Does anyone know if it is typical for an ADC to distort when driven near full
scale (0dB)?

I've got an Edirol UA5 USB audio interface. I put a 1600Hz sine wave in the
line input (played from the line out of a Creative Labs Nomad Jukebox 3),
turned the level up to just under 0dB and recorded a few seconds of audio.

When I looked at the spectrum (Adobe Audition 1.5!) I see several harmonics
(multiples of 1600Hz), and the largest one was only -40dB down or so. I
noticed these extra peaks went away once I lowered the input to say, -6dB.
So, it seems to exhibit some kind of clipping/overloading behavior, but the
input signal was not over. It was something like -1 or -2dB.

I looked at the circuit in this box. It is based on the AK4524 ADC. The
circuit seems to follow the standard design spec. The chip operates on +5V.
The driving opamps run on +/-5V. The input signal to the ADC seems to be
about 0.7V RMS at full scale.

So, could it be that these ADCs are really not very good? Or is the
supporting circuitry poor? Should I expect a more expensive box, like MOTU or
Presonus, to do it right? I guess it pays to check these things out.

Thanks for any help...
Richard

wrote in message ...

I've got an Edirol UA5 USB audio interface. I put a 1600Hz sine wave in
the
line input (played from the line out of a Creative Labs Nomad Jukebox 3),
turned the level up to just under 0dB and recorded a few seconds of audio.

When I looked at the spectrum (Adobe Audition 1.5!) I see several
harmonics
(multiples of 1600Hz), and the largest one was only -40dB down or so. I
noticed these extra peaks went away once I lowered the input to say, -6dB.
So, it seems to exhibit some kind of clipping/overloading behavior, but
the
input signal was not over. It was something like -1 or -2dB.

When you turned the signal down and the peaks went away, did you turn it
down using the controls on the Edirol, or on the Nomad Jukebox?

Peace,
Paul

wrote:
Does anyone know if it is typical for an ADC to distort when driven near full
scale (0dB)?

Yes, if they're crummy enough. And there are many that are indeed
crummy enough.

I've got an Edirol UA5 USB audio interface. I put a 1600Hz sine wave in the
line input (played from the line out of a Creative Labs Nomad Jukebox 3),
turned the level up to just under 0dB and recorded a few seconds of audio.

When I looked at the spectrum (Adobe Audition 1.5!) I see several harmonics
(multiples of 1600Hz), and the largest one was only -40dB down or so.

I woudln't be surprised if most, if not all of that came from your
Jukebox. The analog output stage isn't anything to write home about,
particularly at high levels - and you'd probably need to play back a
nearly full scale file at nearly maximum level in order to get enogh
signal to drive your Edirol interface to near full scale.

In other words, your test equipment may not be as good as the equipment
that you're testing. But there have been some reports that the Edirol
interfaces don't sound all that great.

Bottom line - how does it sound under normal working conditions? If you
hadn't tried to test it, would you be happy with its performance?

wrote:

Does anyone know if it is typical for an ADC to distort when driven
near full scale (0dB)?

Typical? C'mon, you know full well the answer to that is "no." You're
in denial. You bought a cheap box that totally doesn't suck for what it
is, but what it is ain't sooper dooper. Either buy a better box or
accept it and move on.



When I looked at the spectrum (Adobe Audition 1.5!) I see several
harmonics (multiples of 1600Hz), and the largest one was only -40dB
down or so. I noticed these extra peaks went away once I lowered the
input to say, -6dB.

Two words: harmonic distortion. The analog input stage is made to that
price point, and goes all to flooey with hot signals.

No biggie though. Just record at -6. With 90+dB of range to work with,
giving up 6 is not an issue. If you absolutely just have to gotta must
make it full scale, normalize when you're done.

--
"It CAN'T be too loud... some of the red lights aren't even on yet!"
- Lorin David Schultz
in the control room
making even bad news sound good

(Remove spamblock to reply)

"Paul Stamler" writes:

wrote in message ...

I've got an Edirol UA5 USB audio interface. I put a 1600Hz sine wave in
the
line input (played from the line out of a Creative Labs Nomad Jukebox 3),
turned the level up to just under 0dB and recorded a few seconds of audio.

When I looked at the spectrum (Adobe Audition 1.5!) I see several
harmonics
(multiples of 1600Hz), and the largest one was only -40dB down or so. I
noticed these extra peaks went away once I lowered the input to say, -6dB.
So, it seems to exhibit some kind of clipping/overloading behavior, but
the
input signal was not over. It was something like -1 or -2dB.

When you turned the signal down and the peaks went away, did you turn it
down using the controls on the Edirol, or on the Nomad Jukebox?

