Is my LT44 transformer suitable for audio (de)coupling?
Andy
ffect of an earth loop.
An old audio transformer I have come across is marked as "LT44" and I
that that LT44 is sometimes used to indicate a 1:1 (or 600-ohm:600-ohm)
transformer.
However when I measure the resistence of the tarnsformer I get 650-ohms
on the primary and 80-ohmns on the secondary which has a tap at 40-ohms.
That doesn't feel right.
A slip of paper with the transformer says LT.44 transistor driver
transformer and says "impediance ratio is 20K ohm to 1K ohm C.T.".
(1) Why is my measurement of DC resistemce so very different from the
impendence on the specifications?
(2) Is this LT44 useless as an audio transformer? I seem to remember
something about number of turns in the windings being important for an
audio coupling transformer as well as the impedences.
(3) Is it wrong to use "LT44" as an accepted shorthand for a 600-
ohm:600-ohm audio coupling transformer?
Arfa Daily
"Andy" <nom...@nomail.com> wrote in message
news:Xns97B054C7...@127.0.0.1...
DC resistance is a very different animal from the AC resistance measurement
called impedance, which is a complex interaction of many factors, including
DC resistance, capacitive and inductive reactance, and frequency. Its value
varies considerably with frequency, and may change suddenly up or down, as
resonance is approached. Just as an example, a TV antenna might have a
folded dipole with a centre impedance of 300 ohms at the frequency of
interest. However, if you put your DC ohm-meter across there, that
folded-round piece of 1/4 inch metal rod, is going to read just about short
circuit, not 300 ohms. Either side of the design frequency, its impedance
will not be 300 ohms.
The transformer that you have is not "useless as an audio transformer", it's
just not the type of audio transformer that you were hoping for. What you
have there is as it says on your piece of paper - a "driver" transformer. It
is an inter-stage coupling transformer to go between an audio driver stage,
and a pair of transistors in a push-pull output stage. It serves the dual
purpose of transforming the high output impedance of the driver stage to a
lower one more suitable for driving the bases of the output pair, and also
phase splitting due to the centre tap on the secondary. The 1 : 1 600-600
transformer that you are looking for is more usually known as an isolating
transformer. The effect of placing it in an audio line should be virtually
zero, except to isolate the signal grounds from one another, which I guess
is what you are wanting to do.
According to my catalogues " LT44 " is exactly what you have got - a 20k to
1k CT driver transformer, so the term LT44 cannot be taken as shorthand for
a 600-600 isolation transformer.
Arfa
Dave Plowman (News)
Andy <nom...@nomail.com> wrote:
> I want to use an audio transformer to see if it helps to reduce the
> ffect of an earth loop.
Correct way to do it.
> An old audio transformer I have come across is marked as "LT44" and I
> that that LT44 is sometimes used to indicate a 1:1 (or 600-ohm:600-ohm)
> transformer.
Not that I've heard of, but I'm from the UK.
> However when I measure the resistence of the tarnsformer I get 650-ohms
> on the primary and 80-ohmns on the secondary which has a tap at 40-ohms.
> That doesn't feel right.
DC measurements can give a guide to the turns ratio but not the impedance.
> A slip of paper with the transformer says LT.44 transistor driver
> transformer and says "impediance ratio is 20K ohm to 1K ohm C.T.".
Right. That's an inter stage driver much used on early portable radios to
drive the push pull output stage. There'd likely be a matching output one
for a low impedance speaker - although some used approx 40 ohm ones driven
direct.
> (1) Why is my measurement of DC resistemce so very different from the
> impendence on the specifications?
Have a read of a book on basic AC theory or do a Google on it.
> (2) Is this LT44 useless as an audio transformer? I seem to remember
> something about number of turns in the windings being important for an
> audio coupling transformer as well as the impedences.
No - it is an audio transformer, but will drop the level somewhat. This
may not matter in practice. If from a small transistor portable may not
have a very good spec.
> (3) Is it wrong to use "LT44" as an accepted shorthand for a 600-
> ohm:600-ohm audio coupling transformer?
600- 600 ohm transformers are used on balanced pro equipment. Or more
likely once were in valve days. Most balanced audio these days is low
impedance out high in - same as domestic.
Ground isolating transformers these days will be usually 10k in and out.
An easy solution is to buy a line level transformer from a good car audio
shop used to isolate the ground to an add on power amp. It will be stereo
too. In the UK they cost about 5 gbp so if things are normal $5 or less in
the US. Make sure it is a 1:1 though as some are designed to match off the
speaker outputs in the head unit.
--
*The severity of the itch is proportional to the reach *
Dave Plowman da...@davenoise.co.uk London SW
To e-mail, change noise into sound.
Phil Allison
"Dave Plowman (News)"
> Andy
>
>> An old audio transformer I have come across is marked as "LT44" and I
>> that that LT44 is sometimes used to indicate a 1:1 (or 600-ohm:600-ohm)
>> transformer.
>
> Not that I've heard of, but I'm from the UK.
** Maplin sell them ......
>> However when I measure the resistence of the tarnsformer I get 650-ohms
>> on the primary and 80-ohmns on the secondary which has a tap at 40-ohms.
>
>> That doesn't feel right.
>
> DC measurements can give a guide to the turns ratio but not the impedance.
** That is quite wrong.
With small to medium audio transformers, the rated impedance of a winding =
10 to 20 times the DC resistance.
This fact derives directly from the *efficiency* of such transformers being
around 80% to 90 %.
The LT44 complies.
........ Phil
I.F.
"Phil Allison" <phila...@tpg.com.au> wrote in message
news:4b6b4oF...@individual.net...
Maplin sell them (and the LT700 PP O/P transformer that goes with it), but I
think the actual manufacturer might possibly be Eagle Electronics.
Ban
It is probably from an old microphone amplifier.
I think it will be quite suitable for the test you want to do. It will also
work the other way round, connect the 20k to your input and the 1k side with
a 3.3k to 10k resistor in series to the output of your setup. The resistor
will attenuate the signal and thus compensate for the gain of the
transformer.
--
ciao Ban
Apricale, Italy
Andy
> 600- 600 ohm transformers are used on balanced pro equipment. Or
> more likely once were in valve days. Most balanced audio these days
> is low impedance out high in - same as domestic.
>
> Ground isolating transformers these days will be usually 10k in and
> out.
Wow, now that's got me confused!
I am not using old tube equipment and I'm not using pro stuff either.
Instead of 600-600 ohm, you seem to be saying that I need 10K-10K ohm.
Can you or someone else explain this a little please.
10K-10K seems very different from the 600-600 which I was about to get!
Arfa Daily
pro ' standard for audio transmission lines, and is usually, although not
always, a balanced line system. Mics with XLR connection systems are often
600 ohm matched, and balanced, although may be 10k if they don't contain a
600 ohm line matching transformer.
You don't say exactly what pieces of kit you are trying to isolate from one
another, but assuming it's a couple of items operating at ' standard ' line
levels - such as a CD player, or tuner, or cassette deck, feeding into the
corresponding input of an amplifier, or the aux in, then the likely
impedance both ends will be around 47k. Obviously, 600 ohms is not a
terribly good match to 47k, but it's probably about the best that you're
going to do with readily available
1 : 1 line matching transformers. I would suggest that you just give it a
try. There will almost certainly be some effect on the overall frequency
response of the system, but depending on what you listen to, or how critical
your ear is, this may not be an issue, and might possibly be largely
overcome-able by judicious use of the system tone and equalizer controls.
I have used 600 ohm 1 : 1 line matching transformers on several occasions,
to do what you are trying to accomplish, and have never had any real
problems with the finished audio.
Arfa
Arfa Daily
"Arfa Daily" <arfa....@ntlworld.com> wrote in message
news:yIy3g.12469$Vn....@newsfe2-gui.ntli.net...
line level inputs, as opposed to 47k for a phono input, but everything else
said, still applies.
Arfa
Rich Grise
Well, what exactly is it you're trying to accomplish? Your original post
was quite vague, and seems to have been lost in a sea of speculation -
What are you trying to connect to what?
And if you have a ground loop, the solution is to diagnose and repair
the loop itself, rather than apply band-aids.
What do you have now, and what are you trying to do?
Thanks,
Rich
Dave Plowman (News)
Arfa Daily <arfa....@ntlworld.com> wrote:
> Mics with XLR connection systems are often 600 ohm matched, and
> balanced, although may be 10k if they don't contain a 600 ohm line
> matching transformer.
