Page 1 of 1

Ultrapure Sinewave Oscillator

Posted: Sat Aug 20, 2011 6:15 pm
by mediatechnology
I just recevied the print edition of EDN and noticed the front cover article titled "Test 18-Bit ADCs with an Ultrapure Sine-wave Oscillator."

Somehow I knew it was written by the late Jim Williams. It was. :(

I can almost always recognize Jim's work by its title, the schematic art and the vintage oscillograms. Stephen Woodward's schematics stand out the same way.

They're getting a THD+N of around -105.

http://www.edn.com/design/test-and-meas ... 8-bit-ADCs

Image
"Test 18-Bit ADCs with an Ultrapure Sine-wave Oscillator," Jim Williams, EDN, August 11, 2011.

LTC had another application note AN67 for an ultrapure oscillator using "super op amps." It begins on page 62.

https://www.analog.com/media/en/technic ... /an67f.pdf

Re: Ultrapure Sinewave Oscillator

Posted: Sat Aug 20, 2011 6:53 pm
by JR.
RIP... him and RAP

JR

edit- in the back of my head I am thinking about mixed technology where a micro driving DPOTs could get the osc loop dialed in just to the edge of amplitude stability, so the dynamic range of the leveling loop can be necked back to almost nil... but we are chasing the echos of the footprints of angels on the pinhead...

Re: Ultrapure Sinewave Oscillator

Posted: Thu Aug 25, 2011 12:38 pm
by mediatechnology
Just stumbled upon Linear Technology AN-43 "Marrying Gain and Balance" written by Jim Williams.

Has some Wien bridge circuits along with an article on distortion measurement by Hofer.

WIlliams, "Wien Bridge Oscillators" https://www.analog.com/media/en/technic ... /an43f.pdf (page 29)

Hofer, "Understanding Distortion Measurements" https://www.analog.com/media/en/technic ... /an43f.pdf (page 44)

Willliams, "The Wien Bridge and Mr. Hewlett" https://www.analog.com/media/en/technic ... /an43f.pdf (page=43)

Re: Ultrapure Sinewave Oscillator

Posted: Thu Aug 25, 2011 2:27 pm
by JR.
I hate being a digital tout, but if you can get a many bit DAC and the ability to tweak the sine wave look up table, you could correct the linearity of the DAC which is probably already pretty good.

This quickly gets into a chicken/egg thing with how do you measure the DAC linearity to tweak it.

I'm making a 9 bit or so sine wave in my tuner and it's clean enough for the drum heads.

FWIW I'm using a PWM to make my sine wave so I guess a high enough frequency PWM timebase could also make a pretty pure sine sine wave... Quickly back of the envelope math, a 20 MHz processor could only update a 16b PWM at 300 Hz rate.. so we want to go way way faster clock rates for more bit resolution and higher frequency sine waves.

JR