Peace,
Paul

I turned down the gain on the Edirol UA5 and the problem went away. So, I
believe it was the UA5 that was the problem. I also put the same input into a
Sony HiMD (NH800, recording PCM) and did not see any distortion!

Richard

"Mike Rivers" writes:

wrote:
Does anyone know if it is typical for an ADC to distort when driven near full
scale (0dB)?

Yes, if they're crummy enough. And there are many that are indeed
crummy enough.

I've got an Edirol UA5 USB audio interface. I put a 1600Hz sine wave in the
line input (played from the line out of a Creative Labs Nomad Jukebox 3),
turned the level up to just under 0dB and recorded a few seconds of audio.

When I looked at the spectrum (Adobe Audition 1.5!) I see several harmonics
(multiples of 1600Hz), and the largest one was only -40dB down or so.

I woudln't be surprised if most, if not all of that came from your
Jukebox. The analog output stage isn't anything to write home about,
particularly at high levels - and you'd probably need to play back a
nearly full scale file at nearly maximum level in order to get enogh
signal to drive your Edirol interface to near full scale.

In other words, your test equipment may not be as good as the equipment
that you're testing. But there have been some reports that the Edirol
interfaces don't sound all that great.

Bottom line - how does it sound under normal working conditions? If you
hadn't tried to test it, would you be happy with its performance?

As I said in a post before, I think it was the Edirol UA5. When I turned the
gain down on the Edirol, the problem went away. Also, I saw no problem with
the same signal into a Sony HiMD (NH800, PCM recording) with the gain turned
up to full scale.

So, it seems it is either a crummy ADC or a crummy gain stage on the Edirol
UA5. I'm just wondering what kind of money I have to spend to get something
that works.

Before testing this I was a little suspicious of the performance, but I can't say whether that was clarity of the signal or distortion at high levels.

Richard

wrote:

When I turned the
gain down on the Edirol, the problem went away.

So, it seems it is either a crummy ADC or a crummy gain stage on the Edirol
UA5. I'm just wondering what kind of money I have to spend to get something
that works.

Even the most expensive converters get a little funny within the last
half dB short of full scale. How far do you have to turn down the level
before it "works" for you? If you're happy running it with peaks
staying below -2 dBFS, just run it like that and save your money.
Consider that your maximum operating level. Sure, you'll throw away a
few dB of dynamic range, but unless you're recording at less than
16-bit resolution, you'll never miss it. Really.

Now, if it doesn't get clean until you drop the peak level to -10 dBFS
or so, then you probably should replace the Edirol. But if you're
looking for the last dB, probably something else will be better, but no
promises as to how much better. If you're not getting good results
under normal working conditions, I'd look for something else in the
same price/features range and put it to the test. You may wear out the
welcome mat at your dealer's buying and returning, but that's what
dealers are for. Your call as to what to do and how much you're willing
to spend.
..

wrote in message ...

Does anyone know if it is typical for an ADC to distort when driven near
full
scale (0dB)?

Yes.

I've got an Edirol UA5 USB audio interface. I put a 1600Hz sine wave in
the
line input (played from the line out of a Creative Labs Nomad Jukebox 3),
turned the level up to just under 0dB and recorded a few seconds of audio.

When I looked at the spectrum (Adobe Audition 1.5!) I see several
harmonics
(multiples of 1600Hz), and the largest one was only -40dB down or so. I
noticed these extra peaks went away once I lowered the input to say, -6dB.

Yes. That last dB or so before FS is a problematical area for many
converters. The better ones can be very clean until they are just a few
tenths of a dB below FS.

So, it seems to exhibit some kind of clipping/overloading behavior, but
the
input signal was not over. It was something like -1 or -2dB.

What you seem to be describing is some kind of "soft clipping" effect.

I looked at the circuit in this box. It is based on the AK4524 ADC. The
circuit seems to follow the standard design spec. The chip operates on
+5V.
The driving opamps run on +/-5V. The input signal to the ADC seems to be
about 0.7V RMS at full scale.

Shouldn't be a problem. With modern op amps, it should be possible to get a
clean 2.8 volts RMS out of a circuit running on +/- 5.


So, could it be that these ADCs are really not very good?

Typical of cheap circutry, or someone is going down the garden path to soft
clipping.

Or is the
supporting circuitry poor? Should I expect a more expensive box, like
MOTU or
Presonus, to do it right? I guess it pays to check these things out.

If you want to see an audio interface done *right* check out the LynxTWO.