In broadcasting the most common combination with balanced mics is an
approx 150-300 ohm output going into an about 1.2k input on the mixer.
Older STC ribbons were sometimes 30 ohms. Accurate 600 ohm matching only
really was used into very long lines. Like telephones. But most long lines
these days will be digital.
--
*Stable Relationships Are For Horses. *
Don Pearce
<da...@davenoise.co.uk> wrote:
>In article <yIy3g.12469$Vn....@newsfe2-gui.ntli.net>,
> Arfa Daily <arfa....@ntlworld.com> wrote:
>> Mics with XLR connection systems are often 600 ohm matched, and
>> balanced, although may be 10k if they don't contain a 600 ohm line
>> matching transformer.
>
>In broadcasting the most common combination with balanced mics is an
>approx 150-300 ohm output going into an about 1.2k input on the mixer.
>Older STC ribbons were sometimes 30 ohms. Accurate 600 ohm matching only
>really was used into very long lines. Like telephones. But most long lines
>these days will be digital.
Shouldn't that be 12k in? A pair of 6.2k ohms in series feeding the
phantom power would give you that. Even with phantom switched off, the
resistors are still there, so the input impedance doesn't get any
higher for dynamics.
d
--
Pearce Consulting
http://www.pearce.uk.com
Phil Allison
"Don Pearce"
"Dave Plowman (News)"
>>
>>In broadcasting the most common combination with balanced mics is an
>>approx 150-300 ohm output going into an about 1.2k input on the mixer.
>
> Shouldn't that be 12k in?
** Silly guessing what can be found easily by inspection of makers specs
and schematics.
> A pair of 6.2k ohms in series feeding the
> phantom power would give you that.
** Yawn - so what ????
........ Phil
Dave Plowman (News)
Don Pearce <don...@pearce.uk.com> wrote:
> >In broadcasting the most common combination with balanced mics is an
> >approx 150-300 ohm output going into an about 1.2k input on the mixer.
> >Older STC ribbons were sometimes 30 ohms. Accurate 600 ohm matching
> >only really was used into very long lines. Like telephones. But most
> >long lines these days will be digital.
> Shouldn't that be 12k in? A pair of 6.2k ohms in series feeding the
> phantom power would give you that. Even with phantom switched off, the
> resistors are still there, so the input impedance doesn't get any
> higher for dynamics.
No - the phantom 'stand off' resistors don't go in series with the signal
path. They are from the plus side of the phantom supply to each leg of the
balanced input - very simply.
--
*Why isn't there a special name for the back of your knee?
Dave Plowman (News)
Phil Allison <phila...@tpg.com.au> wrote:
> > DC measurements can give a guide to the turns ratio but not the
> > impedance.
> ** That is quite wrong.
> With small to medium audio transformers, the rated impedance of a
> winding = 10 to 20 times the DC resistance.
> This fact derives directly from the *efficiency* of such transformers
> being around 80% to 90 %.
> The LT44 complies.
It might well do. But it's not a rule of thumb I'd rely on for all
transformers.
--
*If a thing is worth doing, wouldn't it have been done already?
Don Pearce
<da...@davenoise.co.uk> wrote:
>In article <44513df9...@news.plus.net>,
> Don Pearce <don...@pearce.uk.com> wrote:
>> >In broadcasting the most common combination with balanced mics is an
>> >approx 150-300 ohm output going into an about 1.2k input on the mixer.
>> >Older STC ribbons were sometimes 30 ohms. Accurate 600 ohm matching
>> >only really was used into very long lines. Like telephones. But most
>> >long lines these days will be digital.
>
>> Shouldn't that be 12k in? A pair of 6.2k ohms in series feeding the
>> phantom power would give you that. Even with phantom switched off, the
>> resistors are still there, so the input impedance doesn't get any
>> higher for dynamics.
>
>No - the phantom 'stand off' resistors don't go in series with the signal
>path. They are from the plus side of the phantom supply to each leg of the
>balanced input - very simply.
That's right - but from the point of view of the balanced signal, they
are in series, from one side to the other.
Phil Allison
"Dave Plowman (News)"
> Phil Allison
>> > DC measurements can give a guide to the turns ratio but not the
>> > impedance.
>
>
>> ** That is quite wrong.
>
>> With small to medium audio transformers, the rated impedance of a
>> winding = 10 to 20 times the DC resistance.
>
>> This fact derives directly from the *efficiency* of such transformers
>> being around 80% to 90 %.
>
>> The LT44 complies.
>
> It might well do. But it's not a rule of thumb I'd rely on for all
> transformers.
** Silly answer.
Contrary to your false claim above - it is a quite reliable guide.
....... Phil
Phil Allison
"Dave Plowman (News)"
> No - the phantom 'stand off' resistors don't go in series with the signal
> path.
** No-one said they did.
They *are* in series across the mic pre's input.
> They are from the plus side of the phantom supply to each leg of the
> balanced input - very simply.
** Very simple thoughts are your forte.
........ Phil
Arny Krueger
> I want to use an audio transformer to see if it helps to
> reduce the ffect of an earth loop.
>
> An old audio transformer I have come across is marked as
> "LT44" and I that that LT44 is sometimes used to indicate
> a 1:1 (or 600-ohm:600-ohm) transformer.
>
> However when I measure the resistence of the tarnsformer
> secondary which has a tap at 40-ohms.
>
> That doesn't feel right.
>
> A slip of paper with the transformer says LT.44
> transistor driver transformer and says "impediance ratio
> is 20K ohm to 1K ohm C.T.".
One thing is sure - this is a transformer with a turns ratio of something
like 4 to 1 to 8:1.
The secondary has a center tap. It might have been designed to be used as a
driver transformer in a transistor push-pull amplifier. I think these were
used in transistor radios in the 1960s.
The resistances have a ratio of about 8:1, but the specs of 20K to 1K
suggest a ratio of 4.5:1 since impedance ratio relates to the square of the
turns ratio.
> (1) Why is my measurement of DC resistemce so very
> different from the impendence on the specifications?
It looks like you got the specifications wrong. No way is this a 1:1
transformer.
> (2) Is this LT44 useless as an audio transformer? I seem
> to remember something about number of turns in the
> windings being important for an audio coupling
> transformer as well as the impedences.
Whether the transformer is good for audio depends on facts not in evidence,
such as power handling levels and frequency response. However, what you say
does not suggest good audio performance.
> (3) Is it wrong to use "LT44" as an accepted shorthand
> for a 600- ohm:600-ohm audio coupling transformer?
Yes, this is no way a 1:1 or 600:600 transformer.
Arny Krueger
> I want to use an audio transformer to see if it helps to
> reduce the ffect of an earth loop.
Go straight to Radio Shack and get their "Ground Isolator".
I've tested them on the bench, and they aren't all that bad.
I just did some bench tests of Radio Shack's "Ground Isolator" 270-054 using
test signals that maxed out around 2.5 v RMS. ZSource = 150 ohms, ZLoad =
5Kohms.
The measured performance was truely amazing for a pair of transformers case
and cables selling for only $16.65.
All IM, THD, and noise artifacts were at least 80 dB down with most in
the -100 dB range or better. Frequency response showed a 2 dB peak at 20 Hz
and then 10 dB down at 10 Hz. There was a 3 dB peak at about 51 KHz falling
to about 10 dB down around 100 KHz. +0.5 dB at 20 KHz.
I repeated the tests with the secondary loaded with 1.5K, and the peak at 51
Khz became well-damped with only about 0.6 dB rise.
Bob
Arfa Daily wrote:
> "Andy" <nom...@nomail.com> wrote in message
> news:Xns97B0E0C9...@127.0.0.1...
> > On 25 Apr 2006, Dave Plowman (News)<da...@davenoise.co.uk> wrote:
> >
> >> 600- 600 ohm transformers are used on balanced pro equipment. Or
> >> more likely once were in valve days. Most balanced audio these days
> >> is low impedance out high in - same as domestic.
> >>
> >> Ground isolating transformers these days will be usually 10k in and
> >> out.
>
> assuming it's a couple of items operating at ' standard ' line
> levels - such as a CD player, or tuner, or cassette deck, feeding into the
> corresponding input of an amplifier, or the aux in, then the likely
> impedance both ends will be around 47k. Obviously, 600 ohms is not a
> terribly good match to 47k,
As others have pointed out audio distribution in modern sound systems
is not matched for maximum power transfer.
