Maker Pro
Maker Pro

ICs for professional audio applications

T

ted

Jan 1, 1970
0
Can somebody recommend ICs names or families for use as
programmable-gain high quality audio amplifiers? (as in high end
microphone or line amplifiers)

I am specifically looking for a very low noise op amp that can be gain
controlled from an external digital source, either in parallel form or
via I2C.

I am aware of the SSM2017/2019 device, which could be gain controlled
by the addition of external resistors and CMOS analogue gates. But I
would prefer an integrated solution..:)-(

I am also aware of the Crystal CS3310 (and the Wolfson equivalent) but
for various reasons, we cannot use these...:)-(

I am also aware of the "programmable resistors" manufactured by Xicor,
Analogue devices and Dallas. But it is not too clear from the spec
sheets what kind of noise of distortion these devices generate if used
in high end audio applications.

Has anybody used these for audio???

Or any other recommendations or suggestions anybody??

TIA

ted
 
I

Ian Bell

Jan 1, 1970
0
ted said:
Can somebody recommend ICs names or families for use as
programmable-gain high quality audio amplifiers? (as in high end
microphone or line amplifiers)

I am specifically looking for a very low noise op amp that can be gain
controlled from an external digital source, either in parallel form or
via I2C.

I am aware of the SSM2017/2019 device, which could be gain controlled
by the addition of external resistors and CMOS analogue gates. But I
would prefer an integrated solution..:)-(

I am also aware of the Crystal CS3310 (and the Wolfson equivalent) but
for various reasons, we cannot use these...:)-(

I am also aware of the "programmable resistors" manufactured by Xicor,
Analogue devices and Dallas. But it is not too clear from the spec
sheets what kind of noise of distortion these devices generate if used
in high end audio applications.

Has anybody used these for audio???

Or any other recommendations or suggestions anybody??

TIA

ted

NE5532. never been bettered.

Ian
 
W

Walter Harley

Jan 1, 1970
0
Ian Bell said:
ted said:
Can somebody recommend ICs names or families for use as
programmable-gain high quality audio amplifiers? (as in high end
microphone or line amplifiers)

I am specifically looking for a very low noise op amp that can be gain
controlled from an external digital source, either in parallel form or
via I2C. [...]
NE5532. never been bettered.

Unfortunately, although it's a lovely chip for certain applications, it does
not do what the OP was asking for.

Ted, you might try looking at the offerings on http://www.thatcorp.com .

FWIW, I don't think you're going to find that pro audio gear uses
gain-controlled chips in mic preamp applications, though I could be wrong.
But they are certainly common in line-level applications.

BTW, whenever you say "very low noise," certain alarm bells should go off in
your brain... that isn't really an adequate spec. There are devices that
have very low current noise, and there are devices that have very low
voltage noise, and there are devices that are pretty good (but not great) at
both. The right answer for a mic preamp may not be the same as the right
answer for a line level amp; it depends on source impedance and circuit
topology.
 
J

John Woodgate

Jan 1, 1970
0
When are the chip makers going to realise that the standard for Pro
Audio is +/- 15..18 volts - headroom is vital.

You don't understand; this TI 'professional audio' is the sort that goes
into PCs. It's FAR more sophisticated than your sort; witness the fact
that you can run a 1500 W surround-sound system from a wall-wart. No
headroom problems.
 
A

Allan Herriman

Jan 1, 1970
0
Nice part!
When are the chip makers going to realise that the standard for Pro
Audio is +/- 15..18 volts - headroom is vital.

This part seems to be able to handle signals of more than 1Vrms at the
input. How much more input-referred headroom would a microphone
amplifier need to be classed as "Pro Audio"?

Surely dynamic range rather than headroom (per se) is vital (?)

The typical performance graphs show an equivalent input noise of about
-128dBu for gains > 30dB with a 150ohm input termination.

I think that's a noise figure of about 2.8dB, which isn't too bad.

Regards,
Allan.
 
A

Allan Herriman

Jan 1, 1970
0
You don't understand; this TI 'professional audio' is the sort that goes
into PCs. It's FAR more sophisticated than your sort; witness the fact
that you can run a 1500 W surround-sound system from a wall-wart. No
headroom problems.

John, the datasheet says:

"Four programmable digital outputs are provided for controlling the
external switching of input pads, phantom power, high pass filters,
and polarity reversal functions."

It seems that TI marketing thinks it's meant to go into professional
equipment.

Regards,
Allan.
 
B

Ban

Jan 1, 1970
0
me said:
Nice part!
When are the chip makers going to realise that the standard for Pro
Audio is +/- 15..18 volts - headroom is vital.
cheers
Mike

So put a 5532 with 10dB of gain behind it. You can just lower the gain of
the preamp and do not need to amplify to +24dBm, with the A/D-converters
having anyway not that range.
I think 5Vrms is absolutly sufficient. The advantage is: a low voltage
process has lower consumption, noise and price.
 
J

John Woodgate

Jan 1, 1970
0
I read in sci.electronics.design that Allan Herriman <allan.herriman.hat
[email protected]> wrote (in <k4piuv87pj98qptsll6rqb15jkj5kiog
[email protected]>) about 'ICs for professional audio applications', on Wed, 24
Dec 2003:
John, the datasheet says:

"Four programmable digital outputs are provided for controlling the
external switching of input pads, phantom power, high pass filters,
and polarity reversal functions."