If you look at the AES standards for measuring computer multimedia audio,
they specify (off the top of my head) that measurements be taken at 3 dB
below the point that results in 1% THD or obvious clipping.

http://www.aes.org/publications/standards/

AES17

Hi Richard,

If you look at the datasheet, you can see there is a PGA inside the 4524.
When set at 0db gain, the max input (P-P) is defined as .58 * Vref where
Vref is limited to the range of 3.0V to Va. To determine max input, you
need to measure the Vref pin to see if they use the 5 volt supply or a more
stable ref voltage. At 5 volt Vref, you would expect a max input of about
1Vrms. That is reasonable for a consumer -10dbV system. The datasheet
indicates S/(N+D) of better than 84db at -.5dbFS with the PGA set at 0db
(notice they don't specify this figure at full analog input voltage).

Also interesting is that the PGA does not offer gain reduction (only gain
boost 0 to +18db). I'm not sure how you are turning the gain down. Are
there any amplifier stages ahead of the ADC on the PCB? I wonder if you are
measuring with the PGA set at 0db? Perhaps the internal PGA has problems
swinging enough to drive the ADC full scale. Might be a production
tolerance issue within the chip. In any event, you stated that the highest
harmonic was -40db down. This is about 1% distortion. Not great, but not
terrible. How fast does this reduce as you drop the signal level to the
card (not the gain)? I suspect that even a .5 db drop will result in a
large drop in the distortion level. Given that, you still have plenty of
headroom to deal with.

Mark

wrote in message ...
Does anyone know if it is typical for an ADC to distort when driven near
full
scale (0dB)?

I've got an Edirol UA5 USB audio interface. I put a 1600Hz sine wave in
the
line input (played from the line out of a Creative Labs Nomad Jukebox 3),
turned the level up to just under 0dB and recorded a few seconds of audio.

When I looked at the spectrum (Adobe Audition 1.5!) I see several
harmonics
(multiples of 1600Hz), and the largest one was only -40dB down or so. I
noticed these extra peaks went away once I lowered the input to say, -6dB.
So, it seems to exhibit some kind of clipping/overloading behavior, but
the
input signal was not over. It was something like -1 or -2dB.

I looked at the circuit in this box. It is based on the AK4524 ADC. The
circuit seems to follow the standard design spec. The chip operates on
+5V.
The driving opamps run on +/-5V. The input signal to the ADC seems to be
about 0.7V RMS at full scale.

So, could it be that these ADCs are really not very good? Or is the
supporting circuitry poor? Should I expect a more expensive box, like
MOTU or
Presonus, to do it right? I guess it pays to check these things out.

Thanks for any help...
Richard

In article , wrote:
Does anyone know if it is typical for an ADC to distort when driven near full
scale (0dB)?

I've got an Edirol UA5 USB audio interface. I put a 1600Hz sine wave in the
line input (played from the line out of a Creative Labs Nomad Jukebox 3),
turned the level up to just under 0dB and recorded a few seconds of audio.

When I looked at the spectrum (Adobe Audition 1.5!) I see several harmonics
(multiples of 1600Hz), and the largest one was only -40dB down or so. I
noticed these extra peaks went away once I lowered the input to say, -6dB.
So, it seems to exhibit some kind of clipping/overloading behavior, but the
input signal was not over. It was something like -1 or -2dB.

This is bad behaviour. And it's not clipping either, it's the onset
of distortion before the clipping point.

I looked at the circuit in this box. It is based on the AK4524 ADC. The
circuit seems to follow the standard design spec. The chip operates on +5V.
The driving opamps run on +/-5V. The input signal to the ADC seems to be
about 0.7V RMS at full scale.

So, could it be that these ADCs are really not very good? Or is the
supporting circuitry poor? Should I expect a more expensive box, like MOTU or
Presonus, to do it right? I guess it pays to check these things out.

This is basically the cheapest ADC circuit that it's possible to get,
and it's preceded by cheez-whiz op-amps running at OUTRAGEOUSLY low
rail voltages. It's a wonder you get any linearity at all with something
like this. What you see is typical of the kind of thing I complain about
as cost-cutting gone wild.

Just relax and keep your levels down. And save up for a decent set of
outboard converters. There is a reason the Prism costs more than the
Edirol.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

(Scott Dorsey) writes:

In article , wrote:
Does anyone know if it is typical for an ADC to distort when driven near full
scale (0dB)?

I've got an Edirol UA5 USB audio interface. I put a 1600Hz sine wave in the
line input (played from the line out of a Creative Labs Nomad Jukebox 3),
turned the level up to just under 0dB and recorded a few seconds of audio.