> I have used 600 ohm 1 : 1 line matching transformers on several occasions,
> to do what you are trying to accomplish, and have never had any real
> problems with the finished audio.
A 1:1 transformer is not doing matching. They just help with common
mode noise and ground loops.
Bob
Dave Plowman (News)
Phil Allison <phila...@tpg.com.au> wrote:
> > No - the phantom 'stand off' resistors don't go in series with the signal
> > path.
> ** No-one said they did.
> They *are* in series across the mic pre's input.
> > They are from the plus side of the phantom supply to each leg of the
> > balanced input - very simply.
> ** Very simple thoughts are your forte.
I'm not going to bother with you anymore. You're an ignorant rude prat.
--
*It is easier to get older than it is to get wiser.
Dave Plowman (News)
Don Pearce <don...@pearce.uk.com> wrote:
> >No - the phantom 'stand off' resistors don't go in series with the
> >signal path. They are from the plus side of the phantom supply to each
> >leg of the balanced input - very simply.
> That's right - but from the point of view of the balanced signal, they
> are in series, from one side to the other.
True - but this would just effectively reduce the impedance of the circuit.
--
*If you can read this, thank a teecher
GregS
>"Andy" <nom...@nomail.com> wrote in message
>news:Xns97B054C7...@127.0.0.1
>
>> I want to use an audio transformer to see if it helps to
>> reduce the ffect of an earth loop.
>
>
>Go straight to Radio Shack and get their "Ground Isolator".
>
>I've tested them on the bench, and they aren't all that bad.
I in the past have been using their 600 ohm 1:1 transformer
to make such a device, with much success. I still want
to measure it. Buying two transformers, a box, and wiring,
is much more complicated than purchashing this premade "Ground Isolator"
I do have reservations of having the unit attenuate the signal
voltage.
greg
Phil Allison
"Dave Plowman (News)"
>
>> ** Very simple thoughts are your forte.
>
> I'm not going to bother with you anymore.
** Is that a promise ?
How about not bothering SEB too ?
> You're an ignorant rude prat.
** ROTFL !
So says another know nothing, pommy wanker.
....... Phil
Arny Krueger
news:e2nut1$2i7$1...@usenet01.srv.cis.pitt.edu
There are no doubt some losses.
Primary DCR = 136 ohms
Secondary DCR = 175 ohms
Turns ratio probably follows this pattern, IOW a slight step up perhaps to
minimize apparent loss.
I would guess its working impedance at around 3 K ohms. This is the range
where the resonance is nicely damped and the signal handling ability allows
for good dynamic range.
Jasen Betts
sound the two resistores in series are parallel with the input to me
Bye.
Jasen
Don Pearce
wrote:
The resistors are in series with each other, that combination is in
parallel with the input. Because the resistors are 6.2k, the
combination applies 12.4k as the maximum possible input impedance of a
microphone preamp.
Dave Plowman (News)
Jasen Betts <ja...@free.net.nz> wrote:
> >>No - the phantom 'stand off' resistors don't go in series with the
> >>signal path. They are from the plus side of the phantom supply to each
> >>leg of the balanced input - very simply.
> >
> > That's right - but from the point of view of the balanced signal, they
> > are in series, from one side to the other.
> sound the two resistores in series are parallel with the input to me
Correct.
But being a total of approx 13k will have little effect across 150 ohms.
--
*Microsoft broke Volkswagen's record: They only made 21.4 million bugs.
Don Pearce
<da...@davenoise.co.uk> wrote:
>In article <74bb.4451...@clunker.homenet>,
> Jasen Betts <ja...@free.net.nz> wrote:
>> >>No - the phantom 'stand off' resistors don't go in series with the
>> >>signal path. They are from the plus side of the phantom supply to each
>> >>leg of the balanced input - very simply.
>> >
>> > That's right - but from the point of view of the balanced signal, they
>> > are in series, from one side to the other.
>
>> sound the two resistores in series are parallel with the input to me
>
>Correct.
>
>But being a total of approx 13k will have little effect across 150 ohms.
True - I was just trying to correct your 1.2k, which while hardly a
typo was certainly a slip of the decimal point.
Phil Allison
"Don Pearce"
** Another slimy, brain dead pommy CUNT
> True - I was just trying to correct your 1.2k, which while hardly a
> typo was certainly a slip of the decimal point.
** It was no error at all -
YOU STINKING, FUCKING POMMY MORON
........ Phil
Dave Plowman (News)
Don Pearce <don...@pearce.uk.com> wrote:
> >But being a total of approx 13k will have little effect across 150 ohms.
> True - I was just trying to correct your 1.2k, which while hardly a
> typo was certainly a slip of the decimal point.
No - that's the input impedance of a Neve desk - one of the classic
designs. Others too. More modern ones may be higher.
--
*If a pig loses its voice, is it disgruntled?
Don Pearce
<da...@davenoise.co.uk> wrote:
>In article <44530f3d...@text.usenet.plus.net>,
> Don Pearce <don...@pearce.uk.com> wrote:
>> >But being a total of approx 13k will have little effect across 150 ohms.
>
>> True - I was just trying to correct your 1.2k, which while hardly a
>> typo was certainly a slip of the decimal point.
>
>No - that's the input impedance of a Neve desk - one of the classic
>designs. Others too. More modern ones may be higher.
Never knew that - thanks. I wonder why?
Don Pearce
<da...@davenoise.co.uk> wrote:
>In article <44530f3d...@text.usenet.plus.net>,
> Don Pearce <don...@pearce.uk.com> wrote:
>> >But being a total of approx 13k will have little effect across 150 ohms.
>
>> True - I was just trying to correct your 1.2k, which while hardly a
>> typo was certainly a slip of the decimal point.
>
>No - that's the input impedance of a Neve desk - one of the classic
>designs. Others too. More modern ones may be higher.
The current Neve mic amp (5012) has a 10k input impedance.
Don Pearce
<da...@davenoise.co.uk> wrote:
>In article <44530f3d...@text.usenet.plus.net>,
> Don Pearce <don...@pearce.uk.com> wrote:
>> >But being a total of approx 13k will have little effect across 150 ohms.
>
>> True - I was just trying to correct your 1.2k, which while hardly a
>> typo was certainly a slip of the decimal point.
>
>No - that's the input impedance of a Neve desk - one of the classic
>designs. Others too. More modern ones may be higher.
One more thing - the Neve mic pre has a pretty poor noise performance.
At -128dBu equivalent at the input, that is about 6dB above pure
thermal noise. That is 4 or 5 dB more noise than they should be
achieving.
Matti Adolfsen
> The resistors are in series with each other, that combination is in
> parallel with the input. Because the resistors are 6.2k, the
> combination applies 12.4k as the maximum possible input impedance of a
> microphone preamp.
>
In a professional microphone input you have in parallel:
- Phantom supply (2 x 6.2 k 0.5% resistors to +48V)
- switchable symmetric 20 dB pad (about 2 k input impedance to 200 ohms
output)
- and the microphone amplifier
To get the best possible noise figure from a mic preamp, you must match
the amp input impedance to the microphone. This is 600 ohms in most
cases. but:
if you are going to split one microphone to several mixers (FOH,
monitors, recording), there parallelled impedance of these mic preamps
should be higher than 600 ohms. In practice, many preamps are designed
to 2 k impedance, some monitor amps are as high as 4 k (the loss in
noise figure does not matter much on a noisy stage anyway..)
Hope this helps
P.S. back to the original question: several consumer units are not happy
when loaded with 600 ohms transformers. These output impedance of a CD
player might vary between 30 ohms to several k ohms, usuall having
serial capacitors in the output: this means high impedance at low
frequencies. A 10 k transformer might have better chance not to cut the
bass.
Don Pearce
<matti.adol...@pp.inet.fi> wrote:
>Don Pearce wrote:
>
>> The resistors are in series with each other, that combination is in
>> parallel with the input. Because the resistors are 6.2k, the
>> combination applies 12.4k as the maximum possible input impedance of a
>> microphone preamp.
>>
>
>In a professional microphone input you have in parallel:
>- Phantom supply (2 x 6.2 k 0.5% resistors to +48V)
>- switchable symmetric 20 dB pad (about 2 k input impedance to 200 ohms
>output)
>- and the microphone amplifier
>
As far as the DC path is concerned, you are right. The two resistors
are in parallel. But as seen by the microphone signal on the balanced
pair, they are in series.
Once you are in a position to need a 20dB pad, the impedance is no
longer important, because internal noise is no longer an issue.
