with the EF86 pentode.

Cheers

Ian

[Eeyore]

Ian Thompson-Bell wrote:

Tynan AgviÐr wrote:
Mr. Graham, and Mr. Turner(and anyone else)

does adding a mic amp section for each channel make a project like this
much harder?

The pendulum MDP-1 is one of my favorite mic amps, but it has all that
stuff that I do not need nor want. I just want

Mic Gain( I use a lot of ribbon mics and some low sens so 60-70 gain is a
good number for me)
Panpot
phantom power on/off
attenuator (for really loud sources)

and thats pretty much it.

Phase reverse??

Good pont. That may be useful.


How does one achieve the 48v phantom standard using tubes, and are there
any tube types that are more suitable for audio applications?

With tubes only you basically need an input transformer and as you know
there are plenty of good ones out there. Phantom power is then easily
applied to the primary of the transformer - it's just another power supply.

As to tube types more suitable to audio applications there is a strong
potential for another flame war here. That said, the ECC81/82/83 series
are popular triode choices along with the EF86 pentode.

I'd imagine the ECC83/12AX7 would be the tube of choice here if the triode
characteristic sound is what's required.

Graham

[Ian Thompson-Bell]

Tynan AgviÐr wrote:
Eeyore wrote in
:

Suffers in what way ? A properly implemented variable attenuator /
volume control should have NO effect whatever on sound quality. If it
did, there was a design defect.

Graham





Yes Sir, Thats what I meant..(I meant to say that it raises the noise , but
said raised the SNR(somehow I equated raising the SNR with raising the
noise level ..duh), but yes, I understand that concept at least, though not
much more.)

and about attenuators..didnt you say that attenuation gives a poorer
performance? What are the ways to adjust ouput/gain instead of attenuating?
Please go easy on me, I am not technically minded at all. I am just a dumb
musician.




This is a complex subject which depends on where the sources of noise
are and where in the amplification chain the attenuator is placed. Fot
the sake of argument suppose we place an attenuator before a unity gain
buffer. Suppose the self noise of the buffer is -90dBu.

With the attenuator set to maximum volume (no attenuation) suppose we
feed it a signal at 0dBu. The noise at the output of the buffer depends
both on its own noise *and* the noise in the signal. Suppose the noise
in the signal is at -70dBu, then the noise after the buffer will be near
enough -70dBu because the buffer noise is much lower. Suppose now we
turn down the attenuator to give 20dB attenuation. The signal is now
-20dBu and the signal in the noise is -90dBu. This noise gets adds to
the buffer noise at -90dBu which increases the noise at the output by
3dB to -87dBu. The signal to noise has just got 3dB worse.

This is a bit of an artificial example but it serves to illustrate the
problem. As a rule, S/N ratio of a mic preamp is dominated by the first
stage noise.

In mixer design there is always a balance between headroom (how much
bigger the signal can get before unacceptable distortion) and the noise
floor. This means that is it sometimes (in fact pretty much always)
necessary to include an attenuator at some point in the circuit to
maintain headroom. This is usually when signal levels are very high so
the system noise is not usually a problem.

In a mic preamp for example, the first amplification stage will often
have its gain set by altering negative feedback. However, the range of
gains available is usually limited between 20dB to 60dB gain. With a
10:1 ratio transformer giving another 20dB of gain a total of 80dB gain
is possible. If you want a total input level range from 0dBu to -80dBu
then some of this will involve gain and some attenuation because a 0dBu
signal will get a minimum of 40dB gain (20dB from the transformer and
20dB from the preamp) which would attempt to produce an output of +40dBu
(about 77 volts rms!!) not to mention the distortion the first tube
would produce because its grid signal is 7 vrms. To ensure there is
sufficient headroom you would probably design the preamp for a nominal
output level of 0dBu which means we need to introduce some attenuation
at the input of the preamp in order to handle signal levels between 0dBu
and -40dBu. The classic way to achieve all this is with a multi-way
rotary switch which controls both the preamp gain and the input attenuation.

Cheers

Ian

[Ian Thompson-Bell]

Tynan AgviÐr wrote:


so when I work with a designer, how to I go about conveying my ideas of
what sounds good, component wise? I mean, how does one go about selecting
transformers, tubes, etc?(The builder should be able to select items based
on what I tell them about my sound preferences, yes/?)

No way. Sounds good is purely subjective and the tonal qualities of a
tube mic preamp will be affected by so many factors it is not possible
to say 'how good' it will sound in advance.

I know of the
Lundahl that I mentioned because my electronaut m63 uses it(it had a Jensen
before that, and I did not like the sound as much).

Obviously if you have particular components you prefer that will help.

Cheers

Ian

[John Byrns]

In article ,
Ian Thompson-Bell wrote:

Tynan AgviÐr wrote:

How does one achieve the 48v phantom standard using tubes, and are there
any tube types that are more suitable for audio applications?