It seems that TI marketing thinks it's meant to go into professional
equipment.
Oh well then, they screwed up even worse that I gave them debit for.
(;-)
 
J

John Woodgate

Jan 1, 1970
0
I read in sci.electronics.design that Allan Herriman <allan.herriman.hat
[email protected]> wrote (in <enniuvgsj48s4sp94klhgv425e0qs9nl
[email protected]>) about 'ICs for professional audio applications', on Wed, 24
Dec 2003:
This part seems to be able to handle signals of more than 1Vrms at the
input. How much more input-referred headroom would a microphone
amplifier need to be classed as "Pro Audio"?

Even so, the +/-5 V supply requirement means two more regulators. And I
wonder how it fares with transients due to 48 V phantom power. It may
need a number of external protection components.
 
G

Geo

Jan 1, 1970
0
I am also aware of the "programmable resistors" manufactured by Xicor,
Analogue devices and Dallas. But it is not too clear from the spec
sheets what kind of noise of distortion these devices generate if used
in high end audio applications.

Has anybody used these for audio???

Tried the Dallas digital resistors but the "zipper" noise breakthrough is just
too much for professional use (application was level-setting).
The thatcorp chips (forget the numbers) worked ok using a processor to drive
d-a. They have an APP note 102 "Digital Gain Control with Analog VCAs"
Obviously not an "integrated" solution but the digital stuff should really be
stuck in a tin box with just the analogue level control emerging IMO.


Geo
 
S

Stefan Heinzmann

Jan 1, 1970
0
John said:
I read in sci.electronics.design that Allan Herriman <allan.herriman.hat
[email protected]> wrote (in <enniuvgsj48s4sp94klhgv425e0qs9nl
[email protected]>) about 'ICs for professional audio applications', on Wed, 24
Dec 2003:




Even so, the +/-5 V supply requirement means two more regulators. And I
wonder how it fares with transients due to 48 V phantom power. It may
need a number of external protection components.

I haven't seen any audio ADC or DAC that is directly compatible with +/-
15V either. The trend is to just have a +5V rail for analog and forget
about the -5V rail. Even smaller voltages are becoming common even for
professional products. The reason is simple: It is not economical to
come up with an IC fabrication process that is good for high speed
digital circuitry and withstands in excess of 30V on the analog side.
Remember, it all has to go onto the same chip, and the circuitry in ADCs
and DACs is predominantly digital, in particular for sigma-delta
converters. So the digital side is what you optimize for.

A similar argument can be made for digitally controlled gain amplifiers.
If you don't like this, you need a voltage controlled amplifier and a
separate control DAC. Separating this allows you to use separate IC
processes for the two functions.

Furthermore, what matters inside your circuit is dynamic range and
THD+N. The actual voltage levels are secondary. Absolute levels matter
at external interfaces, and there you have appropriate buffer amplifiers
to get your levels where you want them. In professional audio you need
them anyway because you deal with balanced interfaces. Inside your
circuit you will rarely want to be balanced.

Cheers
Stefan
 
A

Andy Peters

Jan 1, 1970
0
Nice part!
When are the chip makers going to realise that the standard for Pro
Audio is +/- 15..18 volts - headroom is vital.

The part is designed to talk directly to an ADC input. Since most
ADCs these days run off of +5V supplies, I guess they felt there was
no real need to run off of higher-voltage rails. It is a mic preamp,
after all. If one expects very hot input signals, do what all of the
pro audio manufacturers do and put a pad before the input. The thing
has four general-purpose output bits whose state is set by the serial
input word -- use it to control a pad.

--a
 
A

Andy Peters

Jan 1, 1970
0
John Woodgate said:
Even so, the +/-5 V supply requirement means two more regulators. And I
wonder how it fares with transients due to 48 V phantom power. It may
need a number of external protection components.

The data sheet does mention protection.

A preamp run off of +/-15V rails also needs protection from 48V phantom.

=-a
 
T

Tony Roe

Jan 1, 1970
0
Although the PGA2500 is designed to feed an ADC directly, there are a couple of
issues that confuse me about this connection:

Even though powered by only +/-5V, the pre-amp output can damage an ADC powered
from 0/5V, so I presume that either:
- the pre-amp outputs need to be biased to the ADC common mode voltage, with
clamp diodes to prevent signals <0V, or
- the signals need to be AC coupled, with clamp diodes to prevent both >5V and
<0V signals), or
- the signals need to be attenuated 2:1 wrt a high quality +5V reference.

An analog LPF is shown in all the apnotes I've seen on the balanced connection
(series R in each signal, and a cap between the 2 signals). It provides
anti-alias filtering, and it's also supposed to absorb the HF noise from the
digital parts of the ADC. But in the latter function it's quite inadequate due
to both the series resistors and the high opamp output impedance at 50MHz-ish
frequencies - to do that function needs a separate cap to ground on each signal.

Does anyone have any clues to enlighten me on these apparent anomalies? Or
preferred or alternative ways to fix them?

The part is designed to talk directly to an ADC input. Since most
ADCs these days run off of +5V supplies, I guess they felt there was
no real need to run off of higher-voltage rails. It is a mic preamp,
after all. If one expects very hot input signals, do what all of the
pro audio manufacturers do and put a pad before the input. The thing
has four general-purpose output bits whose state is set by the serial
input word -- use it to control a pad.

--a

Tony (remove the "_" to reply by email)
 
Top