>To get the best possible noise figure from a mic preamp, you must match
>the amp input impedance to the microphone. This is 600 ohms in most
>cases. but:
>
Microphones tend to be somewhere around 100 ohms in most cases. And to
get best noise performance, you don't match impedance. You make sure
the impedance of the microphone gives the best balance between current
noise and voltage noise from the amplifier. The amplifier will be
designed to put this point somewhere around the 100 ohms you expect
from a mic.
600 ohms is a figure used for line level connections.
>if you are going to split one microphone to several mixers (FOH,
>monitors, recording), there parallelled impedance of these mic preamps
>should be higher than 600 ohms. In practice, many preamps are designed
>to 2 k impedance, some monitor amps are as high as 4 k (the loss in
>noise figure does not matter much on a noisy stage anyway..)
>
If you want to split one mic the best way, run it into a single
preamp, then make the split after that, so that the best noise
performance is maintained.
>Hope this helps
>
>P.S. back to the original question: several consumer units are not happy
>when loaded with 600 ohms transformers. These output impedance of a CD
>player might vary between 30 ohms to several k ohms, usuall having
>serial capacitors in the output: this means high impedance at low
>frequencies. A 10 k transformer might have better chance not to cut the
>bass.
Transformers don't have "an impedance" as such. They transform the
impedance of whatever they are connected to by the square of the turns
ratio. So what impedance is presented as a result of putting in a
transformer depends on what the impedance of the equipment is.
Phil Allison
"Don Pearce"
>
> One more thing - the Neve mic pre has a pretty poor noise performance.
> At -128dBu equivalent at the input, that is about 6dB above pure
> thermal noise.
** Absolute bullshit.
Thermal noise in a 20kHz BW and 150 ohms
= 0.22 uV or -130.9 dBu.
>That is 4 or 5 dB more noise than they should be
> achieving.
** Wrong, it is a little over 2 dB.
Get you calcs right - dickhead.
( You forgot to allow for the 0.775 volts bit. )
BTW
This is what Rupert has to say on the matter:
http://www.rupertneve.com/notes_mic.html
http://www.rupertneve.com/notes_noise.html
NOTE
his comments re 1000 to 1200 ohms input load being "normal".
.......... Phil
Phil Allison
"Matti Adolfsen"
>
> To get the best possible noise figure from a mic preamp, you must match
> the amp input impedance to the microphone.
** Absolutely FALSE !!!
The best signal to noise ratio is obtained when the mic impedance is MUCH
less that the load impedance. This allows the mic to generate the greatest
signal voltage at the input while noise is a function of actual mic
impedance in the audio band.
> This is 600 ohms in most cases.
** Crapology.
The vast majority of *dynamic* mics are of 250 ohms or less impedance.
Condenser mics generate their own noise - far more than the vast majority
of mic pres do - so it rarely matters what impedance they are, far as
noise performance goes.
........ Phil
Dave Plowman (News)
Don Pearce <don...@pearce.uk.com> wrote:
> >No - that's the input impedance of a Neve desk - one of the classic
> >designs. Others too. More modern ones may be higher.
> Never knew that - thanks. I wonder why?
They had input balance to unbalance transformers. So you might as well get
the best practical matching? I suspect later units with electronic
balancing are more like your 10k or so.
--
*Someday, we'll look back on this, laugh nervously and change the subject
Dave Plowman (News)
Matti Adolfsen <matti.adol...@pp.inet.fi> wrote:
> To get the best possible noise figure from a mic preamp, you must match
> the amp input impedance to the microphone. This is 600 ohms in most
> cases. but:
Could you name a modern mic with an actual 600 ohm (balanced) output
impedance?
Or indeed any?
--
*It's o.k. to laugh during sexŚ.Ś.just don't point!
Ian Bell
It is a balance between minimising the load on the the mic and keeping noise
under control. The optimum noise performance of the original descrete class
A Neve mic pres was about 4.8K. A 2:1 transformer gives you 6dB of noise
free gain and an input impedance of 1.2K plus all the other benefits of
truly balanced floating inputs.
Ian
Jasen Betts
>>To get the best possible noise figure from a mic preamp, you must match
>>the amp input impedance to the microphone. This is 600 ohms in most
>>cases. but:
>>
> Microphones tend to be somewhere around 100 ohms in most cases. And to
> get best noise performance, you don't match impedance. You make sure
> the impedance of the microphone gives the best balance between current
> noise and voltage noise from the amplifier. The amplifier will be
> designed to put this point somewhere around the 100 ohms you expect
> from a mic.
>
> 600 ohms is a figure used for line level connections.
I own two dynamic microphones (consumer grade ones with unbalanced outputs
an 3.5mm plugs) both are 600 ohms.
the few pro or semi-pro mics I've handled were also labeled 600 ohms.
> Transformers don't have "an impedance" as such. They transform the
> impedance of whatever they are connected to by the square of the turns
> ratio. So what impedance is presented as a result of putting in a
> transformer depends on what the impedance of the equipment is.
as long as the mutual inductance is above the impedance of the source and sink
yes - a VHF balun won't make a good audio DI unit.
Bye.
Jasen
David
"Jasen Betts" <ja...@free.net.nz> wrote in message
news:50f7.4453...@clunker.homenet...
Hello Jasen,
Excuse me for butting in, but there is a difference between the advertised
impedance and the actual impedance.
If you measure the AC microphone impedance, you will see it is usually lower
that the advertised value and also varies across frequency.
Further, unbalanced microphones are typically higher in impedance than
balanced. 600 in the norm for unbalanced and 200 for balanced.
I good rule of thumb used for years is the preamp should present a load of
at least 5x the rated microphone impedance. This rule prevents frequency
coloration of the signal due to the frequency dependence of the actual
source impedance. Dynamic microphones have the most variation over
frequency, and capacitor or electret microphones, the least.
A nit on your statement quoted above, I assume you meant 'magnetizing
inductance' rather than 'mutual inductance'..
David
Dave Plowman (News)
Jasen Betts <ja...@free.net.nz> wrote:
> the few pro or semi-pro mics I've handled were also labeled 600 ohms.
Hmm. If you go to the AKG or Sennheiser sites and check out the specs
there most are around 150 ohms with a recommended input impedance of
greater than 1k.
--
*Some days we are the flies; some days we are the windscreen.*
Matti Adolfsen
> "Matti Adolfsen"
>
>> To get the best possible noise figure from a mic preamp, you must match
>> the amp input impedance to the microphone.
>
>
> ** Absolutely FALSE !!!
> The best signal to noise ratio is obtained when the mic impedance is MUCH
> less that the load impedance. This allows the mic to generate the greatest
> signal voltage at the input while noise is a function of actual mic
> impedance in the audio band.
Check your source information, there might be errors in your statement.
I'm talking about the thermal noise created in the resistive part of
the dynamic microphone. This is the metohd used by most european
broadcast companies to evaluate mic preamp performance. Best noise
figure can be obtained if the power matching is correct: source and load
impedances must be matched: Maximum power transfer from microphone to
preamp, and reduction of the noise source impedance to 50% of the
original. This theory is used in RF design too.
Btw, the best noise figure I've measured so far was from a TELEFUNKEN
V-76 tube amplifier. 0.5dB this preamp has high turns ratio input
transformer (about 1:50) feeding a pentode amplifier.
>> This is 600 ohms in most cases.
> ** Crapology.
Not so nice terminology, but this time correct information. Correct
impedance can be found from manufacturers websites and international
standards. I checked my sources:
- MIC input source impedance from 0 to 200 ohms (N10, page 6)
- MIC input source impedance for noise measrurements: 200 ohms (N10,
page 14)
- Nominal impedance of NEUMANN U-87: 200 ohms.
- Minimum load impedance for U-87 microphone: 1 k ohms (www.neumann.com).
Sources:
- Elements of infrared technology, page 236
- N10 Nordic technical recommendation 3rd edition, oct 1983
Further reading:
- Mathematical noise modeling and analysis of some popular preamplifier
circuit topologies. T.F. Darling, J.AES Vol 35, 1987
- Noise of sources, John Maxwell, National Semiconductor 1977
Matti
Phil Allison
"Matti Adolfsen"
>
>>> To get the best possible noise figure from a mic preamp, you must match
>>> the amp input impedance to the microphone.
>>
>>
>> ** Absolutely FALSE !!!
>
>> The best signal to noise ratio is obtained when the mic impedance is MUCH
>> less that the load impedance. This allows the mic to generate the
>> greatest signal voltage at the input while noise is a function of actual
>> mic impedance in the audio band.
