Sign in with
Sign up | Sign in
Your question

THD = 0.05% believable?

Tags:
Last response: in Home Audio
Share
Anonymous
March 24, 2005 10:07:18 PM

Archived from groups: rec.audio.tech (More info?)

Having built a little bare-bones musical instrument preamp, I started
playing with the RMAA program. Without going all the way to my basement
for my voltmeter to measure actual signal levels, I just ran the tests.
At first blush, my preamp measures 0.05% THD.

My question is whether this is believable to those who have been doing
these kinds of measurements? Is this in the ballpark for what is
basically a collection of op amps (TLC2202 and TLC2264)? Mainly I just
want a sanity check on my numbers.

Thanks!

More about : thd believable

Anonymous
March 25, 2005 3:46:34 AM

Archived from groups: rec.audio.tech (More info?)

In article <1111720038.902899.25290@g14g2000cwa.googlegroups.com>,
elephantcelebes@yahoo.com wrote:

> Having built a little bare-bones musical instrument preamp, I started
> playing with the RMAA program. Without going all the way to my basement
> for my voltmeter to measure actual signal levels, I just ran the tests.
> At first blush, my preamp measures 0.05% THD.
>
> My question is whether this is believable to those who have been doing
> these kinds of measurements? Is this in the ballpark for what is
> basically a collection of op amps (TLC2202 and TLC2264)? Mainly I just
> want a sanity check on my numbers.
>
> Thanks!

Some op-amps have over 140dB gain. Run them in a low gain closed loop
and the distortion drops to nearly zero. 0.001% isn't unusual.

The catch is that the gain drops rapidly with frequency. The circuit
transitions from closed loop gain to open loop gain, which can have
wicked distortion. Older and cheaper op-amps start to distort badly
around 10KHz even with low closed loop gains. High grade audio op-amps
have a high gain, a fast slew rate, and reduced distortion even in open
loop mode. They may be too expensive for large mass-produced audio
equipment but they're cheap in the scope of hobby projects. Mid-grade
op-amps work flawlessly as long as the closed loop gain isn't too high.
Anonymous
March 25, 2005 5:31:21 AM

Archived from groups: rec.audio.tech (More info?)

In <1111720038.902899.25290@g14g2000cwa.googlegroups.com>, on 03/24/05
at 07:07 PM, elephantcelebes@yahoo.com said:

>Having built a little bare-bones musical instrument preamp, I started
>playing with the RMAA program. Without going all the way to my
>basement for my voltmeter to measure actual signal levels, I just ran
>the tests. At first blush, my preamp measures 0.05% THD.

>My question is whether this is believable to those who have been doing
>these kinds of measurements? Is this in the ballpark for what is
>basically a collection of op amps (TLC2202 and TLC2264)? Mainly I just
>want a sanity check on my numbers.

"RMAA"? I'm not familiar with that. Is it a computer program that uses
the internal sound card as the signal generator and measurement system?

I have relatively little respect for run of the mill sound cards. The
0.05% implies fairly low noise. I'm not surprised at the preamp having
0.05% THD+N, but I am somewhat surprised that your sound card would
have a residual of 0.05%.

What frequency did you use to measure the 0.05%.?

The input filter for the line input of the cound card limits the
usefulness of high frequency measurements. (If the input has a sharp
20KHz filter, a 15KHz THD measurement doesn't make much sense because
all the harmonics are chopped.)

Unfortunately, the THD measurement does not correleate well with a
listening jury's opinion. Sometimes a 0.1% unit will sound much better
than a 0.05% unit.

-----------------------------------------------------------
spam: uce@ftc.gov
wordgame:123(abc):<14 9 20 5 2 9 18 4 at 22 15 9 3 5 14 5 20 dot 3 15
13> (Barry Mann)
[sorry about the puzzle, spammers are ruining my mailbox]
-----------------------------------------------------------
Related resources
Can't find your answer ? Ask !
Anonymous
March 25, 2005 8:49:51 AM

Archived from groups: rec.audio.tech (More info?)

<elephantcelebes@yahoo.com> wrote in message
news:1111720038.902899.25290@g14g2000cwa.googlegroups.com
> Having built a little bare-bones musical instrument preamp, I
started
> playing with the RMAA program. Without going all the way to my
> basement for my voltmeter to measure actual signal levels, I just
ran
> the tests. At first blush, my preamp measures 0.05% THD.
>
> My question is whether this is believable to those who have been
doing
> these kinds of measurements?

Yes. Almost everything modern measures in the 0.01 to 0.1% range at
any level much below clipping.

