Follow-Up to the "Friday's Tip of the Week" for Week 59 - "Why Noise Is Your Friend"

On the Sonar-as-test-equipment theme, I have a project setup with a sine wave generator and an oscilloscope plug-in that is surprisingly useful for examining cross-over distortion when biasing a valve (tube, sorry) guitar amplifier. I put the sine wave into the amp and pick up the output via a Koch Dummybox. A purist would probably argue that the Dummybox is not identical to a physical loudspeaker, but it serves the purpose of getting the output of the amp into Sonar and it a whole lot easier on the ears.

That is a good approach for the wahs. That also looks like a good review, I'll have to go back and read it.

ChristopherM
[Apologies to OP for this distraction]

Actually I find the "SONAR as test equipment" very interesting. Carry on...

ampfixer
 
I typically would use a 100 mv sine wave at a fixed frequency between 800 Hz and 1 KHz. I need to test all the frequencies at once. Is Pink noise the way to do this? If so what can I use as a noise generator?
 

What you could do is add an audio track with the volume right down.  Then insert this free plugin from Meldaproductions:

https://www.meldaproduction.com/plugins/product.php?id=MNoiseGenerator

I say put the volume down before adding this plugin is because this effect is generating noise at all times, even when your project is not playing.  If you leave the volume up, you get a nice dose of Artic Blast as soon as you add the plugin.

ampfixer
I have modelling software to design EQ circuits. It shows how the output frequency will be effected by the EQ controls, but it's theoretical. Listening to the finished product doesn't usually reflect what the simulator predicted.

What modelling software are you using? As most EQ circuits in guitar amps are entirely passive, they should not present too much difficulty in modelling. I've used LTSpice with reasonable success on passive EQ. Even on valve/tube based circuits, it gives usable results, although the end results will be limited by the accuracy of the spice models of the valves. You can even use wav files as input and generate wav files as output (although it's a bit tedious). I haven't found a role for Sonar in any of this, I would add.

ampfixer
The amplifier output could be anywhere from 5 to 25 volts AC when pushed into clipping.

For the sake of the soundcard's health, I would use a simple potential divider (just a couple of resistors with adequate power ratings) to scale the output. I would then push that into the soundcard via a small capacitor just to ensure that no stray DC from the amp can get to the soundcard. As the resulting RC network is itself a high-pass filter, you might need to select values that do not compromise your results. And this is probably a good moment to point out that I only do things this way via Sonar because I don't have access to hardware 'scope or signal generator these days.

ampfixer
I typically would use a 100 mv sine wave at a fixed frequency between 800 Hz and 1 KHz. I need to test all the frequencies at once. Is Pink noise the way to do this? If so what can I use as a noise generator?

The (free) Melda plug, MNoisegenerator, can produce various types of noise, including pink, and there is also another, MOscillator, that handles other signal generation. I use an ancient 32-bit plug S(M)exoscope as a 'scope. That said, I'm still not really clear on what benefit you will get from noise, as opposed to a swept sine wave. The output will be noise filtered by your EQ and I'm not sure how you will make a meaningful interpretation of that. Nevertheless, I'll be interested to see what you find. I'd also be interested if anyone knows of a decent 64-bit Oscilloscope plug-in, ideally free.
 

Add another reason for why I love this forum...