With tubes only you basically need an input transformer and as you know
there are plenty of good ones out there. Phantom power is then easily
applied to the primary of the transformer - it's just another power supply.

As to tube types more suitable to audio applications there is a strong
potential for another flame war here. That said, the ECC81/82/83 series
are popular triode choices along with the EF86 pentode.

Don't forget the 5879 pentode, my personal favorite. Two of these make
a great microphone amplifier. The 5879 can also be strapped for triode
operation if desired.


Regards,

John Byrns

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

[RickH]

On Jan 11, 2:07*pm, "Iain Churches" wrote:
--
Iain
Aural perception is a skill that requires study and careful development over
along period of time. Few have it as a natural gift."Arny Krueger" wrote in message


Mixing in the digital domain for the most part sounds compromised,
there is nothing quite as open and deep as a good analog mix. Take a
listen to the 1960's Frank Sinatra / Antonio Carlos Jobim album, the
mix on that puts ALL digital recordings of late to shame. It was most
likely mixed on a custom tube mixer, incredible as it seems but I dont
see digital domain mixing (unless you've spent $80,000 on the
equipment) matching the best analog mixers which are much cheaper.
The OP is on the right track if he is attempting to match the quality
of that Frank/Jobim album for example.

[Arny Krueger]

"RickH" wrote in message


Mixing in the digital domain for the most part sounds
compromised,

Not enough noise and distortion for you, I take it?

there is nothing quite as open and deep as a good analog mix.

Why?

Take a
listen to the 1960's Frank Sinatra / Antonio Carlos Jobim album, the
mix on that puts ALL digital recordings of late to shame.

Talk about a BS comparison - where is the comparable digitally mixed
recording of 1960's Frank Sinatra / Antonio Carlos Jobim to compare with?

What the writer is doing is comparing a nice legacy album with a creation of
his imagination. He imagines that the same album produced digitally would
sound crappy, and that supposed to be definitive proof?

LOL!

It was most likely mixed on a custom tube mixer, incredible as it
seems but I dont
see digital domain mixing (unless you've spent $80,000 on the
equipment) matching the best analog mixers which are much
cheaper.

The best analog mixers don't cost less than $80,000. It is easy to get up
into the quarter-half $million range if you're talking about a top quality
analog mixer.

I'll bet money you've never mixed anything of consequence digtally, and
you're talking out the back of your neck.

The OP is on the right track if he is attempting to match the quality
of that Frank/Jobim album for example.

In any performance the talent and quality of execution provided by the
performers is of the essence.

The next time I see Frank and Tony hanging out looking for a gig, I'll send
them your way... ;-)

[Eeyore]

Ian Thompson-Bell wrote:

Tynan AgviÐr wrote:
Eeyore wrote

Suffers in what way ? A properly implemented variable attenuator /
volume control should have NO effect whatever on sound quality. If it
did, there was a design defect.

Graham


Yes Sir, Thats what I meant..(I meant to say that it raises the noise , but
said raised the SNR(somehow I equated raising the SNR with raising the
noise level ..duh), but yes, I understand that concept at least, though not
much more.)

and about attenuators..didnt you say that attenuation gives a poorer
performance? What are the ways to adjust ouput/gain instead of attenuating?
Please go easy on me, I am not technically minded at all. I am just a dumb
musician.




This is a complex subject which depends on where the sources of noise
are and where in the amplification chain the attenuator is placed. Fot
the sake of argument suppose we place an attenuator before a unity gain
buffer. Suppose the self noise of the buffer is -90dBu.

With the attenuator set to maximum volume (no attenuation) suppose we
feed it a signal at 0dBu. The noise at the output of the buffer depends
both on its own noise *and* the noise in the signal. Suppose the noise
in the signal is at -70dBu, then the noise after the buffer will be near
enough -70dBu because the buffer noise is much lower. Suppose now we
turn down the attenuator to give 20dB attenuation. The signal is now
-20dBu and the signal in the noise is -90dBu. This noise gets adds to
the buffer noise at -90dBu which increases the noise at the output by
3dB to -87dBu. The signal to noise has just got 3dB worse.

This is a bit of an artificial example but it serves to illustrate the
problem. As a rule, S/N ratio of a mic preamp is dominated by the first
stage noise.

In mixer design there is always a balance between headroom (how much
bigger the signal can get before unacceptable distortion) and the noise
floor. This means that is it sometimes (in fact pretty much always)
necessary to include an attenuator at some point in the circuit to
maintain headroom. This is usually when signal levels are very high so
the system noise is not usually a problem.

