From a discussion on S.E.D

...Jim Thompson
--
| James E.Thompson, P.E. | mens |
| Analog Innovations, Inc. | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| Phoenix, Arizona Voice480)460-2350 | |
| E-mail Address at Website Fax480)460-2142 | Brass Rat |
| http://www.analog-innovations.com | 1962 |

I love to cook with wine. Sometimes I even put it in the food.

I thought you were going to move those 2.2k too.
that poor op amp is sitting there now with its supply all exposed.
and you need to put the gain controll back somewhere too.
it needs to cope with 1.5v ip aparantly, but not at full gain ofc
whatever could cuase 1.5 from a mic who knows ...

Colin =^.^=

Jim Thompson wrote:

From a discussion on S.E.D

Congratulations. Your R15 has just degraded the noise figure.

Graham

colin wrote:

whatever could cuase 1.5 from a mic who knows ...

Rock and roll.

Graham

"colin"

I thought you were going to move those 2.2k too.
that poor op amp is sitting there now with its supply all exposed.


** TL071 op-amps do have pretty a fair PSSR - you know.

and you need to put the gain controll back somewhere too.


** Not just for seeing what the full gain noise figure is.


it needs to cope with 1.5v ip aparantly, but not at full gain ofc
whatever could cuase 1.5 from a mic who knows ...


** Dynamic mics can be made to produce such an output level - if parked in
front of a loudspeaker box, snare drum or a singer with the right technique.
Many condenser mics can output much more.

But the lowest gain setting is needed to set the output level of the pre-amp
to around 1 volt rms, with very loud sources.

This will be achieved with about 160 mV rms from the mic or a 135 dB SPL in
the mid/upper audio band.

Less SPL is needed at low frequencies due to "proximity effect " causing up
to 20 dB bass boost with most cardioid mics.



....... Phil

Eeyore wrote:

Jim Thompson wrote:

From a discussion on S.E.D

Congratulations. Your R15 has just degraded the noise figure.

Correction. not so. I'd missed the value change.

Graham

"Eesyore"

Jim Thompson wrote:

From a discussion on S.E.D

Congratulations. Your R15 has just degraded the noise figure.

Correction. not so. I'd missed the value change.



** The noise ( at 1.36 nV rt/Hz ) is just a little worse than for the
Project 66 schematic.

Note: that figure is for a shorted input rather than a 200 ohms source.

Current in the transistors is the same, but there is one resistor change
that adds a pinch.

The 33 pF caps produce an earlier HF roll off than project 66 does.




........ Phil

Phil Allison wrote:

"Eesyore"

Jim Thompson wrote:

From a discussion on S.E.D

Congratulations. Your R15 has just degraded the noise figure.

Correction. not so. I'd missed the value change.


** The noise ( at 1.36 nV rt/Hz ) is just a little worse than for the
Project 66 schematic.

Note: that figure is for a shorted input rather than a 200 ohms source.

Current in the transistors is the same, but there is one resistor change
that adds a pinch.

The 33 pF caps produce an earlier HF roll off than project 66 does.

-3dB @ 32 kHz in fact. A bit low for my liking for sure. That'll be intrusive at
20kHz too..

Graahm

" Eesyore"


** The noise ( at 1.36 nV rt/Hz ) is just a little worse than for the
Project 66 schematic.

Note: that figure is for a shorted input rather than a 200 ohms source.

Current in the transistors is the same, but there is one resistor change
that adds a pinch.

The 33 pF caps produce an earlier HF roll off than project 66 does.

-3dB @ 32 kHz in fact. A bit low for my liking for sure. That'll be
intrusive at
20kHz too..



** In a recording or PA system there are very many amplification stages,
each one with a LF and HF roll off curve, and all of them ACCUMULATE !!

If the overall aim is a flat response +/- 1 dB from 25Hz to 20 kHz -
then each stage needs to be a whole lot flatter that that aim.

However, given the input device is a dynamic microphone, few of which have
good response beyond 15 kHz , a -1dB point at 16 kHz is not so bad.

