Thread: HP 3582A ?
View Single Post
  #14   Report Post  
Old November 28th 06, 01:30 PM posted to rec.audio.pro,rec.audio.opinion,rec.audio.tech,rec.radio.amateur.equipment
Arny Krueger Arny Krueger is offline
external usenet poster
 
First recorded activity by RadioBanter: Nov 2006
Posts: 25
Default HP 3582A ?

"Straydog" wrote in message
.com

*If* the thing under test is amenable to it, there's
little or no hope of improving on a PC soundcard


Are you talking about any PC _sound_card?


Let's put it this way. A lot of good audio development and test was done
with equipment such as the HP 333/334. Conventional wisdom is that if you
can't measure it with a 333, you can't hear it either. A 333 pretty well
ran out of gas around 0.02% THD. IOW if you hooked a perfect signal source
up to a 333 and nulled it carefully, that would be a good residual. The
residual THD of a typical Realtek audio interface on a modern motherboard is
less than that.

and Matlab as a combo. Might
need a breadboard to adapt the signals, but if it's in
range, it'll outrun just about anything for the money.


All you need is a a modern PC and a freeware audio testing program - RMA55.

A PC _sound_card? Meant for _audio_ frequency ranges?


Unh huh.

Like single diget Hz to 20-30 kHz?


Well up to 20 KHz or so.

You think I'd rather
have that than a decent DC to RF (X mHz) oscilloscope 10
v/div down to 1-10 mv/div sensitivity?


Try verifying the specs of a $39.95 DVD players audio outputs with a scope.

In the real world of general purpose troubleshooting and development you
would ideally have both.

And, can measure DC.


That's what voltmeters are for, chum.

Maybe if you want to do some FFT or other DSP on an
audio signal you can have the soundcard.


You seem to be seeing the light.

I'll pick the dedicated gear with specs that fit the ap.


Of course.

You're assuming you can trust the test equipment.
Sometimes, you can't....


You mean you don't need to _assume_ the soundcard is
working or you don't need to worry about sw bugs or
invalid process or tollerances or other spec-dependent
limitations? So, if the computer says "X" then it is
absolutely true, valid, straight from the horses mouth?


Measurements are what they are. Their utility is always in the
interpretation.

No ands, ifs, or buts? Do you know what the term
"parasitic" means as it applies to chips?


It means, "Where is my dual-trace scope?"