In a mic preamp for example, the first amplification stage will often
have its gain set by altering negative feedback. However, the range of
gains available is usually limited between 20dB to 60dB gain. With a
10:1 ratio transformer giving another 20dB of gain a total of 80dB gain
is possible. If you want a total input level range from 0dBu to -80dBu
then some of this will involve gain and some attenuation because a 0dBu
signal will get a minimum of 40dB gain (20dB from the transformer and
20dB from the preamp) which would attempt to produce an output of +40dBu
(about 77 volts rms!!) not to mention the distortion the first tube
would produce because its grid signal is 7 vrms. To ensure there is
sufficient headroom you would probably design the preamp for a nominal
output level of 0dBu which means we need to introduce some attenuation
at the input of the preamp in order to handle signal levels between 0dBu
and -40dBu. The classic way to achieve all this is with a multi-way
rotary switch which controls both the preamp gain and the input attenuation.

Cheers

Ian

Nicely explained.

Graham

[Eeyore]

John Byrns wrote:

Ian Thompson-Bell wrote:
Tynan AgviÐr wrote:

How does one achieve the 48v phantom standard using tubes, and are there
any tube types that are more suitable for audio applications?


With tubes only you basically need an input transformer and as you know
there are plenty of good ones out there. Phantom power is then easily
applied to the primary of the transformer - it's just another power supply.

As to tube types more suitable to audio applications there is a strong
potential for another flame war here. That said, the ECC81/82/83 series
are popular triode choices along with the EF86 pentode.

Don't forget the 5879 pentode, my personal favorite. Two of these make
a great microphone amplifier. The 5879 can also be strapped for triode
operation if desired.

What bothers me about using a pentode is that you won't get the 'triode sound'
which is what in my estimation most 'tubeophiles' are seeking.

If I wanted to build a truly transparent mic amp I'd use semiconductors out of
preference since such a crcuit will certainly technically outperfrom any
equivalent tube circuitry in terms of outright linearity, noise, bandwidth and
response flatness by several orders of magnitude.

Graham

[robert casey]

Patrick Turner wrote:

robert casey wrote:

Mixing in the way you suggest will result in intermodulation !



This time I agree with you. Ran some spice simulations of the circuit,
with 3 6CG7 triodes wired up as described. Applied to each one a 100mv
sine wave. One triode got 1KHz, the 2nd 1.1KHz, and the 3rd 1.3KHz.
Got a lot of intermod, some only 70dB down, at 2.7KHz, 2.9Khz and 3KHz.




The natural level of IMD within the main signal being handled is far
greater than that caused by cross talk.
Your analysis is very poor, and you have NOT stated all conditions of
the test you say you have simulated

Upon further review, I found my mistake. I didn't run the simulation
long enough (DUH!), and the Forrier analysis was thus bogus. Ran it a
lot longer, and the IMD is at least 100dB down.

[John Byrns]

In article ,
Eeyore wrote:

John Byrns wrote:

Ian Thompson-Bell wrote:
Tynan AgviÐr wrote:

How does one achieve the 48v phantom standard using tubes, and are
there
any tube types that are more suitable for audio applications?


With tubes only you basically need an input transformer and as you know
there are plenty of good ones out there. Phantom power is then easily
applied to the primary of the transformer - it's just another power
supply.

As to tube types more suitable to audio applications there is a strong
potential for another flame war here. That said, the ECC81/82/83 series
are popular triode choices along with the EF86 pentode.

Don't forget the 5879 pentode, my personal favorite. Two of these make
a great microphone amplifier. The 5879 can also be strapped for triode
operation if desired.

What bothers me about using a pentode is that you won't get the 'triode
sound'
which is what in my estimation most 'tubeophiles' are seeking.

You say most "tubeophiles" are seeking the "triode sound", can you
explain what you mean when you refer to the "triode sound", and contrast
it with the sound of pentodes?

If I wanted to build a truly transparent mic amp I'd use semiconductors out
of
preference since such a crcuit will certainly technically outperfrom any
equivalent tube circuitry in terms of outright linearity, noise, bandwidth
and
response flatness by several orders of magnitude.

How much "transparency" do you really need? Once you achieve a certain
level of transparency you reach a point of diminishing returns, and
other factors become more important in a recording than overkill in the
microphone amplifier and mixer.


Regards,

John Byrns

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

[John Byrns]

In article ,
robert casey wrote:

Patrick Turner wrote:

robert casey wrote:

Mixing in the way you suggest will result in intermodulation !

This time I agree with you. Ran some spice simulations of the circuit,
with 3 6CG7 triodes wired up as described. Applied to each one a 100mv
sine wave. One triode got 1KHz, the 2nd 1.1KHz, and the 3rd 1.3KHz.
Got a lot of intermod, some only 70dB down, at 2.7KHz, 2.9Khz and 3KHz.

The natural level of IMD within the main signal being handled is far
greater than that caused by cross talk.
Your analysis is very poor, and you have NOT stated all conditions of
the test you say you have simulated

Upon further review, I found my mistake. I didn't run the simulation
long enough (DUH!), and the Forrier analysis was thus bogus. Ran it a
lot longer, and the IMD is at least 100dB down.