The treble EQ pot can always be given a tiny nudge if the sound engineer has
a mind to .....




........ Phil

"Phil Allison" wrote in message
...

"colin"

I thought you were going to move those 2.2k too.
that poor op amp is sitting there now with its supply all exposed.


** TL071 op-amps do have pretty a fair PSSR - you know.

Yes thats true for DC, the gain of that stage isnt that high either,
but its usualy not shown what the PSRR is for higher frequency,
it usually deteriotes and can be quite bad at higher frequencies.

Colin =^.^=

Phil Allison wrote:

" Eesyore"

** The noise ( at 1.36 nV rt/Hz ) is just a little worse than for the
Project 66 schematic.

Note: that figure is for a shorted input rather than a 200 ohms source.

Current in the transistors is the same, but there is one resistor change
that adds a pinch.

The 33 pF caps produce an earlier HF roll off than project 66 does.

-3dB @ 32 kHz in fact. A bit low for my liking for sure. That'll be
intrusive at 20kHz too..


** In a recording or PA system there are very many amplification stages,
each one with a LF and HF roll off curve, and all of them ACCUMULATE !!

If the overall aim is a flat response +/- 1 dB from 25Hz to 20 kHz -
then each stage needs to be a whole lot flatter that that aim.

However, given the input device is a dynamic microphone, few of which have
good response beyond 15 kHz , a -1dB point at 16 kHz is not so bad.

The treble EQ pot can always be given a tiny nudge if the sound engineer has
a mind to .....

Given that live sound is far from the only application, I like to see no more
than -0.2dB from any individual stage @ 20kHz. As you say, they accumulate very
quickly.

Graham

colin wrote:

"Phil Allison" wrote in message
"colin"

I thought you were going to move those 2.2k too.
that poor op amp is sitting there now with its supply all exposed.

** TL071 op-amps do have pretty a fair PSSR - you know.

Yes thats true for DC, the gain of that stage isnt that high either,
but its usualy not shown what the PSRR is for higher frequency,
it usually deteriotes and can be quite bad at higher frequencies.

The supply rails they operate on are clean in typical usage. Over a decade ? ago
I removed all such RC decoupling from my clients' products since all they did
was add cost and create a manufacturing problem since the series Rs have to
mounted off the PCB to appease the safety guys.

Graham

"colin"

** TL071 op-amps do have pretty a fair PSSR - you know.

Yes thats true for DC, the gain of that stage isnt that high either,
but its usually not shown what the PSRR is for higher frequency,
it usually deteriorates and can be quite bad at higher frequencies.


** AFAIK, no PSRR curve is published for the TL071, but for the very
similar LF351 - the PSRR drops from over 100dB at DC to about 60 dB at 20
kHz.

Maybe a good thing Project 66 includes *whole audio band* DC supply
decoupling ( 10ohm + 470 uF ) for the benefit of both the input AND op-amp
stages of the mic pre.

The usual 7815 / 7915 ICs regs can be noisy, one or two V of audio band
noise is not that unusual. Even if the op-amp's PSRR ratio were only 60 dB
right cross the audio band, this noise level would be attenuated to 2 uV
times the noise gain of the stage - say 10 times. That makes it about 20 uV
rms at very most.

Compared to the residual output noise of the op-amp and input differential
stage even at low gain, it is not very significant.




....... Phil

Phil Allison wrote:

The usual 7815 / 7915 ICs regs can be noisy, one or two V of audio band
noise is not that unusual.

I think you mean millivolts.

Graham

I think you mean millivolts.


** Obvious to anyone who did not snip the entire context like


Graham Stevenson, ASD ****ED POMMY **** DID





........ Phil

Phil Allison wrote:

I think you mean millivolts.

** Obvious to anyone who did not snip the entire context like

How *you* snipped the context !

Damn you're silly sometimes.

Graham

I think you mean millivolts.


** Obvious to anyone who did not snip the entire context like


Graham Stevenson, ASD ****ED POMMY **** DID

"Eeysore the Total ASSHOLE "

The supply rails they operate on are clean in typical usage.


