Sound Quality Vs. Measurements

Status
Not open for further replies.
Administrator
Joined 2004
Paid Member
Aim an IR detector at the chip. Run a 10 Hz (or 5 Hz or whatever) signal through it. Do you see that signal at the IR detector?
OK, interesting idea, but an LF sine wave might be my choice. I'd rather use something with more dynamics, on the order of 20-30dB peak over average like a good recording.
With that signal I'm not sure what I'd be looking for, thermally.

But I'll consider TL's point, that hot IS different, tho not because of thermal distortion. Too bad he won't discuss it here.
 
Sy,

Here's an experiment that will convince the doubters: grind the top off a TO3 device on a heatsink. Aim an IR detector at the chip. Run a 10 Hz (or 5 Hz or whatever) signal through it. Do you see that signal at the IR detector?

Before you would do this, you would have to demonstrate first that IR detection system was capable to actually track any such temperature change as postulated...

Can you provide proof it does?

And would you mind referencing the publication of your experiment and controls?

Ciao T
 
Hi,

Which ones?

The ones you reference in your post # 281.

I take you performed the experiment before telling others to perform it?

Also, what the sensor does is one thing, you need to show whole system.

My IR based temperature measuring gun certainly is not fast enough to show a 10Hz sinewave...

Ciao T
 
Administrator
Joined 2004
Paid Member
Silicon solar cells have a bandwidth of 20Khz or better. They are used as pickups for optical soundtracks on film. IR detectors pick up the remote control stuff that is even higher than that.

I just don't know what the signal would look like with high dynamic range music going thru a transistor. Could be fun to try.
 
The question is, for what purpose. ;)

Current source with huge emitter degradation is not a current mirror. And not a complementary pair of source followers, where difference in voltage bias in milivolts cause current change in amperes.


Can't edit...

If you take emitter followers, things are even worse. Since designers of transistor amps are stuck with voltage amplification paradigm, they are forced to set bias of output emitter followers by voltage between bases. For stabilization they use thermal feedback, that shows all usual stability issues. No need to measure temperatures, just measure currents, how they vary with signal and it's envelope.
 
Sy,

IR photodiodes are routinely used with 38kHz subcarrier systems. If you have a gun-type thermometer, the signal can be taken off the amplifier following the detector.

Why do not just disclose the exact experimental set-up you used for the experiment you described in post # 281 and what controls you employed to make sure your results where valid, to save us all the time?

Ciao T
 
Hi,

I'm not sure I understand you. Or perhaps you're not understanding me. Are you saying that experiments should not be proposed?

Of course experiments should be proposed.

However, it is reasonable to expect of those proposing them to have actually carried them out themselves, if they claim that said experiments will provide proof of their position.

Anyway, from the above I conclude you did not carry out the experiment yourself and that your claim for the experiment "to convince doubters" is based on your belief in the outcome, rather than actually having carried out the experiment?

Ciao T
 
Spectral resolution is pretty straightforward- with a typical computer setup, better than 1Hz is easy. The sidebands that result from FM or plain old amplitude distortion are pretty characteristic. $200 gets you nearly 100 kHz of bandwidth, better than -150dB noise floor, and better than 1Hz resolution.

This capability cost six figures when I was at Nicolet. I am still totally amazed at how easy and reliable modern implementations are.

Hear ya. I have an e-mu 192KHz, 24 bit acquisition and a hundred bucks of software. Beats the old Scientific Atlantic analyzers that were 200K in the 70's.

What I can't find is a really nice interpretation of a delayed sweep storage scope. I bought about three, and none of then are up to the task.

You may get a kick. many years ago I sent Nicolet, HP, and Tek running. I needed a pre-triggered two channel scope that could capture a half nanosecond half volt glitch. I could do in on phosphor with the Polaroid, but no one could do in electronically back then. Everyone was using clock jitter to get that kind of resolution. Circa 1980. I finally got the picture using two meter probes as a delay line and a 1/2 meter as the trigger.
 
Status
Not open for further replies.