Distortion Measurement

[joaquins]

  I keep reading "Low Drop Out" (regulator) very often in this topic 

  Funny story from Saul Walker, the army asked his team to build a clock distribution in a big war ship, the signal should be transmitted all over the ship, 1MHz (IIRC) TKD+N<-80dB. Being fixed freq you can filter it, they said no problem, they couldn't pass about 74dB. They went back to the bosses saying they couldn't get to the specs... Oh, that's a typo they said, it should said 40dB!!!!

  For LDOs (the ones on this topic) there's also the lamp option, older technology (HP someone?) but still pretty good, probably can get better than the JFET option and makes for the simplest circuit once you got it trimmed for the right lamp.

JS

 

[jdbakker]

[...] For LDOs (the ones on this topic) there's also the lamp option, older technology (HP someone?) but still pretty good, probably can get better than the JFET option and makes for the simplest circuit once you got it trimmed for the right lamp.

Simplicity yes, distortion no. The application note I linked to earlier starts out with William Hewlett's lamp-stabilised oscillator (see appendix C and note 4 on page 29 in particular), and goes on to show that you can get better performance with a JFET or a Vactrol for the control element.

JDB.
[A longer version of the Wien-oscillator article appears in Analog Circuit Design - Art, Science and Personalities, a book well worth a read]

 

[benb]

There's a nice big thread on oscillators at this link. So far, everything mentioned here (Cordell's oscillator/tester, the HP lightbulb circuit, Jim Williams appnote) has also been discussed below, as well as lots of other stuff.

http://www.diyaudio.com/forums/equipment-tools/205304-low-distortion-audio-range-oscillator.html

 

[merlin]

Just posting these here for reference. Apologies if someone's already pointed them out.
http://www.janascard.cz/PDF/An%20ultra%20low%20distortion%20oscillator%20with%20THD%20below%20-140%20dB.pdf
http://vkaudiotest.co.uk/VK-1%20Audio%20Oscillator.htm

 

[mike-wsm]

http://ukhhsoc.torrens.org/makers/Radford/Test_Equipment/Low_Distortion_Oscillators/index.html

 

[mike-wsm]

Thanks for all the kind suggestions. The design is making very slow progress, I did a lot of number crunching  getting the attenuator right and then I had to break off for a while to do some much needed work on domestic stuff. Latest state of the art is at http://mike-wsm.org.uk/ldo.html I am hoping for a respectable 0.001 per cent thd+n at 1kHz and good figures over the range 10Hz to 100kHz.

 

[mike-wsm]

Alas I had a severe stroke in May last year, had to give away all my stuff and abandon work for ever on this project. However my old Radford designs have never been matched so they remain the sole examples of how to do the job, not one-offs but fast tuneable production propositions. The current circuits have not been built and tested, but I have left them at http://mike-wsm.org.uk/g3.html for anyone to have a go. Good luck!

 

[madswitcher]

Hi Mike

So sorry to hear about your stroke and you having to give up on your project.  You have one big fan of Radford here as I bought a DMS 4 - the autoranging version - at the Dunstable Downs Radio Rally for £15.  I am really an old time fart in that prefer to be able to relate to what is going on when measuring and am still on the look-out for a Radford LDO on Ebay.

The DMS4 needed a bit of work on it (a dud chip) and I swopped some of the op-amps for some modern lower-noise versions and screened the case on the inside.  I calibrated it by the book and it works like a charm- it just sits there and does what it says on the tin, down to about 0.0006% with the new chips etc.  I just love the auto lock-on function: I then use the distortion output into a spectrum analyser on the PC to find out what is where. 

So belatedly, thank you for such an elegant design - and I wish you well.

Best regards

Mike

 

[JohnRoberts]

Alas I had a severe stroke in May last year, had to give away all my stuff and abandon work for ever on this project. However my old Radford designs have never been matched so they remain the sole examples of how to do the job, not one-offs but fast tuneable production propositions. The current circuits have not been built and tested, but I have left them at http://mike-wsm.org.uk/g3.html for anyone to have a go. Good luck!

Sorry to hear about the stroke, getting old isn't for wimps.

Hopefully somebody will pick up your work. Be well, or as well as you can.

JR