So from that can we conclude that it is not necessary to use either a
cathode follower buffer, or a virtual earth scheme in a tube based mixer
of this sort?


Regards,

John Byrns

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

[Eeyore]

John Byrns wrote:

Eeyore wrote:
John Byrns wrote:
Ian Thompson-Bell wrote:
Tynan AgviÐr wrote:

How does one achieve the 48v phantom standard using tubes, and are
there any tube types that are more suitable for audio applications?


With tubes only you basically need an input transformer and as you know
there are plenty of good ones out there. Phantom power is then easily
applied to the primary of the transformer - it's just another power
supply.

As to tube types more suitable to audio applications there is a strong
potential for another flame war here. That said, the ECC81/82/83 series
are popular triode choices along with the EF86 pentode.

Don't forget the 5879 pentode, my personal favorite. Two of these make
a great microphone amplifier. The 5879 can also be strapped for triode
operation if desired.

What bothers me about using a pentode is that you won't get the 'triode
sound' which is what in my estimation most 'tubeophiles' are seeking.

You say most "tubeophiles" are seeking the "triode sound", can you
explain what you mean when you refer to the "triode sound", and contrast
it with the sound of pentodes?

The harmonic character of the distortion products created by the non-linear
transfer characteristic. The transfer characteristic of the triode and pentode
are markedly different.

http://www.google.com/search?n&q=%22triode+equation

Graham

[Eeyore]

John Byrns wrote:

Eeyore wrote:

If I wanted to build a truly transparent mic amp I'd use semiconductors out
of preference since such a crcuit will certainly technically outperfrom any
equivalent tube circuitry in terms of outright linearity, noise, bandwidth
and response flatness by several orders of magnitude.

How much "transparency" do you really need? Once you achieve a certain
level of transparency you reach a point of diminishing returns, and
other factors become more important in a recording than overkill in the
microphone amplifier and mixer.

How much transparency do you want ? Why take risks on choosing an inadequate
figure ? If the circuit can be made so linear as to provide say 0.001% THD (at
the same time as having excellent noise and frequency response) why not do so ?
It doesn't cost much. Only a few dollars.

Graham

[Eeyore]

John Byrns wrote:

robert casey wrote:
Patrick Turner wrote:
robert casey wrote:

Mixing in the way you suggest will result in intermodulation !

This time I agree with you. Ran some spice simulations of the circuit,
with 3 6CG7 triodes wired up as described. Applied to each one a 100mv
sine wave. One triode got 1KHz, the 2nd 1.1KHz, and the 3rd 1.3KHz.
Got a lot of intermod, some only 70dB down, at 2.7KHz, 2.9Khz and 3KHz.

The natural level of IMD within the main signal being handled is far
greater than that caused by cross talk.
Your analysis is very poor, and you have NOT stated all conditions of
the test you say you have simulated

Upon further review, I found my mistake. I didn't run the simulation
long enough (DUH!), and the Forrier analysis was thus bogus. Ran it a
lot longer, and the IMD is at least 100dB down.

So from that can we conclude that it is not necessary to use either a
cathode follower buffer, or a virtual earth scheme in a tube based mixer
of this sort?

If you don't buffer, the mixing will have to be done at relatively high
impedance and that has both cross-talk and noise issues.

Graham

[Eeyore]

John Byrns wrote:

Eeyore wrote:

What bothers me about using a pentode is that you won't get the 'triode
sound' which is what in my estimation most 'tubeophiles' are seeking.

You say most "tubeophiles" are seeking the "triode sound", can you
explain what you mean when you refer to the "triode sound", and contrast
it with the sound of pentodes?

Scott Reynolds and Marshall Leach model vacuum tubes as voltage-controlled
current sources whose output current is a weighted sum of controlling element
voltages raised to the three-halves power. The controlling elements are different
for triodes and pentodes. (Since the suppressor grid has little effect on the
transfer characteristics, pentodes are modeled as tetrodes.) The plate current
equation for triodes is

IP = (EG+EP /µ)3/2/kG1 for EG+EP /µ = 0
= 0 otherwise.

The plate current equation for pentodes is

IP = (2/kG1p)(EG+EG2 /µ)3/2 arctan(EP /kVB ) for EG+EG2 /µ = 0
= 0 otherwise.

where IP is plate current, EG is control grid voltage, EP is plate voltage, EG2
is screen grid voltage (all voltages with respect to the cathode), µ is the
amplification factor, and kG1 is a factor used to fit the equation to data. The
pentode equation differs from the triode equation in that the screen grid
replaces the plate as a controlling element, and an arctangent term (from Scott
Reynolds' model) is added to model a response curve "knee" whose location is
proportional to kVB. Equations (1) and (2) are forms of the Langmuir-Childs law,
which can be derived from fundamental physics.3 Plate curves for a 12AX7 triode
with µ = 93 (below the specified value but a good overall fit to equation (1))
are illustrated below in Figure 1.

http://www.normankoren.com/Audio/Tub...e_article.html

Graham

[ian]

John Byrns wrote:
In article ,
Ian Thompson-Bell wrote:

Tynan AgviÐr wrote:

How does one achieve the 48v phantom standard using tubes, and are there
any tube types that are more suitable for audio applications?