>
>
> Check your source information,
** Check the source of your's - fuckwit.
It needs wiping immediately.
> I'm talking about the thermal noise created in the resistive part of the
> dynamic microphone.
** So am I.
But I actually know what I am talking about.
> Best noise figure can be obtained if the power matching is correct:
** Absolute BULLSHIT !!
> source and load impedances must be matched:
** Absolute BULLSHIT !!
> Maximum power transfer from microphone to preamp, and reduction of the
> noise source impedance to 50% of the original. This theory is used in RF
> design too.
** RF theory must not be applied to audio..
The world of audio uses "max voltage transfer" when linking sources to
loads.
You are a moronic ass.
PISS OFF.
......... Phil
Pooh Bear
"Dave Plowman (News)" wrote:
> In article <44530f3d...@text.usenet.plus.net>,
> Don Pearce <don...@pearce.uk.com> wrote:
> > >But being a total of approx 13k will have little effect across 150 ohms.
>
> > True - I was just trying to correct your 1.2k, which while hardly a
> > typo was certainly a slip of the decimal point.
>
> No - that's the input impedance of a Neve desk - one of the classic
> designs. Others too. More modern ones may be higher.
A typical modern console has around a 2k ohm impedance mic input.
'Rule of thumb' is to load with 10x source impedance.
Graham
Pooh Bear
"Dave Plowman (News)" wrote:
> In article <44530f3d...@text.usenet.plus.net>,
> Don Pearce <don...@pearce.uk.com> wrote:
> > >But being a total of approx 13k will have little effect across 150 ohms.
>
> > True - I was just trying to correct your 1.2k, which while hardly a
> > typo was certainly a slip of the decimal point.
>
> No - that's the input impedance of a Neve desk - one of the classic
> designs. Others too. More modern ones may be higher.
A typical modern console has around a 2k ohm impedance mic input.
Pooh Bear
Don Pearce wrote:
Please do your sums properly Don before making gaffes like that !
Graham
Pooh Bear
"Dave Plowman (News)" wrote:
> In article <NnE4g.57$h96...@read3.inet.fi>,
> Matti Adolfsen <matti.adol...@pp.inet.fi> wrote:
> > To get the best possible noise figure from a mic preamp, you must match
> > the amp input impedance to the microphone. This is 600 ohms in most
> > cases. but:
>
> Could you name a modern mic with an actual 600 ohm (balanced) output
> impedance?
>
> Or indeed any?
Loads of the inexpensive but half decent Asian mics are genuinely 600 ohms.
Graham
Phil Allison
"Poopie Bear"
> Don Pearce wrote:
>>
>> One more thing - the Neve mic pre has a pretty poor noise performance.
>> At -128dBu equivalent at the input, that is about 6dB above pure
>> thermal noise. That is 4 or 5 dB more noise than they should be
>> achieving.
>
> Please do your sums properly Don before making gaffes like that !
>
** How hysterical.
Dopey Drawers Pearce's worst gaffes ARE his erroneous sums.
Plus his INSANE insistence that RF and audio are the same !!!
.......... Phil
Don Pearce
Thank you! I did make a gaffe. The actual figure for the Neve noise
figure is about 3dB. That is still unforgivably poor for high end kit
- it is in fact no better than my little Behringer. Ten years ago I
was designing satellite receivers working up at 12GHz. The noise
figure I was working to was 0.3dB.
The last audio preamp I made had a noise figure of about 0.5dB,
because I was willing to use multiple parallel discrete transistors
for the input circuitry. Making it any better than this would have
been possible, but unwarranted because unlike the satellite receiver,
it wasn't pointing at a cold sky, but a warm microphone.
Phil Allison
"Don Pearce"
Poopie Bear
>
>> One more thing - the Neve mic pre has a pretty poor noise performance.
>>> At -128dBu equivalent at the input, that is about 6dB above pure
>>> thermal noise. That is 4 or 5 dB more noise than they should be
>>> achieving.
>>
>>Please do your sums properly Don before making gaffes like that !
>>
>
** A *fucking stupid* one.
Like hundreds of others and another one, right now !
> The actual figure for the Neve noise
> figure is about 3dB.
** The published curve shows it is typically less than 2 dB.
> That is still unforgivably poor for high end kit
> - it is in fact no better than my little Behringer.
** It is SFA additional noise in practice.
PLUS none at all when a typical condenser mic is used, as is the norm.
> Ten years ago I
> was designing satellite receivers working up at 12GHz.
** Who gives a rat's fuck ?
Go stick you irrelevant & erroneous RF shite up you fucking ARSE !
You are NOTHING but a trouble making, posturing, pommy
ARROGANT PIG - Don Pearce.
> The last audio preamp I made had a noise figure of about 0.5dB,
** Bet the ASININE FUCKWIT matched the source and load impedances.
Then miscalculated the **REAL** noise figure.
ROTFLMAO !!!
......... Phil
Dave Plowman (News)
Pooh Bear <rabbitsfriend...@hotmail.com> wrote:
> "Dave Plowman (News)" wrote:
> > In article <44530f3d...@text.usenet.plus.net>,
> > Don Pearce <don...@pearce.uk.com> wrote:
> > > >But being a total of approx 13k will have little effect across 150
> > > >ohms.
> >
> > > True - I was just trying to correct your 1.2k, which while hardly a
> > > typo was certainly a slip of the decimal point.
> >
> > No - that's the input impedance of a Neve desk - one of the classic
> > designs. Others too. More modern ones may be higher.
> A typical modern console has around a 2k ohm impedance mic input.
Thanks for that. I did have a quick glance at some specs for modern mixers
but those I looked at seemed shy of quoting the mic input impedance.
> 'Rule of thumb' is to load with 10x source impedance.
Yup.
> Graham
--
*To be intoxicated is to feel sophisticated, but not be able to say it.
Pooh Bear
Don Pearce wrote:
> On Mon, 01 May 2006 08:38:56 +0100, Pooh Bear
> <rabbitsfriend...@hotmail.com> wrote:
>
> >Don Pearce wrote:
> >
> >> On Fri, 28 Apr 2006 15:36:00 +0100, "Dave Plowman (News)"
> >> <da...@davenoise.co.uk> wrote:
> >>
> >> >In article <44530f3d...@text.usenet.plus.net>,
> >> > Don Pearce <don...@pearce.uk.com> wrote:
> >> >> >But being a total of approx 13k will have little effect across 150 ohms.
> >> >
> >> >> True - I was just trying to correct your 1.2k, which while hardly a
> >> >> typo was certainly a slip of the decimal point.
> >> >
> >> >No - that's the input impedance of a Neve desk - one of the classic
> >> >designs. Others too. More modern ones may be higher.
> >>
> >> One more thing - the Neve mic pre has a pretty poor noise performance.
> >> At -128dBu equivalent at the input, that is about 6dB above pure
> >> thermal noise. That is 4 or 5 dB more noise than they should be
> >> achieving.
> >
> >Please do your sums properly Don before making gaffes like that !
> >
> >Graham
>
> Thank you! I did make a gaffe.
No problem, we all goof up from time to time. ;-)
> The actual figure for the Neve noise
> figure is about 3dB. That is still unforgivably poor for high end kit
> - it is in fact no better than my little Behringer.
Indeed
> Ten years ago I
> was designing satellite receivers working up at 12GHz. The noise
> figure I was working to was 0.3dB.
>
> The last audio preamp I made had a noise figure of about 0.5dB,
> because I was willing to use multiple parallel discrete transistors
> for the input circuitry.
Care to name which ones you were using ?
> Making it any better than this would have
> been possible, but unwarranted because unlike the satellite receiver,
> it wasn't pointing at a cold sky, but a warm microphone.
Back in the days when I was at Neve, the then V series ( Mks 1 and 2 ) consoles (
and just about everything else except the digital console ) had a mic pre using a
step up transformer and a 5534. The quoted noise for that was a rather poor -126dBu
and it didn't actually measure any better either IIRC ! I was somewhat surprised to
say the least.
The recent mic pres I've done ( quite economy types ) manage about -128.5 - as long
as you factor in the extra little bit to account for the true noise equivalent
bandwidth of the measurement set : -3dB @ 22kHz 4th order is about 23kHz NEB.