** ******** they are.


Over a decade ? ago
I removed all such RC decoupling from my clients' products since all they
did
was add cost and create a manufacturing problem since the series Rs have
to
mounted off the PCB to appease the safety guys.


** When there is no series resistor in the DC rail to a channel board -
any electro
used as a local bypass will nicely couple residual supply rail * noise* to
the signal
earth pattern of the PCB.

Can easily amount to hundreds of microvolts of wide band noise at critical
points
appering in series with the wanted audio signal.

What ASININE STUPIDITY !!!

Studiomaster ****e was alwasy FULL of dumb errors like that.

Guess who did that ???

Graham Stevenson.

Lard Arse, DJ Charlatan.





....... Phil

Phil Allison wrote:

"Eeysore the Total ASSHOLE "

The supply rails they operate on are clean in typical usage.

** ******** they are.

They are in my products.

Over a decade ? ago I removed all such RC decoupling from my clients'
products since all they did was add cost and create a manufacturing problem
since the series Rs have to mounted off the PCB to appease the safety guys.

** When there is no series resistor in the DC rail to a channel board -
any electro used as a local bypass will nicely couple residual supply rail *
noise* to
the signal earth pattern of the PCB.

Can easily amount to hundreds of microvolts of wide band noise at critical
points appering in series with the wanted audio signal.

I know that you IDIOT. That's why *THERE AREN'T ANY*. Didn't you even read what
I said ?

As for using them to decouple supply noise, that noise ends up as current in the
ground too. A little bit of effort paid to making the supply rails clean pays
off very nicely.


What ASININE STUPIDITY !!!

Studiomaster ****e was alwasy FULL of dumb errors like that.

Guess who did that ???

Graham Stevenson.

Lard Arse, DJ Charlatan.

IDIOT.

I said I removed the electrolytics. The whole damn lot of them. Lock stock and
barrel. There are local film caps near the op-amps for stability.

Graham

"Eeysore the Total ASSHOLE "


The supply rails they operate on are clean in typical usage.

** ******** they are.

They are in my products.


** Shame your remark was about " typical" products.

You context shifting ****WIT !!



Over a decade ? ago I removed all such RC decoupling from my clients'
products since all they did was add cost and create a manufacturing
problem
since the series Rs have to mounted off the PCB to appease the safety
guys.

** When there is no series resistor in the DC rail to a channel board -
any electro used as a local bypass will nicely couple residual supply
rail *
noise* to the signal earth pattern of the PCB.

Can easily amount to hundreds of microvolts of wide band noise at
critical
points appering in series with the wanted audio signal.


I know that you IDIOT. That's why *THERE AREN'T ANY*. Didn't you even read
what
I said ?


** Shame you did NOT say there were no bypass Cs at all.

You context shifting ****WIT !!




What ASININE STUPIDITY !!!

Studiomaster ****e was always FULL of dumb errors like that.

Guess who did that ???

Graham Stevenson.

Lard Arse, DJ Charlatan.

IDIOT.

I said I removed the electrolytics.


** Shame you did not - you LYING ****.




....... Phil

"Jim Thompson"


** Hmmm .....

Input noise is shown 1.136 nV across the audio band.

Multiply by 141 ( ie sq.rt 20,000 ) = 160 nV.

The self noise of a 200 ohm source = 240 nV.

Gives a total noise = 288 nV.

Just a bit more than " project 66 " tests at 270 uV.

---------------------------------------


The gain graph shows the figure is close to 59 dB.

Just one hour after Jim posed his Q about a gain equation

- I posted this in reply :


" ** The overall gain is: 6.82 times 4000 / Rg "


In this example, Rg = 27 + 4 = 31 ohms

So the predicted voltage gain = 880 times.

Makes the gain 58.9 dB.




....... Phil

Phil Allison wrote:

"Eeysore the Total ASSHOLE "

I said I removed the electrolytics.

** Shame you did not - you LYING ****.

Just read more carefully next time OK ? I happen to entirely agree with you wrt
sticking PSU noise currents into the ground connection. It's crazy to do so yet
so many ppl do.

Graham