With tubes only you basically need an input transformer and as you know
there are plenty of good ones out there. Phantom power is then easily
applied to the primary of the transformer - it's just another power supply.

As to tube types more suitable to audio applications there is a strong
potential for another flame war here. That said, the ECC81/82/83 series
are popular triode choices along with the EF86 pentode.

Don't forget the 5879 pentode, my personal favorite. Two of these make
a great microphone amplifier. The 5879 can also be strapped for triode
operation if desired.


Regards,

John Byrns


Not familiar with these. Can't find a data sheet on the www but from
what I have found the seem a bit like the EF86

Cheers

Ian.

[ian]

Eeyore wrote:
Nicely explained.

Graham



Thanks

Ian

[ian]

Eeyore wrote:

I'd imagine the ECC83/12AX7 would be the tube of choice here if the triode
characteristic sound is what's required.

Graham


They would be my first choice - there are plenty of different
manufacturers so changing tone by changing manufacturer is an option too.

Cheers

Ian

[John Byrns]

In article ,
Eeyore wrote:

John Byrns wrote:

Eeyore wrote:
John Byrns wrote:
Ian Thompson-Bell wrote:
Tynan AgviÐr wrote:

How does one achieve the 48v phantom standard using tubes, and are
there any tube types that are more suitable for audio applications?


With tubes only you basically need an input transformer and as you
know
there are plenty of good ones out there. Phantom power is then easily
applied to the primary of the transformer - it's just another power
supply.

As to tube types more suitable to audio applications there is a
strong
potential for another flame war here. That said, the ECC81/82/83
series
are popular triode choices along with the EF86 pentode.

Don't forget the 5879 pentode, my personal favorite. Two of these make
a great microphone amplifier. The 5879 can also be strapped for triode
operation if desired.

What bothers me about using a pentode is that you won't get the 'triode
sound' which is what in my estimation most 'tubeophiles' are seeking.

You say most "tubeophiles" are seeking the "triode sound", can you
explain what you mean when you refer to the "triode sound", and contrast
it with the sound of pentodes?

The harmonic character of the distortion products created by the non-linear
transfer characteristic. The transfer characteristic of the triode and
pentode
are markedly different.

http://www.google.com/search?n&q=%22triode+equation

Graham, you are avoiding the question, I know that triodes and pentodes
have markedly different transfer characteristics. I didn't ask you
about transfer characteristics, I asked you what you mean by "triode
sound", and how it differs from the sound of other devices?


Regards,

John Byrns

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

[John Byrns]

In article ,
Eeyore wrote:

John Byrns wrote:

Eeyore wrote:

If I wanted to build a truly transparent mic amp I'd use semiconductors
out
of preference since such a crcuit will certainly technically outperfrom
any
equivalent tube circuitry in terms of outright linearity, noise,
bandwidth
and response flatness by several orders of magnitude.

How much "transparency" do you really need? Once you achieve a certain
level of transparency you reach a point of diminishing returns, and
other factors become more important in a recording than overkill in the
microphone amplifier and mixer.

How much transparency do you want ? Why take risks on choosing an inadequate
figure ? If the circuit can be made so linear as to provide say 0.001% THD
(at
the same time as having excellent noise and frequency response) why not do so
?
It doesn't cost much. Only a few dollars.

The dollars all add up in the end. What is the point in providing a THD
of say 0.001%? Surely there are more important things to worry about
and to spend ones money on.


Regards,

John Byrns

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

[John Byrns]

In article ,
Eeyore wrote:

John Byrns wrote:

robert casey wrote:
Patrick Turner wrote:
robert casey wrote:

Mixing in the way you suggest will result in intermodulation !

This time I agree with you. Ran some spice simulations of the circuit,
with 3 6CG7 triodes wired up as described. Applied to each one a 100mv
sine wave. One triode got 1KHz, the 2nd 1.1KHz, and the 3rd 1.3KHz.
Got a lot of intermod, some only 70dB down, at 2.7KHz, 2.9Khz and 3KHz.

The natural level of IMD within the main signal being handled is far
greater than that caused by cross talk.
Your analysis is very poor, and you have NOT stated all conditions of
the test you say you have simulated

Upon further review, I found my mistake. I didn't run the simulation
long enough (DUH!), and the Forrier analysis was thus bogus. Ran it a
lot longer, and the IMD is at least 100dB down.

So from that can we conclude that it is not necessary to use either a
cathode follower buffer, or a virtual earth scheme in a tube based mixer
of this sort?

If you don't buffer, the mixing will have to be done at relatively high
impedance and that has both cross-talk and noise issues.