Graham
Pooh Bear
"Dave Plowman (News)" wrote:
> In article <4455B9FB...@hotmail.com>,
> Pooh Bear <rabbitsfriend...@hotmail.com> wrote:
>
> > "Dave Plowman (News)" wrote:
>
> > > In article <44530f3d...@text.usenet.plus.net>,
> > > Don Pearce <don...@pearce.uk.com> wrote:
> > > > >But being a total of approx 13k will have little effect across 150
> > > > >ohms.
> > >
> > > > True - I was just trying to correct your 1.2k, which while hardly a
> > > > typo was certainly a slip of the decimal point.
> > >
> > > No - that's the input impedance of a Neve desk - one of the classic
> > > designs. Others too. More modern ones may be higher.
>
> > A typical modern console has around a 2k ohm impedance mic input.
>
> Thanks for that. I did have a quick glance at some specs for modern mixers
> but those I looked at seemed shy of quoting the mic input impedance.
>
> > 'Rule of thumb' is to load with 10x source impedance.
>
> Yup.
Actually, as a follow-up, following some discussion in r.a.p , my most recent
mic pre featured a modest increase in input Z to 2.5k. I note from the links
posted here that Rupert has gone as far as going to 10k now, which did
surprise me a little. I'd like to spend some time listening to the effect of
loading on various mics actually.
Graham
Don Pearce
<rabbitsfriend...@hotmail.com> wrote:
Yes - I have a box of old MAT-01s from PMI. They are strictly reserved
for such projects. I don't know if they are still available.
>> Making it any better than this would have
>> been possible, but unwarranted because unlike the satellite receiver,
>> it wasn't pointing at a cold sky, but a warm microphone.
>
>Back in the days when I was at Neve, the then V series ( Mks 1 and 2 ) consoles (
>and just about everything else except the digital console ) had a mic pre using a
>step up transformer and a 5534. The quoted noise for that was a rather poor -126dBu
>and it didn't actually measure any better either IIRC ! I was somewhat surprised to
>say the least.
>
The 5534 is not bad, but I wouldn't say it is the quietest way of
doing things. I had to make a very small preamp (just one op amp) for
a high impedance (50k) microphone. I searched for ages for quiet op
amp before I realised that an OP27 is optimized pretty well perfectly
at this impedance, with an excess noise of only about 1dB. Amazing!
>The recent mic pres I've done ( quite economy types ) manage about -128.5 - as long
>as you factor in the extra little bit to account for the true noise equivalent
>bandwidth of the measurement set : -3dB @ 22kHz 4th order is about 23kHz NEB.
>
>Graham
I really wish noise was expressed as a noise figure, rather than a
level. That way it wouldn't matter what impedance you were using, you
would simply have a figure of merit that told you how much worse the
pre was than theoretically perfect.
Phil Allison
"Poopie Fuckwit Bear"
>
> Actually, as a follow-up, following some discussion in r.a.p , my most
> recent
> mic pre featured a modest increase in input Z to 2.5k.
** Wank, wank ,wank ,wank, wank ,wank .....
> I note from the links
> posted here that Rupert has gone as far as going to 10k now, which did
> surprise me a little.
** The simple reason was given - you blind as a bat ass.
The mic input on that unit doubles as a balanced line input.
....... Phil
Pooh Bear
Don Pearce wrote:
I recall the beast.
Somewhere I think I have some of those similar Nat Semi parts that featured multiple
devices on-die. Forget the part number now. Oh no - I *was* right - the LM394 - just
checked in case. Though that would be an IC but the M just means monolithic. They're not
even insanely expensive now !
Also took a look at some esoteric fet data a while back. Noise somewhere down in the
500pV/sqrt Hz region. Interfet is the company.
> >> Making it any better than this would have
> >> been possible, but unwarranted because unlike the satellite receiver,
> >> it wasn't pointing at a cold sky, but a warm microphone.
> >
> >Back in the days when I was at Neve, the then V series ( Mks 1 and 2 ) consoles (
> >and just about everything else except the digital console ) had a mic pre using a
> >step up transformer and a 5534. The quoted noise for that was a rather poor -126dBu
> >and it didn't actually measure any better either IIRC ! I was somewhat surprised to
> >say the least.
>
> The 5534 is not bad, but I wouldn't say it is the quietest way of
> doing things.
Indeed not. They could have used something from AD or PMI and instantly improved the
noise figure.
> I had to make a very small preamp (just one op amp) for
> a high impedance (50k) microphone. I searched for ages for quiet op
> amp before I realised that an OP27 is optimized pretty well perfectly
> at this impedance, with an excess noise of only about 1dB. Amazing!
They're good op-amps. Never had the budget to design them into anything though.
:-(
> >The recent mic pres I've done ( quite economy types ) manage about -128.5 - as long
> >as you factor in the extra little bit to account for the true noise equivalent
> >bandwidth of the measurement set : -3dB @ 22kHz 4th order is about 23kHz NEB.
> >
> >Graham
>
> I really wish noise was expressed as a noise figure, rather than a
> level. That way it wouldn't matter what impedance you were using, you
> would simply have a figure of merit that told you how much worse the
> pre was than theoretically perfect.
Hmmm, I wonder how that would go down with those who 'cheat' by using a 150 ohm source
instead of 200. I note that Mackie ( I think ) is now quoting noise with the input
*shorted* too.
Graham
Don Pearce
<rabbitsfriend...@hotmail.com> wrote:
>> Yes - I have a box of old MAT-01s from PMI. They are strictly reserved
>> for such projects. I don't know if they are still available.
>
>I recall the beast.
>
>Somewhere I think I have some of those similar Nat Semi parts that featured multiple
>devices on-die. Forget the part number now. Oh no - I *was* right - the LM394 - just
>checked in case. Though that would be an IC but the M just means monolithic. They're not
>even insanely expensive now !
>
Yes that is very similar. The great thing about using these discretes
as front ends is that the knee frequency for 1/f noise is way lower
than the transistors in the average op-amp. So not only don't they
hiss - they don't rumble either. Three of them in parallel is about
right for a mic at about 150 to 200 ohms.
>Also took a look at some esoteric fet data a while back. Noise somewhere down in the
>500pV/sqrt Hz region. Interfet is the company.
>
But what about the current noise? You need to multiply that by the
source impedance to add in its effect. I've looked at a few FET input
op amps with amazingly low voltage noise, and this always more than
makes up the difference.
>>
>> I really wish noise was expressed as a noise figure, rather than a
>> level. That way it wouldn't matter what impedance you were using, you
>> would simply have a figure of merit that told you how much worse the
>> pre was than theoretically perfect.
>
>Hmmm, I wonder how that would go down with those who 'cheat' by using a 150 ohm source
>instead of 200. I note that Mackie ( I think ) is now quoting noise with the input
>*shorted* too.
>
Yup, you really do have to read specs with a cynical eye these days.
Dave Plowman (News)
Don Pearce <don...@pearce.uk.com> wrote:
> Thank you! I did make a gaffe. The actual figure for the Neve noise
> figure is about 3dB. That is still unforgivably poor for high end kit
> - it is in fact no better than my little Behringer.
Heh heh. But I'll guarantee which one sounds better...
--
*OK, who stopped payment on my reality check?
Phil Allison
"Don Pearce"
Poopie Bear
>
> Yes that is very similar. The great thing about using these discretes
> as front ends is that the knee frequency for 1/f noise is way lower
> than the transistors in the average op-amp. So not only don't they
> hiss - they don't rumble either. Three of them in parallel is about
> right for a mic at about 150 to 200 ohms.
** More totally asinine crapology from the Pommy RF Fool.
Audio band white noise is totally dominated by high frequency noise !!
Hence - there is no audible "rumble noise" from mic pres based on op-amps
or transistors.
> But what about the current noise?
** With a FET ?
With a 200 ohms source ??
What drugs is this FUCKING JERK taking ???
Or not taking ??
>>Hmmm, I wonder how that would go down with those who 'cheat' by using a
>>150 ohm source
>>instead of 200. I note that Mackie ( I think ) is now quoting noise with
>>the input
>>*shorted* too.
** Some condenser mics have very low output Zs - less than 20 ohms.
Not wrong to quite the noise for that case.
.......... Phil
Dave Plowman (News)
Don Pearce <don...@pearce.uk.com> wrote:
> Yes that is very similar. The great thing about using these discretes
> as front ends is that the knee frequency for 1/f noise is way lower
> than the transistors in the average op-amp. So not only don't they
> hiss - they don't rumble either. Three of them in parallel is about
> right for a mic at about 150 to 200 ohms.
Last time I played with this idea I found it very sensitive to RF
interference.
