Page 4 of 17

Re: Super Low Distortion Ultra Pure Audio Oscillators Revisited

Posted: Thu Jan 18, 2018 4:33 pm
by psupine
Yes, yes, I hear your mocking tone. I have a whole family who can sing that song for me. :lol:

It's a journey innit. I'd rather build something and learn along the way rather than buy a magic box that does stuff that I can't fix when it breaks.

I admit that surface mount thing is an arbitrary choice. My usual through-hole palette is AD797, LME49720, OPA2134, OP07, NE5532 and TL072, depending on where it is in the circuit. They are not the newest and greatest, but those first three are still pretty good chips, and the last three are pretty cheap for what they do. I've certainly built surface mount designs when I did this kind of stuff for a living, but at home without the microscope and rework station I really can't be bothered with it. I guess that limits the ultimate performance, but I'm ok with that.

Re: Super Low Distortion Ultra Pure Audio Oscillators Revisited

Posted: Thu Jan 18, 2018 4:54 pm
by mediatechnology
I think the ambitious part of building the Cordell Analyzer are obtaining and wiring the multipole switches.

Re: Super Low Distortion Ultra Pure Audio Oscillators Revisited

Posted: Thu Jan 18, 2018 5:30 pm
by psupine
Since I have to re-roll the PCB anyway, I was going to use Elektro Bauelemente PCB mount wafers for S2 and S4 and NSF open wafers for S1, S3 and S5. All the switch parts I need are available from RS Components, rather than trying to track down NOS rotary switches from 1980. I'm also going to use balanced output and input. I managed to find some through-hole packaged LM1496 demodulators, but this is really off thread now. So back on thread ...

I am exploring the Cordell State Variable Oscillator with LTSpice, substituting Linear opamps to match the 5534 and 318 main features of BW and SR. It behaves nicely at 65Hz 650Hz 6500Hz but at 65kHz with the JFET fully off, the oscillator blows up. I'm sure it works fine in reality, but I've previously found that if LTSpice doesn't like circuit it's because I've done something wrong. My next step I suppose is to try to import the correct opamp models and explore that.

Or maybe I should just breadboard it ...

Re: Super Low Distortion Ultra Pure Audio Oscillators Revisited

Posted: Thu Jan 18, 2018 6:40 pm
by JR.
psupine wrote: Thu Jan 18, 2018 5:30 pm

Or maybe I should just breadboard it ...
Not a bad idea... I didn't have spice, and couldn't trust it at first.

sorry I am not mocking, just reflecting where I am in my journey (far down the long road.).

JR

Re: Super Low Distortion Ultra Pure Audio Oscillators Revisited

Posted: Wed Feb 27, 2019 9:50 pm
by Jacko
Image

Re: Super Low Distortion Ultra Pure Audio Oscillators Revisited

Posted: Thu Feb 28, 2019 5:11 am
by mediatechnology
Nice circuit. I wasn't familiar with the VCA810.

http://www.ti.com/lit/ds/symlink/vca810.pdf

I'm curious if I can make one with a 2180 and conserve an op amp.

Re: Super Low Distortion Ultra Pure Audio Oscillators Revisited

Posted: Thu Mar 28, 2019 1:13 pm
by terkio
This thread revived an old time where I struggled with LF oscillators for measuring distortion with the double T bridged filter.
The best I achieved was nothing to praise, distortion measurements capable of 0.1 %, that was good enough for me.
I remember the struggle to achieve an amplitude stable and pure sine signal, two conflicting objectives. At that time the trick was to use a low power filament bulb.
From this challenge, I said, never again and thought the solution was to go digital, with the sine table in a ROM and dump it to a DAC at a chosen speed.
Nowdays with 24 bits DAC and GHz clock frequencies at affordable cost, isn't it the way to go, to synthetize perfect low frequency signals ? Not going to extremes, I presume a PIC or an Arduino could fill the bill, if more power is needed, a $50 Raspberry Pi

Re: Super Low Distortion Ultra Pure Audio Oscillators Revisited

Posted: Thu Mar 28, 2019 1:29 pm
by JR.
The VCA810 is yet another part I never encountered.

@terkio... yup, I messed around with a microchip DSP that had 16b dac built into it... I used it to make multiple sine waves simultaneously for a prototype drum tuner... when overlaying 8-10 different pitch sine waves simultaneously you have to reduce the amplitude of each to avoid saturation, but it worked great. The concept worked (used FFT to read for drum resonances) but it would have been too cool for drummers who think my current tuner is too complicated and too expensive. :lol:

JR

Re: Super Low Distortion Ultra Pure Audio Oscillators Revisited

Posted: Thu Mar 28, 2019 1:42 pm
by terkio
@JR. Wonderful, I am glad to see a DSP with a DAC built in, a perfect solution for this and many more.

EDIT: I spotted dsPIC33F 4 bucks !
Can you recommend a kit with a DSP and DAC, I could play with, plug and play ?

Re: Super Low Distortion Ultra Pure Audio Oscillators Revisited

Posted: Thu Mar 28, 2019 3:47 pm
by JR.
terkio wrote: Thu Mar 28, 2019 1:42 pm @JR. Wonderful, I am glad to see a DSP with a DAC built in, a perfect solution for this and many more.

EDIT: I spotted dsPIC33F 4 bucks !
Can you recommend a kit with a DSP and DAC, I could play with, plug and play ?
Sorry about the veer...

I am not a big fan of microchip development tools but have used PIC for years... so DSPic was an incremental step for me.

I am already using a 16b pic 24F family in my second generation drum tuner so I was able to drop a DSPic into the Pic24F foot print of my production design with only minor modifications, and use my current development environment.

In my current drum tuner I use the PWM output from the PIC24F to make sine waves... not very hifi but good enough for exciting drumhead resonances. I actually make 3 octave spaced sine waves for a non-FFT measurement algorithm of my own design to perform rough scans faster. The DSP with DAC would make higher quality sine waves, and perform better measurements using FFT.

But I have lost motivation to make a next generation, better and more expensive DSP/FFT platform.

JR