What does the presence or absence of buffering have to do with the level
at which mixing is done? Crosstalk only becomes an issue if a single
input must feed more than one bus, in which case a buffer serves a
useful purpose. What does buffering have to do with "noise issues"?


Regards,

John Byrns

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

[John Byrns]

In article ,
Eeyore wrote:

John Byrns wrote:

Eeyore wrote:

What bothers me about using a pentode is that you won't get the 'triode
sound' which is what in my estimation most 'tubeophiles' are seeking.

You say most "tubeophiles" are seeking the "triode sound", can you
explain what you mean when you refer to the "triode sound", and contrast
it with the sound of pentodes?

Scott Reynolds and Marshall Leach model vacuum tubes as voltage-controlled
current sources whose output current is a weighted sum of controlling element
voltages raised to the three-halves power. The controlling elements are
different
for triodes and pentodes. (Since the suppressor grid has little effect on the
transfer characteristics, pentodes are modeled as tetrodes.) The plate
current
equation for triodes is

IP = (EG+EP /µ)3/2/kG1 for EG+EP /µ = 0
= 0 otherwise.

The plate current equation for pentodes is

IP = (2/kG1p)(EG+EG2 /µ)3/2 arctan(EP /kVB ) for EG+EG2 /µ = 0
= 0 otherwise.

where IP is plate current, EG is control grid voltage, EP is plate voltage,
EG2
is screen grid voltage (all voltages with respect to the cathode), µ is the
amplification factor, and kG1 is a factor used to fit the equation to data.
The
pentode equation differs from the triode equation in that the screen grid
replaces the plate as a controlling element, and an arctangent term (from
Scott
Reynolds' model) is added to model a response curve "knee" whose location is
proportional to kVB. Equations (1) and (2) are forms of the Langmuir-Childs
law,
which can be derived from fundamental physics.3 Plate curves for a 12AX7
triode
with µ = 93 (below the specified value but a good overall fit to equation
(1))
are illustrated below in Figure 1.

http://www.normankoren.com/Audio/Tub...e_article.html

I incorporated the work of those two gentleman into my spice models more
than a decade ago IIRC. But I don't understand how you are making the
jump from transfer curves to the "triode sound"? Can you explain what
you mean by the "triode sound"?


Regards,

John Byrns

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

[Eeyore]

John Byrns wrote:

Eeyore wrote:
John Byrns wrote:

You say most "tubeophiles" are seeking the "triode sound", can you
explain what you mean when you refer to the "triode sound", and contrast
it with the sound of pentodes?

The harmonic character of the distortion products created by the non-linear
transfer characteristic. The transfer characteristic of the triode and
pentode are markedly different.

http://www.google.com/search?n&q=%22triode+equation

Graham, you are avoiding the question, I know that triodes and pentodes
have markedly different transfer characteristics. I didn't ask you
about transfer characteristics, I asked you what you mean by "triode
sound", and how it differs from the sound of other devices?

Clearly the specific non-linearities of the transfer characteristic will affect
the sound. Do you not accept that ? Various devices have a 'sonic fingerprint' if
you like in this regard.

What are you trying to get at ?

Graham

[Eeyore]

John Byrns wrote:

Eeyore wrote:
John Byrns wrote:
Eeyore wrote:

If I wanted to build a truly transparent mic amp I'd use semiconductors
out of preference since such a crcuit will certainly technically outperfrom

any equivalent tube circuitry in terms of outright linearity, noise,
bandwidth and response flatness by several orders of magnitude.

How much "transparency" do you really need? Once you achieve a certain
level of transparency you reach a point of diminishing returns, and
other factors become more important in a recording than overkill in the
microphone amplifier and mixer.

How much transparency do you want ? Why take risks on choosing an inadequate
figure ? If the circuit can be made so linear as to provide say 0.001% THD
(at the same time as having excellent noise and frequency response) why not do
so
? It doesn't cost much. Only a few dollars.

The dollars all add up in the end.

About $3.50 for an LME49720 and 4810 in place of more humble offerings like the
NE5532 or NJM4580 in a 3 op-amp mic amp and about 6c extra for 2 additional
ultra-linearing transistors for the discrete front end on a design I have in mind..

No, you won't find them in a Behringer but that's not the kind of product we're
discussing here.


What is the point in providing a THD of say 0.001%? Surely there are more
important things to worry about and to spend ones money on.

What is the point in NOT making the very best mic amp you can ?

We're talking a tiny fraction of the cost of an input transformer. The added cost
of a couple of extra transistors and highly linear op-amps is negligible in this
respect.

I repeat.. Why would you knowingly make something for high-end recording worse than
it need be ?

Graham

[Eeyore]

John Byrns wrote:

Eeyore wrote:

http://www.normankoren.com/Audio/Tub...e_article.html

I incorporated the work of those two gentleman into my spice models more
than a decade ago IIRC.