>Is this in the ballpark for what is
> basically a collection of op amps (TLC2202 and TLC2264)? Mainly I
just
> want a sanity check on my numbers.

I was looking at the product folder and kind of amused to see 62 pages
of tech data, and seeemingly not one word about nonlinear distortion.
Anonymous
March 25, 2005 8:51:49 AM

Archived from groups: rec.audio.tech (More info?)

"Barry Mann" <zzzz@zzzz.zzz> wrote in message
news:4243c4c2$4$avgroveq$mr2ice@wcnews.cyberonic.com
> In <1111720038.902899.25290@g14g2000cwa.googlegroups.com>, on
03/24/05
> at 07:07 PM, elephantcelebes@yahoo.com said:
>
>> Having built a little bare-bones musical instrument preamp, I
started
>> playing with the RMAA program. Without going all the way to my
>> basement for my voltmeter to measure actual signal levels, I just
ran
>> the tests. At first blush, my preamp measures 0.05% THD.
>
>> My question is whether this is believable to those who have been
>> doing these kinds of measurements? Is this in the ballpark for what
>> is basically a collection of op amps (TLC2202 and TLC2264)? Mainly
I
>> just want a sanity check on my numbers.
>
> "RMAA"? I'm not familiar with that. Is it a computer program that
uses
> the internal sound card as the signal generator and measurement
> system?

Yes.

>
> I have relatively little respect for run of the mill sound cards.
The
> 0.05% implies fairly low noise. I'm not surprised at the preamp
having
> 0.05% THD+N, but I am somewhat surprised that your sound card would
> have a residual of 0.05%.

Here's a little consciousness-raiser for you Barry:

http://www.pcavtech.com/soundcards/lynxtwo/
Anonymous
March 25, 2005 9:39:03 AM

Archived from groups: rec.audio.tech (More info?)

"Barry Mann" <zzzz@zzzz.zzz> wrote in message
news:4243c4c2$4$avgroveq$mr2ice@wcnews.cyberonic.com

> The input filter for the line input of the cound card limits the
> usefulness of high frequency measurements. (If the input has a sharp
> 20KHz filter, a 15KHz THD measurement doesn't make much sense
because
> all the harmonics are chopped.)

That's why God invented high frequency IM measurements, which RMAA
supports well.

I use 18 & 20 KHz with RMAA. RMAA also does IM sweeps.

> Unfortunately, the THD measurement does not correleate well with a
> listening jury's opinion. Sometimes a 0.1% unit will sound much
better
> than a 0.05% unit.

That's why there are other forms of measurement. ;-)
Anonymous
March 25, 2005 10:35:22 AM

Archived from groups: rec.audio.tech (More info?)

On 24 Mar 2005 19:07:18 -0800, elephantcelebes@yahoo.com wrote:

>Having built a little bare-bones musical instrument preamp, I started
>playing with the RMAA program. Without going all the way to my basement
>for my voltmeter to measure actual signal levels, I just ran the tests.
>At first blush, my preamp measures 0.05% THD.
>
>My question is whether this is believable to those who have been doing
>these kinds of measurements? Is this in the ballpark for what is
>basically a collection of op amps (TLC2202 and TLC2264)? Mainly I just
>want a sanity check on my numbers.

Perfectly believable, in fact not that great for an opamp design! :-)
--

Stewart Pinkerton | Music is Art - Audio is Engineering
Anonymous
March 25, 2005 10:39:52 AM

Archived from groups: rec.audio.tech (More info?)

Arny Krueger wrote:
> >Is this in the ballpark for what is
> > basically a collection of op amps (TLC2202 and TLC2264)? Mainly I
> > just want a sanity check on my numbers.
>
> I was looking at the product folder and kind of amused to see 62
pages
> of tech data, and seeemingly not one word about nonlinear distortion.

Those lengthy data sheets were indeed rather annoying. But after
scratching my head over them for a while, I could not really see
anything to discourage me from using those chips in an audio circuit. I
was looking for something with FET input to handle piezoelectric
transducers, and I wanted to run the whole shebang off a single 9-V
battery. Add up enough premium JFET op amps, and you will go through
batteries like the Stones used to go through hotel rooms. The CMOS
chips seemed worth a try.

I was indeed careful about keeping the closed loop gains within reason
-- no more than 20 dB in any section.

According to RMAA, the motherboard sound "card" on my PC sports 24 bit,
192 kHz conversion. This setup seems to be quite functional for my
needs. The only thing that I add is a voltmeter to calibrate the signal
levels.

My noise measurement is a bit more elaborate. I use a 100k resistor as
a noise "standard" that puts out roughly 40 nV/rtHz, collect spectra
with the resistor and with the input shorted, compute the noise from
the quadrature difference of these curves. The numbers that I get are
in very good agreement with the published noise values for the op amps
I am using.