--
*Always borrow money from pessimists - they don't expect it back *
Don Pearce
<da...@davenoise.co.uk> wrote:
>In article <4455c4d1....@news.plus.net>,
> Don Pearce <don...@pearce.uk.com> wrote:
>> Thank you! I did make a gaffe. The actual figure for the Neve noise
>> figure is about 3dB. That is still unforgivably poor for high end kit
>> - it is in fact no better than my little Behringer.
>
>Heh heh. But I'll guarantee which one sounds better...
I wouldn't put money on that if I were you. I've measured the
performance of the Behringer, and I can find no fault with it. It has
problems, of course, but they are limitations imposed by cheap
implementation - certainly not in the sound department.
Phil Allison
"Dave Plowman (No-News)"
> Dumb Pommy Prick
>
>> Yes that is very similar. The great thing about using these discretes
>> as front ends is that the knee frequency for 1/f noise is way lower
>> than the transistors in the average op-amp. So not only don't they
>> hiss - they don't rumble either. Three of them in parallel is about
>> right for a mic at about 150 to 200 ohms.
>
> Last time I played with this idea I found it very sensitive to RF
> interference.
** Matching the source and load impedances RF style does that !!
You Bloody IDIOT !!
......... Phil
Don Pearce
<da...@davenoise.co.uk> wrote:
>In article <4456d74c....@news.plus.net>,
> Don Pearce <don...@pearce.uk.com> wrote:
>> Yes that is very similar. The great thing about using these discretes
>> as front ends is that the knee frequency for 1/f noise is way lower
>> than the transistors in the average op-amp. So not only don't they
>> hiss - they don't rumble either. Three of them in parallel is about
>> right for a mic at about 150 to 200 ohms.
>
>Last time I played with this idea I found it very sensitive to RF
>interference.
Well, there should have been nothing inherently RF sensitive about it.
Like any other sensitive system, it needs all the usual RF-proofing
tweaks to keep it sane.
Ian Bell
>
> One more thing - the Neve mic pre has a pretty poor noise performance.
> At -128dBu equivalent at the input, that is about 6dB above pure
> thermal noise. That is 4 or 5 dB more noise than they should be
> achieving.
>
Even if that were correct it is immaterial in most situations where the gain
is not set to maximum (and that is the ONLY place a mic amp achieves an
equivalent input figure this good). As gain is reduced, output noise begins
to dominate and the old Neve mixers still beat most 'modern' designs in
that respect.
Put simply, equivalent input noise is only part of the story.
ian
Ian Bell
>
> Thank you! I did make a gaffe. The actual figure for the Neve noise
> figure is about 3dB. That is still unforgivably poor for high end kit
> - it is in fact no better than my little Behringer. Ten years ago I
> was designing satellite receivers working up at 12GHz. The noise
> figure I was working to was 0.3dB.
>
> The last audio preamp I made had a noise figure of about 0.5dB,
> because I was willing to use multiple parallel discrete transistors
> for the input circuitry. Making it any better than this would have
> been possible, but unwarranted because unlike the satellite receiver,
> it wasn't pointing at a cold sky, but a warm microphone.
>
Not that is really matters. An improvement in noise figure of 2.5dB will
only improve the signal to noise at very high gains by the same amount.
Unless you are using dynamics on very quiet sources, the self noise of your
condensor mic is the limiting factor not the noise figure of the mic pre.
Ian
Michael A. Terrell
>
> Thank you! I did make a gaffe. The actual figure for the Neve noise
> figure is about 3dB. That is still unforgivably poor for high end kit
> - it is in fact no better than my little Behringer. Ten years ago I
> was designing satellite receivers working up at 12GHz. The noise
> figure I was working to was 0.3dB.
Its a lot easier to reduce the noise figure at a narrow bandwidth at
microwave frequencies.
--
Service to my country? Been there, Done that, and I've got my DD214 to
prove it.
Member of DAV #85.
Michael A. Terrell
Central Florida
Don Pearce
<mike.t...@earthlink.net> wrote:
>Don Pearce wrote:
>>
>> Thank you! I did make a gaffe. The actual figure for the Neve noise
>> figure is about 3dB. That is still unforgivably poor for high end kit
>> - it is in fact no better than my little Behringer. Ten years ago I
>> was designing satellite receivers working up at 12GHz. The noise
>> figure I was working to was 0.3dB.
>
>
> Its a lot easier to reduce the noise figure at a narrow bandwidth at
>microwave frequencies.
Narrow bandwidth? I don't call a bandwidth of nearly 2GHz narrow. And
asitappens, I can achieve these sorts of figure at audio as well.
Michael A. Terrell
2 GHz at 12 GHz is 16.67% which is narrow. Don't tell me that there
is no filtering at all. A circulator or isolator has a usable bandwidth
which keeps out of band noise out of the amp. One of my KU band
receivers is aboard the ISS.
Audio is true broadband, from DC to whatever the upper limit is set
at by the design and limitations of the components if it is DC coupled,
and from less than 100 Hz if its AC coupled.
Dave Plowman (News)
Don Pearce <don...@pearce.uk.com> wrote:
> Narrow bandwidth? I don't call a bandwidth of nearly 2GHz narrow.
Don't these things tend to be measured in octaves, as it were?
--
*A snooze button is a poor substitute for no alarm clock at all *
GregS
Not only that, but at low to DC freqs., the noise increases dramatically
of most devices. When looking at op-amps I have to often look at DC drifts,
which is the largest noise.
greg
Don Pearce
<mike.t...@earthlink.net> wrote:
>Don Pearce wrote:
>>
>> On Mon, 01 May 2006 13:24:48 GMT, "Michael A. Terrell"
>> <mike.t...@earthlink.net> wrote:
>>
>> >Don Pearce wrote:
>> >>
>> >> Thank you! I did make a gaffe. The actual figure for the Neve noise
>> >> figure is about 3dB. That is still unforgivably poor for high end kit
>> >> - it is in fact no better than my little Behringer. Ten years ago I
>> >> was designing satellite receivers working up at 12GHz. The noise
>> >> figure I was working to was 0.3dB.
>> >
>> >
>> > Its a lot easier to reduce the noise figure at a narrow bandwidth at
>> >microwave frequencies.
>>
>> Narrow bandwidth? I don't call a bandwidth of nearly 2GHz narrow. And
>> asitappens, I can achieve these sorts of figure at audio as well.
>>
>> d
>>
>> --
>> Pearce Consulting
>> http://www.pearce.uk.com
>
> 2 GHz at 12 GHz is 16.67% which is narrow. Don't tell me that there
>is no filtering at all. A circulator or isolator has a usable bandwidth
>which keeps out of band noise out of the amp. One of my KU band
>receivers is aboard the ISS.
>
Of course there is filtering. The waveguide coupler as a bandwidth, as
does the mixer. The image filter then has a bandwidth, and finally the
IF amplifier. Despite all this filtering - none of which is lossless -
the noise figure still happens. It really isn't easy even compared to
audio. Well done for having a receiver aboard ISS, but ground
transmitters tend not to be particularly power-limited, so the noise
figure probably isn't quite the issue it is with a terrestrial
receiver. In fact there would be no point because it is pointing at
the earth, which is hot.
> Audio is true broadband, from DC to whatever the upper limit is set
>at by the design and limitations of the components if it is DC coupled,
>and from less than 100 Hz if its AC coupled.
But achieving low noise at audio - however you want to describe its
bandwidth - is actually pretty trivial. The fact is that manufacturers
choose not to implement it because they can make something that is
"reasonable" for a lot less money. The business of achieving low noise
at audio doesn't rely on being able to juggle S parameters, physical
layout, board materials and mechanical design like microwave.
Don Pearce
<da...@davenoise.co.uk> wrote:
>In article <445c0c52....@news.plus.net>,
> Don Pearce <don...@pearce.uk.com> wrote:
>> Narrow bandwidth? I don't call a bandwidth of nearly 2GHz narrow.
>
>Don't these things tend to be measured in octaves, as it were?
Yes, but remember we are talking about scraping away at the last
vestiges of the state of the art here. Optimum noise match varies over
that range, and designing matching structures that keep the complex
impedance of the waveguide transition at the exact conjugate of the
input transistor noise match over such a bandwidth really is far from
simple. By comparison audio is a breeze - I promise.
Phil Allison
"GregS"
>>
>> Audio is true broadband, from DC to whatever the upper limit is set
>>at by the design and limitations of the components if it is DC coupled,
>>and from less than 100 Hz if its AC coupled.
