[DigitalSignal]

This message does not have any commercial purpose. I just want to
report the progress in one of our development. About a year ago I
raised question that whether it is realistic to achieve 150 dB dynamic
range in the measurement. The discussion is he

http://groups.google.com/group/rec.a...gitalsignal999

After a year of development, we made an instrument that can measure up
to 130dB to 140dB. See: www.go-ci.com for product details. We are 10
dB less than what we planned. But it is is still a satisfactory
result. 13dB means that in the same frame of data capture, we can see
signal clearly as high as 10Vpk, or as small as 10uV. With this
progress, no gain setting is needed any more.

[Chris Hornbeck]

On Sat, 8 Dec 2007 19:06:25 -0800 (PST), DigitalSignal
wrote:

After a year of development, we made an instrument that can measure up
to 130dB to 140dB.

Your website seems broken in Mozilla, but maybe I'm atypical.

For another data point, my 1970's Sound Tech analyzers (a 1700B
and a 1701A, both with IM) can measure THD+N to an averaged -115dB
over a signal range to 110KHz, but with reduced resolution above
the audio range. When they're working...

Another difference from the older analog analyzers is that
the old guys deliver a scope display which gives an immediate
qualitative analysis to a trained eye, but *no* connection
for modern analysis (pre-PC, pre-AppleII ferchrissakes). I
can't see enough of the website to tell if your machine
provides something comparable.

Bandwidth is a tougher issue because affordable solutions
must fit into consumer silicon. But if there were a way to
allow measurements a couple of octaves above the audio
range, albeit at reduced resolution, it would be even more
useful for development work.

Much thanks for your work,

Chris Hornbeck

[DigitalSignal]

Chris,

Thanks for the info about bug in Mozilla. I will check.

Your 1701A that measured up to -115dB is amazing. I assume it is an
analog instrument with some kind of tracking filters.

The new product we just released does provide both scope and spectral
display. It combines the time domain acquisition and frequency
analysis together. The upper sampling rate is 102.4kHz which provides
at least 2 octave beyond normal audio range.

The target market is for vibration analysis. Does anybody think this
technique fits into audio application as well?

Many thanks,

[Mike Rivers]

On Dec 8, 10:06 pm, DigitalSignal wrote:

After a year of development, we made an instrument that can measure up
to 130dB to 140dB.

So where can we see one of these marvels? Too bad I didn't know about
it sooner. You just missed the deadline for the annual Test Equipment
issue of Pro Audio Review.

By the way, the web site worked oK for me, using Mozilla Firefox. Web
sites and computers are like that.

[Scott Dorsey]

DigitalSignal wrote:

Your 1701A that measured up to -115dB is amazing. I assume it is an
analog instrument with some kind of tracking filters.

The new product we just released does provide both scope and spectral
display. It combines the time domain acquisition and frequency
analysis together. The upper sampling rate is 102.4kHz which provides
at least 2 octave beyond normal audio range.

The target market is for vibration analysis. Does anybody think this
technique fits into audio application as well?

Take a look at the old B&K gear... they originated the notion of doing
frequency analysis with tracking filters, and they built a number of
units like the 2112 that managed to get outrageously high S/N levels
while being full of 12AX7 tubes, in part due to very careful gain
management (and switching calibrated gain stages in and out.

We can do a lot better today with active filters and digital sampling
systems... and with a digital sampling system it is possible to design
a device to have outrageously high S/N over a very narrow bandwidth.
This is fine if you only need to measure a single frequency at a time.

Oh yeah, and look inside an Audio Precision....
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Paul Stamler]

"DigitalSignal" wrote in message
...
Chris,

Thanks for the info about bug in Mozilla. I will check.

Your 1701A that measured up to -115dB is amazing. I assume it is an
analog instrument with some kind of tracking filters.

The new product we just released does provide both scope and spectral
display. It combines the time domain acquisition and frequency
analysis together. The upper sampling rate is 102.4kHz which provides
at least 2 octave beyond normal audio range.

Well, not really. Sampling rate of 102.4kHz means maximum frequency of
51.2kHz. Two octaves below that would be 12.8kHz, whereas the upper limit of
hearing is usually considered to be 20kHz, with individual variations of
course. To get two octaves above that, you'd need to be sampling at least at
160kHz.

Peace,
Paul

[Richard Crowley]

"Mike Rivers" wrote ...
DigitalSignal wrote:

After a year of development, we made an instrument that can measure
up
to 130dB to 140dB.

So where can we see one of these marvels? Too bad I didn't know about
it sooner. You just missed the deadline for the annual Test Equipment
issue of Pro Audio Review.

Check the bandwidth before refering to it as "audio equipment".

[Mike Rivers]

On Dec 9, 1:25 pm, "Richard Crowley" wrote:

Check the bandwidth before refering to it as "audio equipment".

I suppose it was presumptuous of me, since this was posted in an audio
newsgroup. It looks to me like it goes above 20 kHz for two channels.
That's enough like audio for me, though with 96 kHz becoming the norm
for most amateur work (and professionals, too) we tend to want to look
at response at least to 35 kHz.

Not clear how the sample rate and number of channels relates, but I
assume it does.

[DigitalSignal]

The instrument we did, CoCo-80, has 8 measurement input channels. Each
has an A/D and can simultanously sample up to 102.4kHz. The valid
frequency range is about 46kHz.

[Chris Hornbeck]

On Sat, 8 Dec 2007 23:49:51 -0800 (PST), DigitalSignal
wrote:

Your 1701A that measured up to -115dB is amazing. I assume it is an
analog instrument with some kind of tracking filters.

It would be amazing if it were true, but I've misstated the situation.
The old analog nulling analyzers let the user visually resolve
distortion components down around -115dB, but have a broadband
noise background about 20dB higher.

I'm very sorry to have described them so poorly. Late night
is my only excuse, and not a good one.


The new product we just released does provide both scope and spectral
display. It combines the time domain acquisition and frequency
analysis together. The upper sampling rate is 102.4kHz which provides
at least 2 octave beyond normal audio range.

The target market is for vibration analysis. Does anybody think this
technique fits into audio application as well?

For audio use, you'd be competing with a combination of (usually
existing) computer sound cards and freeware analysis programs,
like Rightmark (sp?). Sound card based measurements have similar
bandwidth limitations to your new machine.

The old Sound Tech's and current AP's can make measurements at
*signal* frequencies to above 100KHz. This would require sampling
rates of over 600KHz just to see the third harmonic. Audio
development work needs to see how electronics works at least a
coupla octaves above the audio range, because lots of what's
interesting is how the electronics deals with out-of-band
signal.

Thanks very much for your efforts, and keep 'em flying,

Chris Hornbeck