Granted, this is largely an academic exercise. The major noise source
in my system is typically ambient, and the main source of distortion is
probably the speaker. My reason for playing with RMAA is not to hang a
magic number on this preamp, but to find out the range of operating
conditions under which I can expect it to behave itself.
Anonymous
March 25, 2005 1:07:11 PM

Archived from groups: rec.audio.tech (More info?)

In <c8WdnTjJ7ZnYcN7fRVn-uw@comcast.com>, on 03/25/05
at 05:51 AM, "Arny Krueger" <arnyk@hotpop.com> said:
[ ... ]

>Here's a little consciousness-raiser for you Barry:

>http://www.pcavtech.com/soundcards/lynxtwo/

Darn!!

Of course, my comment was aimed at the losest possible cost card or on
motherboard facility.

Thanks

-----------------------------------------------------------
spam: uce@ftc.gov
wordgame:123(abc):<14 9 20 5 2 9 18 4 at 22 15 9 3 5 14 5 20 dot 3 15
13> (Barry Mann)
[sorry about the puzzle, spammers are ruining my mailbox]
-----------------------------------------------------------
Anonymous
March 25, 2005 3:33:59 PM

Archived from groups: rec.audio.tech (More info?)

"Barry Mann" <zzzz@zzzz.zzz> wrote in message
news:42442ace$5$avgroveq$mr2ice@wcnews.cyberonic.com
> In <c8WdnTjJ7ZnYcN7fRVn-uw@comcast.com>, on 03/25/05
> at 05:51 AM, "Arny Krueger" <arnyk@hotpop.com> said:
> [ ... ]
>
>> Here's a little consciousness-raiser for you Barry:
>
>> http://www.pcavtech.com/soundcards/lynxtwo/
>
> Darn!!
>
> Of course, my comment was aimed at the losest possible cost card or
on
> motherboard facility.

Well of course, those on-board sound chips perform a bit more poorly
than a high end card street priced at most of a $grand. ;-)

However, the performance of on-board sound chips has undergone ongoing
and considerable improvement. 5 or so years ago they were as a rule
pretty horrible. They have improved considerably to the point where
they are merely mediocre. Putting a number on it, on-board sound chips
have improved from dynamic ranges in the 50 dB range then to the 75-80
dB range now. I suspect that their performance will continue improve
into the 90 dB range within a few years.
Anonymous
March 25, 2005 3:43:11 PM

Archived from groups: rec.audio.tech (More info?)

zzzz@zzzz.zzz (Barry Mann) wrote in news:4243c4c2$4$avgroveq
$mr2ice@wcnews.cyberonic.com:


> "RMAA"? I'm not familiar with that. Is it a computer program that uses
> the internal sound card as the signal generator and measurement system?

Rightmark Audio Analyzer

http://audio.rightmark.org/index_new.shtml

It's free. Doesn't replace an Audio Precision test set,
but it's a lot more affordable. The sound card could use
some sort of attenuator/interface, especially when working
with amplifiers that swing more than a few volts.

--Damon
Anonymous
March 25, 2005 9:37:16 PM

Archived from groups: rec.audio.tech (More info?)

"Damon Hill" <damon1six1@comcast.not> wrote in message
news:Xns96246D119304Ddamon161attbicom@216.196.97.131
> zzzz@zzzz.zzz (Barry Mann) wrote in news:4243c4c2$4$avgroveq
> $mr2ice@wcnews.cyberonic.com:
>
>
>> "RMAA"? I'm not familiar with that. Is it a computer program that
>> uses the internal sound card as the signal generator and
measurement
>> system?
>
> Rightmark Audio Analyzer
>
> http://audio.rightmark.org/index_new.shtml
>
> It's free. Doesn't replace an Audio Precision test set,
> but it's a lot more affordable. The sound card could use
> some sort of attenuator/interface, especially when working
> with amplifiers that swing more than a few volts.

I use a 5,000 ohm, 2 watt Type J AKA Type AB carbon comp
potentiometer.
Anonymous
March 26, 2005 12:38:24 AM

Archived from groups: rec.audio.tech (More info?)

"Arny Krueger" <arnyk@hotpop.com> wrote in news:M46dnSUvTqAxPdnfRVn-
2A@comcast.com:

> "Damon Hill" <damon1six1@comcast.not> wrote in message
> news:Xns96246D119304Ddamon161attbicom@216.196.97.131

>> It's free. Doesn't replace an Audio Precision test set,
>> but it's a lot more affordable. The sound card could use
>> some sort of attenuator/interface, especially when working
>> with amplifiers that swing more than a few volts.
>
> I use a 5,000 ohm, 2 watt Type J AKA Type AB carbon comp
> potentiometer.