>
> Not only that, but at low to DC freqs., the noise increases dramatically
> of most devices.
** All way outside the audio band and not even slightly relevant since noise
is proportional to the square root of the bandwidth.
The band from 0 to 20 Hz has a mere 20 Hz bandwidth !
The audio band is 20,000 Hz wide !!!!!!
So, some 31 times more noise energy audible as high frequency hiss.
> When looking at op-amps I have to often look at DC drifts,
> which is the largest noise.
** Drivel.
DC drift is due to temp variations.
It is separate from noise.
...... Phil
GregS
NO No NO. Look at it on a scope. Look at the charts.
greg
Phil Allison
"Dave Plowman (News)"
>
>> Narrow bandwidth? I don't call a bandwidth of nearly 2GHz narrow.
>
> Don't these things tend to be measured in octaves, as it were?
** Bandwidth in *electronics* is normally quoted in Hz, kHz, MHz, GHz etc.
Octaves are a musical notion, so not used much outside of audio and
acoustics.
....... Phil
Phil Allison
"GregS"
>>** Drivel.
>>
>>DC drift is due to temp variations.
>>
>>It is separate from noise
>
> NO No NO. Look at it on a scope. Look at the charts.
** Noise is noise - not drift.
You ignored all my other points so I guess you have no clue what the fuck
you are on about.
....... Phil
Don Pearce
Tell me about it. I've had to design baffling into a system before now
to control air currents that were causing cyclic temperature drifts.
Generally in an audio amp, though, there is sufficient DC feedback to
render this a non-problem.
GregS
If you want to isolate thermal drift, fine. I mostly talking
about visable noise fluctuating up and down over
seconds and tenths of seconds. Thats Noise.
greg
Phil Allison
"GregS" = A FUCKWIT
>>>>** Drivel.
>>>>
>>>>DC drift is due to temp variations.
>>>>
>>>>It is separate from noise
>>>
>>> NO No NO. Look at it on a scope. Look at the charts.
>>
>>
>>** Noise is noise - not drift.
>>
>>You ignored all my other points so I guess you have no clue what the fuck
>>you are on about.
>
> If you want to isolate thermal drift, fine. I mostly talking
> about visable noise fluctuating up and down over
> seconds and tenths of seconds. Thats Noise.
** ROTFLMAO.
Make up you mind, you dumb ASSHOLE !!
........ Phil
GregS
I made up my mind a long time ago.
greg
Michael A. Terrell
In our analog receivers the DC offset was measured, and a 18 bit ADC
was used to trim the offset to keep it to about a millivolt from
wherever it was set. Some external decoding required a fixed DC offset,
and this was available from the front panel. We also had a 63 dB range
in the output level control, digital with accurate .1 dB steps. This
was on the Microdyne 700 and 1620/1670 series telemetry receiving
equipment.
Michael A. Terrell
>
> Not only that, but at low to DC freqs., the noise increases dramatically
> of most devices. When looking at op-amps I have to often look at DC drifts,
> which is the largest noise.
>
> greg
Tell me about it. ;-) The diversity combiner took the linear 0 to 5
VDC AGC signals to combine the two video signals into a single signal
with a fixed output level as the received levels changed. The AGC
system in the combiner had to have under 1.5 mV error after the 14 op
amps used to set levels and feed them through the venerable MC1496 used
as an analog multiplier.
Don Pearce
Mine was the error amplifier in an ultra low noise RF synthesiser. The
low frequency drift that resulted from the air currents resulted in
phase noise from the oscillator in the .01 to 1Hz range. It was really
hard to fault find.
Michael A. Terrell
My point about microwave is that without filtering you have aditional
noise to deal with in the system. We offered IF bandwidths from 10 KHz
to 20 MHz, and video amps with matching bandwidths to keep as much noise
out of the received signal as possible.
The same receivers are used by NOAA for their weather sat downlinks.
With the improved doppler compensation they are able to lock onto the
sat before it comes over the hrizon five minutes faster than the old
Harris equipment it replaced.
Don Pearce
Ah! I see what you mean. Of course matching a receive filter to the
necessary bandwidth means that you neither lose much-needed signal,
nor let in unwanted noise. That goes for IF and baseband. As for RF,
presumably you have a range of input frequencies to deal with, so you
can't implement a channel-width filter there. But what you can do is a
roofing filter and image filters.
I presume your Doppler compensation was predictive, offsetting the
receiver by 17000MPH worth of frequency in anticipation of
acquisition.
GregS
I forget how this got into this, but I once built a 20 bit, well I don't know how many
bits, but I overlapped two 12 bit stages, and used discrets like a successive
approximation register, and op-amps and got at least 20 bit offset correction.
A/D D/A, but was interesting.
greg
Michael A. Terrell
A lot is still covered under several NDAs I had to sign, but the rest
of it was making the system as phase clean as frequency stable as
possible. Without that, the additional circuitry didn't stand a chance.
There are only a few things you can do to compensate for the doppler
effect, but that added to careful design and implementation works
wonders. I also worked on the both the fixed and mobile earth station
for Italy's space efforts. My title was production and engineering test
tech, but a lot of the engineers ask why I didn't have a degree due to
the way I not only found problems, but how to correct them. I got
bounced all over the Ocala facility, where ever the current hot spots
were. My boss told me that I wouldn't take "NO" for an answer. I
smiled and told him, "I don't take YES if I don't believe them. ;-)
Michael A. Terrell
I had to clean up an older design RF synthesizer because of multiple
engineering changes and parts going obsolete. The unit was inside a
double shield, so it was quite temperature stable, but the uncased
ceramic disk capacitor they originally used as standoffs was no longer
available. They switched to a good quality microwave ceramic capacitor,
and soldered them next to the large vias that were originally used to
solder the caps to the board. This caused the phase noise to go through
the roof. The fix was simple. Solder the holes closed, then solder the
cap where it was in the original design. The PLL was broken into three
band segments to reduce noise, and forward biased diodes were used to
short out part of the inductor to raise the frequency. The rest of the
phase noise was from the two RF output amps ringing due to a minor
change in the layout. I was working to use a Maxim IC to replace the
two amps when I was laid off in Sept. 2001. I can't remember the number
right now, but the two outputs were 180 out of phase to reduce the
current load variations. I also made some changes to the reference
frequency traps to clean up the DC loop control voltage.
Lostgallifreyan
news:4bm5oeF...@individual.net:
> Audio band white noise is totally dominated by high frequency noise !!
>
White noise has equal amounts of all frequencies. We hear the HF dominate
because higher frequencies have more energy. (And because our ears are more
sensitive to it).
GregS
The noise voltage of many op-amps is almost flat from 1K to 100K, but below 1 K
it moves upward. I'm looking at the LF353 as it has a very low level of low frequency
noise, or at least the chart shows that. The AD711 has 18 nvHz at 1 kHz but
60 nvHz at 1 Hz.
greg
Tom MacIntyre
An octave is a doubling of the frequency, so it has logarithmic
implications.
Tom
Tom MacIntyre
wrote:
Ears are most sensitive around 1k, aren't they?
Tom
Lostgallifreyan
news:sa0d52hs3jff6bohm...@4ax.com:
Probably. :) I wasn't thinking hard about that, I was just caught by the
claim that HF dominates white noise in the audio band, when white noise is
defined as being made up of all frequencies present with equal energy in
each.
Tom MacIntyre
wrote:
Hmmm...equal energy means that it would be more energy per octave at
higher frequencies, right?
Tom
Lostgallifreyan
Yes, but that comes down to how you define the scale. The only reason,
following from that, to say that HF dominates in the audio band, is because
of expressing a log scale as lin, as pitch as opposed to frequency. If you
say that the pitch notation is purely based on musical needs, and prefer
the frequency scale for electronics analysis, surely you also have to
discard with it the notion of HF dominance.
The reason the claim that HF dominates seems wrong to me, is I think a
peice of string is its own length, regardless of whether we measue in
inches or centimetres, let alone a nonlinear scale. White noise has a very
specific definition, in which energy is equal across the spectrum, does not
dominate part of it. It's out perception of pitch that does that.
Tom MacIntyre
wrote:
Interesting about the string, because we can simulate musical octaves
by folding a piece of string in half again and again.
Wouldn't a scale from, say, 100 Hz to 1600 Hz sound "interesting" if
we used 100 Hz as the semitone, rather than 2^(1/12) x f? :-)
Tom