Some how I doubt you made any SPICE models ever.


But I don't understand how you are making the jump from transfer curves to the
"triode sound"?

You must be fairly stupid in that case or simply hankering after an argument.


Can you explain what you mean by the "triode sound"?

You're just trolling now. Stop being a ****wit.

Graham

[Eeyore]

John Byrns wrote:

Eeyore wrote:
John Byrns wrote:

So from that can we conclude that it is not necessary to use either a
cathode follower buffer, or a virtual earth scheme in a tube based mixer
of this sort?

If you don't buffer, the mixing will have to be done at relatively high
impedance and that has both cross-talk and noise issues.

What does the presence or absence of buffering have to do with the level
at which mixing is done?

I said IMPEDANCE not level. Can't you bloody well read ?

Stop going off at a tangent to suit your own stupid argumentative agenda.

Graham

[Patrick Turner]

Arny Krueger wrote:

"Iain Churches" wrote in message
ti.fi

The Williamson, which dates from the forties, has a
bandwidth of 300kHz.

Not really. I have here a copy of the original August 1949 wireless world
article here before me.

It's quite obvious that this amp does not meet accepted standards for
stability. It's 300 KHz response is a cheat. It appears that the stability
margin at 300 KHz is about zero degrees. The open loop phase shift is so
great that the negative feedback might actually be verging on being positive
at 300 KHz.

Kroooooooooooooooger is right this time about the Williamson.

And for the Willy amp to actually have 300kHz of BW the OPT must be
a state of the art type which ordinary mortals never could afford in
1955,
and nobody can find or afford now.

So just about all samples of W configured amps with OPT between plain
awful
and mediocre will oscillate at LF without a load, and a 0.22 uF will
also make them oscillate violently at
some low RF.

But with appropriate gain shelving bvelow 30Hz and above 10kHz the W amp
can be made unconditionally
stable regardless od whether loaded or loaded by any kind of reactive
load able to be configured.



Modern feedback systems target stability margins on the order of 45 degrees.
Providing this amount of stability by ordinary means would restrict response
to about 120 KHz.

67kHz is the usual BW which is attained in most well designed amps with
good OPT
in order to get unconditional stability, even if one does have an OPT
with say 20Hz to 300kHz
full power ability. I know, i've wound such OPT, and had to deal with
the excessive BW.


Stability issues are mentioned in the article, along with suggested
work-arounds. This might have passe for high tech in 1949, but not even in
the mid-late 50s.

One clear indication of inadequate stability is the absence of a monotonic
roll-off.

I've also checked out the schematics of more modern versions of this amp,
and their treble frequency response extension is far less, no doubt in order
to provide commercial levels of stability.

The sine-wave output is quite presentable.

Not with the feedback verging on positive.

The same can be said for many bespoke amps available here in Europe using
Lundahl, Sowter or Welter transformers.

Nope. You won't find tubed amps from even the mid-1950s with non-existent
stability margins.

Many tube amps from the 1950 were ****ing awful with regard to
stability,
with many oscilating at LF if you ever left them turned on without a
load,
and then any hint of capacitance in the load would cause RF
oscillations.

See my website for all manner of treatments to avoid instability,
and how to get fabulous sound even when the OPT is pretty poor.

Williamson was one of the most ignored technocrats of the 1950s.
Nearly every prominent manufacturer tried to dumb down OPT design to
avoid the high costs
of doing things right and according to the Book.

The same bravado ignornace and utter BS tradition continues today and
gives tube amps a terrible reputation.
Asian makers are a particular menace!!!
They just mainly regurgitate the worst traditions of the mainstream
lowest common denominator of the western nations
1955 crap.


Patrick Turner.

[Patrick Turner]

Eeyore wrote:

Patrick Turner wrote:

Eeyore wrote:
Patrick Turner wrote:

J-fets CAN be trusted NOT to stuff up the music if they only have to
make a few millivolts of signal.

You may get as much as a VOLT out of low-Z mics when close miking rock and roll
bands.

Yeah, they do scream so loud eh!!!

And drum beats go way high.

So out comes the tube compressor and de-esser.......

In a previous post you said you'd need 80dB gain for mic, so make up
your mind....

It has nothing to do with "making up one's mind". Depending on application, a mic amp
may need as little as 0dB of voltage gain or as much as 80dB.

Not with a volt of mic signal.

Why do you think the likes of Lundahl spec their mic transformers with that kind of
level on the primary ?

Hence the need for attenuation placed sensibly....

You mean proper gain staging. Attenuation is generally not the smartest idea.
Especially when voltage gain is hard to come by with as with tubes.

Graham

I thought the OP guy wanted line mixing, after raising mic signals to
whatever levels.

But without a total block diagram of what is going to be used with all
details, its impossible to design anything seriously here in principle
in a professional way.

Patrick Turner.