That'll work, but I'm not sure about the calibration. I'd
been thinking about a buffered input with switched attenuator.

--Damon
Anonymous
March 26, 2005 12:57:31 AM

Archived from groups: rec.audio.tech (More info?)

"Barry Mann" <zzzz@zzzz.zzz> wrote in message
news:4243c4c2$4$avgroveq$mr2ice@wcnews.cyberonic.com...
> "RMAA"? I'm not familiar with that. Is it a computer program that uses
> the internal sound card as the signal generator and measurement system?

Yes.

> I have relatively little respect for run of the mill sound cards. The
> 0.05% implies fairly low noise. I'm not surprised at the preamp having
> 0.05% THD+N, but I am somewhat surprised that your sound card would
> have a residual of 0.05%.

You haven't measured any lately have you? Even an SB live can do less than
0.01% THD+N with a 1 kHz test signal.
And the best cards are around 0.001%!

> What frequency did you use to measure the 0.05%.?

RMAA uses 1 kHz for THD.

> The input filter for the line input of the cound card limits the
> usefulness of high frequency measurements. (If the input has a sharp
> 20KHz filter, a 15KHz THD measurement doesn't make much sense because
> all the harmonics are chopped.)

Yes, so why do it? You can hear 30 kHz?

> Unfortunately, the THD measurement does not correleate well with a
> listening jury's opinion. Sometimes a 0.1% unit will sound much better
> than a 0.05% unit.

Yes, it's possible depending on spectral distribution. You can use a
software FFT spectrum analyser to check what it is.

MrT.
Anonymous
March 26, 2005 3:41:49 PM

Archived from groups: rec.audio.tech (More info?)

"Arny Krueger" <arnyk@hotpop.com> wrote in news:D 7GdnfOGY_0a1tnfRVn-
jA@comcast.com:

> "Barry Mann" <zzzz@zzzz.zzz> wrote in message
> news:42442ace$5$avgroveq$mr2ice@wcnews.cyberonic.com
>> In <c8WdnTjJ7ZnYcN7fRVn-uw@comcast.com>, on 03/25/05
>> at 05:51 AM, "Arny Krueger" <arnyk@hotpop.com> said:
>> [ ... ]
>>
>>> Here's a little consciousness-raiser for you Barry:
>>
>>> http://www.pcavtech.com/soundcards/lynxtwo/
>>
>> Darn!!
>>
>> Of course, my comment was aimed at the losest possible cost card or
> on
>> motherboard facility.
>
> Well of course, those on-board sound chips perform a bit more poorly
> than a high end card street priced at most of a $grand. ;-)
>
> However, the performance of on-board sound chips has undergone ongoing
> and considerable improvement. 5 or so years ago they were as a rule
> pretty horrible. They have improved considerably to the point where
> they are merely mediocre. Putting a number on it, on-board sound chips
> have improved from dynamic ranges in the 50 dB range then to the 75-80
> dB range now. I suspect that their performance will continue improve
> into the 90 dB range within a few years.
>
>
>

Not sure if the onboard D-A chip or the audio system is at fault but my
company furnshed laptop's audio system is the worst I have encountered to
date. It is a Dell Latitiude D600.

r


--
Nothing beats the bandwidth of a station wagon filled with DLT tapes.
Anonymous
March 26, 2005 8:13:05 PM

Archived from groups: rec.audio.tech (More info?)

"R" <spmaway@ylhoo.com> wrote in message
news:Xns962544210303Cmc2500183316chgoill@10.232.1.1

>
> Not sure if the onboard D-A chip or the audio system

You can determine that using the Audio Rightmark program.

> is at fault but my company furnshed laptop's audio system is the
worst I have
> encountered to date. It is a Dell Latitiude D600.

I can't find out what audio interface is involved.

You might find out from the Device Manager
Anonymous
March 26, 2005 8:14:11 PM

Archived from groups: rec.audio.tech (More info?)

"Damon Hill" <damon1six1@comcast.not> wrote in message
news:Xns9624C7D0057F5damon161attbicom@216.196.97.131
> "Arny Krueger" <arnyk@hotpop.com> wrote in news:M46dnSUvTqAxPdnfRVn-
> 2A@comcast.com:
>
>> "Damon Hill" <damon1six1@comcast.not> wrote in message
>> news:Xns96246D119304Ddamon161attbicom@216.196.97.131
>
>>> It's free. Doesn't replace an Audio Precision test set,
>>> but it's a lot more affordable. The sound card could use
>>> some sort of attenuator/interface, especially when working
>>> with amplifiers that swing more than a few volts.
>>
>> I use a 5,000 ohm, 2 watt Type J AKA Type AB carbon comp
>> potentiometer.
>
> That'll work, but I'm not sure about the calibration.