When I looked at the spectrum (Adobe Audition 1.5!) I see several harmonics
(multiples of 1600Hz), and the largest one was only -40dB down or so. I
noticed these extra peaks went away once I lowered the input to say, -6dB.
So, it seems to exhibit some kind of clipping/overloading behavior, but the
input signal was not over. It was something like -1 or -2dB.

This is bad behaviour. And it's not clipping either, it's the onset
of distortion before the clipping point.


Dear Scott:

Thanks for the detailed response. Yes, I agree it is bad behaviour.

I looked at the circuit in this box. It is based on the AK4524 ADC. The
circuit seems to follow the standard design spec. The chip operates on +5V.
The driving opamps run on +/-5V. The input signal to the ADC seems to be
about 0.7V RMS at full scale.

So, could it be that these ADCs are really not very good? Or is the
supporting circuitry poor? Should I expect a more expensive box, like MOTU or
Presonus, to do it right? I guess it pays to check these things out.

This is basically the cheapest ADC circuit that it's possible to get,
and it's preceded by cheez-whiz op-amps running at OUTRAGEOUSLY low
rail voltages. It's a wonder you get any linearity at all with something
like this. What you see is typical of the kind of thing I complain about
as cost-cutting gone wild.


I'm just wondering if it is the Opamps or the ADC. I guess to find out I
could put a signal directly into the analog in pin on the ADC.

Just relax and keep your levels down. And save up for a decent set of
outboard converters. There is a reason the Prism costs more than the
Edirol.

It seems to distort anywhere above -3dB. I would be happy to run it at -6dB
and below, say, but the problem is there is a really nice "clip" light that
comes on at -3dB. I run this thing "standlone" (without a PC) and rely on the
clip light to set levels. (I set it so the clip light just flickers every now
and then.) I would be happy if I could find a way to make the clip light to
turn on earlier, say at -6 or -9dB. Does anyone know how this clip light
works? I'm thinking it might be reading right out a digital channel
somewhere.

Oh yeah, one more question: What is the simplest ADC circuit I can make that
takes line in and outputs SPDIF optical? In particular, do I need this Cirrus
CS8402 to generate optical out? This does rate conversion and all kinds of
stuff, but can I just clock it at 44.1 and output 44.1/16 directly from the
AK4524? Again, I'm assuming that the AK4524 is really an OK chip. People
have told me the same chip is in reasonable prducts, like the "Apogee MiniME"
ADC box.

--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

Richard

In article , wrote:

I'm just wondering if it is the Opamps or the ADC. I guess to find out I
could put a signal directly into the analog in pin on the ADC.

I'm betting it's a soft-limit inside the ADC chip. But if you are getting
that much distortion, you should see flat-topping on a scope. I see
flat-topping at about 3% distortion on a 1 KC tone, with my eyes and my
ancient Tek scope.

Just relax and keep your levels down. And save up for a decent set of
outboard converters. There is a reason the Prism costs more than the
Edirol.

It seems to distort anywhere above -3dB. I would be happy to run it at -6dB
and below, say, but the problem is there is a really nice "clip" light that
comes on at -3dB. I run this thing "standlone" (without a PC) and rely on the
clip light to set levels. (I set it so the clip light just flickers every now
and then.) I would be happy if I could find a way to make the clip light to
turn on earlier, say at -6 or -9dB. Does anyone know how this clip light
works? I'm thinking it might be reading right out a digital channel
somewhere.

The clip light is SUPPOSED to come on when a certain number of consecutive
full-scale samples appear. How many depends on the manufacturer... some
equipment has DIP switches inside so you can select it. I tend to set
things up for four consecutive FS to light the lamp, but some guys have it
set for as many as eight.

I'd set for peaks around -12dBFS or lower. Give yourself some headroom.
You have plenty of dynamic range, why not use some?

Oh yeah, one more question: What is the simplest ADC circuit I can make that
takes line in and outputs SPDIF optical? In particular, do I need this Cirrus
CS8402 to generate optical out? This does rate conversion and all kinds of
stuff, but can I just clock it at 44.1 and output 44.1/16 directly from the
AK4524? Again, I'm assuming that the AK4524 is really an OK chip. People
have told me the same chip is in reasonable prducts, like the "Apogee MiniME"
ADC box.

There are a bunch of chips that will give you SPDIF output.. check the
Crystal Semi databook (which I think is now the Cirrus). There are a couple
that don't do rate conversion or anything fancy. You can probably get
some of the Crystal Semi evaluation boards for cheap on Ebay if you look,
also, and they would be a good start toward homebrewing various ADC designs.

I'm not sure I'd call the MiniME reasonable, though... but use your ears
and eyes and see for yourself.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."