[Patrick Turner]

Eeyore wrote:

Arny Krueger wrote:

"Iain Churches" wrote

The Williamson, which dates from the forties, has a
bandwidth of 300kHz.

I'll bet you anything it's not a full-power bandwidth ! I'm happy with as little
as 50kHz for that figure even with solid state.

Not really. I have here a copy of the original August 1949 wireless world
article here before me.

It's quite obvious that this amp does not meet accepted standards for
stability. It's 300 KHz response is a cheat. It appears that the stability
margin at 300 KHz is about zero degrees. The open loop phase shift is so
great that the negative feedback might actually be verging on being positive
at 300 KHz.

I suspect the typical loads of the day weren't as wildly reactive as we see now
so they may have got away with it.

Loads are usually rarely completely reactive, ever.

There is usually always some series R or shunt R in series with or
shunting
whatever reactance one is faced with.

Its always been this way.

So Quad ESL57 are about equivalent to having 2uF in series with 1.7
ohms.

2uF alone might cause a poor tube amp to oscilate itself to death, but
often
it won't, simply because at HF the Z becomes too low to prevent output
stage gain at HF.
Any poor amp, SS included might show a 6dB peak in sine wave response at
say 30kHz
loaded by 2uF, at low levels. This shows poor design. High 1kHz clipping
levels
cannot be sustained at HF into 2uF because the load at HF is only say 2
ohms at 40kHz,
and 1 ohm at 80kHz.

With the addition of a series 1.7 ohms, the response at low levels will
usaually become much flatter
as the R prevents the ultimate 90 degree phase lag at any HF at the
output,
and the load minimum at 80kHz is about 1.8 ohms, mainly resistive.

But many amps able to handle the 2uF plus 1.7ohmm ESL equivalent load
will
oscillate badly at HF with a pure 0.22 uF, because this reactive load
doesn't become
low enough Z to simply shunt the output at HF, but has just the right Z
to
cause a pahes shift as it reacts with the OPT leakage inductance which
makes all tube amps
have an inductive output impedance characteristic.

The leakage L in a crummy old Leak amp can be 50mH anode to anode and
with an OPT Z-ratio of 1,000:1
it becomes 50uH, and I leave you all to work out for yourselves how any
amp
feeding speakers will behave with a series 50uH between the amp and
speaker.

In well designed amps the series LL at the output should be 5uH max, and
the amp damped to reduce the
Q of what effectively is a second order filter at HF.

SS amps have to have Zobel network with a series L with parallel
resistor
to ensure that C loads or short circuits do not cause excessive device
currents above 20kHz.

Patrick Turner.


Graham

[Patrick Turner]

"Tynan AgviÐr" wrote:

Mr. Graham, and Mr. Turner(and anyone else)

does adding a mic amp section for each channel make a project like this
much harder?

Of course it does.

If you want each mic amp fully tubed...

But you need to KNOW exactly what you really want,
and be able to specify to a designer what microphone types are likely to
be used
and design accordingly.

Patrick Turner.

The pendulum MDP-1 is one of my favorite mic amps, but it has all that
stuff that I do not need nor want. I just want

Mic Gain( I use a lot of ribbon mics and some low sens so 60-70 gain is a
good number for me)
Panpot
phantom power on/off
attenuator (for really loud sources)

and thats pretty much it.

How does one achieve the 48v phantom standard using tubes, and are there
any tube types that are more suitable for audio applications?

(I just discovered how 48v came to be..Norwegian State Television in 1966
was using a 48 V DV powering system for most of its gear(lights, etc)...
They wanted to take advantage of this readily available store of voltage to
power their capacitor microphones, rather than use a separate power supply
for each one. Georg Neumann and Co capitalized on it and the rest is
history)

[Patrick Turner]

"Tynan AgviÐr" wrote:

You mean proper gain staging. Attenuation is generally not the
smartest idea. Especially when voltage gain is hard to come by with as
with tubes.

Graham

Please forgive me if I sound like a fool. I sound like one because I am one
with this stuff..but
Doesnt attenuation raise the SNR a good bit? I know this is the case with
some other gear I have(on my DAC for example, there is implemented an
attenuator(labeled as a volume control). When engaged at all, the audio
suffers. I had the manufacturer(mytek) disable it completely.

SNR all depends on the levels and the design.

The higher the microphone signal, the lower the amp noise will be in
relation.

Attenuation may or may not worsen SNR, or worsen distortions.

A block diagram of what exactly you will end up having
will avoid poor SNR and high distortions.

Unfortunately, you are not a very technically literate person and
you said you didn't want to deal with any schematic details.
Those of us whose second language IS Schematica understand you.
Your forte might be with guitar, but not with voltage levels and
devices.
Someone working to make what you want sure will have to deal with
technical issues though,
so you need to spell out clearly what mics you plan to use
so that levels and attenuation can be all tailored to suit
a best outcome.

Patrick Turner.