I use a DVM to directly monitor the output of the UUT.

> I'd been thinking about a buffered input with switched attenuator.

It's hard to build a buffer that is as good as many sound cards.
Anonymous
March 27, 2005 12:57:28 AM

Archived from groups: rec.audio.tech (More info?)

R <spmaway@ylhoo.com> wrote:
>
>Not sure if the onboard D-A chip or the audio system is at fault but my
>company furnshed laptop's audio system is the worst I have encountered to
>date. It is a Dell Latitiude D600.

Dell must hire deaf engineers. The onboard audio on my Inspiron 7500
is equally gawdawful, beat for spurious noises and leakage whines only
by the initial release of the Atari ST.


Francois.
Anonymous
March 28, 2005 6:15:50 AM

Archived from groups: rec.audio.tech (More info?)

fps@idiom.com ((null)) wrote in news:1111870648.643508@smirk:

> R <spmaway@ylhoo.com> wrote:
>>
>>Not sure if the onboard D-A chip or the audio system is at fault but my
>>company furnshed laptop's audio system is the worst I have encountered to
>>date. It is a Dell Latitiude D600.
>
> Dell must hire deaf engineers. The onboard audio on my Inspiron 7500
> is equally gawdawful, beat for spurious noises and leakage whines only
> by the initial release of the Atari ST.
>
>
> Francois.
>

The Inspiron I have here is wonderful compared to the D600. I wish someone
made a reasonable PCMCIA sound card.

r
Anonymous
March 28, 2005 10:05:16 AM

Archived from groups: rec.audio.tech (More info?)

"R" <spmaway@ylhoo.com> wrote in message
news:Xns9626CE2114815mc2500183316chgoill@10.232.1.1
> fps@idiom.com ((null)) wrote in news:1111870648.643508@smirk:
>
>> R <spmaway@ylhoo.com> wrote:
>>>
>>> Not sure if the onboard D-A chip or the audio system is at fault
>>> but my company furnshed laptop's audio system is the worst I have
>>> encountered to date. It is a Dell Latitiude D600.
>>
>> Dell must hire deaf engineers. The onboard audio on my Inspiron
7500
>> is equally gawdawful, beat for spurious noises and leakage whines
>> only by the initial release of the Atari ST.
>>
>>
>> Francois.
>>
>
> The Inspiron I have here is wonderful compared to the D600. I wish
> someone made a reasonable PCMCIA sound card.

Echo does - well its a PC Card, but that's probably what you meant.
Anonymous
March 29, 2005 2:50:07 AM

Archived from groups: rec.audio.tech (More info?)

So why is this so? In an age where any given CD player on the market
will have the digital aplitude to sound nearly indistinguishably alike,
how can computer soundcards still be stuggling to even get decent
16-bit sound? Shouldn't "near ideal" D/A convertors be pretty much
obiquitous these days?

I'm not questioning your statement, mind you. I'm just curious to hear
a bit more detail.
Anonymous
March 29, 2005 12:03:44 PM

Archived from groups: rec.audio.tech (More info?)

"dangling entity" <randycat99@earthlink.net> wrote in message
news:1112079007.069195.320480@f14g2000cwb.googlegroups.com

> So why is this so? In an age where any given CD player on the
market
> will have the digital aplitude to sound nearly indistinguishably
> alike, how can computer soundcards still be stuggling to even get
> decent 16-bit sound? Shouldn't "near ideal" D/A convertors be
pretty
> much obiquitous these days?

It's getting to be that way, but its taken an amazingly long time.

It wasn't that long ago that CD players did have converters that were
nearly indistinguishable from straight wire, while sound cards still
had converters that performed at the 8-bit not 16 bit level. I'm
talking 1995-1998.

There are many reasons, including the fact that while CD players need
only have good converters for playback, while audio interfaces do both
recording and playing back. Because normal use involves two trips
through the converters, the economic burden is doubled by the need for
two sets of converters, while the converters individually need to
perform twice as well.

Looking at the SoundBlaster line the Audigy's were the first ones that
were good enough to be nearly indistinguishable. The good news is that
the Live! line were retrofitted with better converters, and the newer
Live! versions also perform at the nearly indistinguishable level.

You experience find real-world examples of this by downloading files
from

http://www.pcabx.com/product/soundcard/
!