Page 3 of 17

Re: Super Low Distortion Ultra Pure Audio Oscillators Revisited

Posted: Thu Sep 07, 2017 4:17 pm
by mlloyd1
Hi Wayne:

Any plans to offer PCBs of this oscillator design?

mlloyd1

Re: Super Low Distortion Ultra Pure Audio Oscillators Revisited

Posted: Thu Sep 07, 2017 6:38 pm
by mediatechnology
Thank you for the inquiry and recently joining us.

I've been kicking around the idea of offering a board though I think I want to add an output buffer with a simple 0, -20, -40 dB pad in addition to the variable level it now has.
I also want to explore Victor's floating TL431 shunt regulator and using an optional battery supply.
It might be possible to achieve a floating output.
So I think I have a little more work to do first.

Re: Super Low Distortion Ultra Pure Audio Oscillators Revisited

Posted: Tue Jan 16, 2018 10:01 am
by JR.
I shared a link to this project before but apparently the designer (Mike) had a stroke so has abandoned this project unfinished.

For the low distortion oscillator crowd, (not me) maybe somebody will pick this up and carry it further down the road.

http://mike-wsm.org.uk/ldo.html

JR

Re: Super Low Distortion Ultra Pure Audio Oscillators Revisited

Posted: Wed Jan 17, 2018 7:03 am
by mediatechnology
Sorry to hear about Mike. Where does he post usually?

One of the things I noticed in his design was a comparator-based detector.

I haven't looked at it too closely but why are there two OTAs?

Re: Super Low Distortion Ultra Pure Audio Oscillators Revisited

Posted: Wed Jan 17, 2018 8:37 am
by JR.
I posted a link to his discussion at GroupDIY (old prodigy). He was involved in the Radford Low distortion oscillator designs.

He first posted about the project in mar of 2017. He said he quit working on it after the stroke, but is still vertical.

JR

Re: Super Low Distortion Ultra Pure Audio Oscillators Revisited

Posted: Wed Jan 17, 2018 9:37 am
by mediatechnology
OK, it looks like he has "I" and "Q" quadrature outputs from the State Variable Filter feeding a squaring circuit (I) with Q feeding the input to the first OTA to form a sample and hold which samples at the peak of Q. The second OTA is the VCA for gain stabilization of the State Variable oscillator.

I've seen that in a number of oscillators since it provides low ripple and fast settling. I've recently downloaded several service manuals I should link to.

I suggested an LM13700-based S/H similar to Mike's over at DIYAudio but it was poo-pooed (as many things are) being too simple and too inexpensive. Unless an IC costs upwards of $5 it simply won't do...

Right now I'm interested in spot tones and Victor's super-simple halfwave rectifier is slugged enough with a long tc that half or full-wave makes no difference in the THD. Its pretty much limited by the op amps and in my case the "temporary" construction.

One thing I do see is radiated HF rectifier switching noise as the rectifier op amp goes open loop. Its way deep in the null. There's a lot to be said for non-switching rectification (e.g. Cordell) because nothing goes open loop.

I need to get this gizmo off the Protoboard because I need it and the companion notch filter to measure the Dual Class-A amp.

I'm hoping I can use this to eliminate the requirement for LME49720s:

Image
A "Super Op Amp" using complimentary differential input "helper" transistors. (c) 2018 Wayne Kirkwood.

See: viewtopic.php?f=6&t=926

Re: Super Low Distortion Ultra Pure Audio Oscillators Revisited

Posted: Wed Jan 17, 2018 10:45 am
by JR.
My memory is a little fuzzy about this but I recall an exchange with Brad Wood (RIP) many years ago where he shared a schematic of a design that used the quadrature and squaring trig identity. I am not sure this is remotely a limiting factor for THD as sundry examples demonstrate...

I still suspect a micro could be incorporated into a smart AGC loop, but I am no longer chasing things my bench can't measure (I don't even have an audio test bench anymore). There are already enough zeros after the decimal point for me. :D

JR

Re: Super Low Distortion Ultra Pure Audio Oscillators Revisited

Posted: Wed Jan 17, 2018 4:41 pm
by psupine
Hi all,

I'm heading down a Cordell THD analyser build. There's a perpetual arms race to make test gear quieter than the gear it's testing, so naturally I am looking at every option I can find that might make it better.

My own personal biases are that I really prefer working with DIPs rather than SOICs which limits what opamps are available. Another bias is that I already have a stock of THAT4301s, so if there is anything to do with VGAs or RMS sensing they will be my first call.

So my first question to the group is this. If I build Bob's design as is, except that I replace the original 5534s with LME49720s, what sort of residual am I likely to see in that oscillator? Is there a point in changing what ain't broke?

Next question (a bit off topic) is about the current move away the LME49720? Is this an end-of-life issue or the problem with HF rectification? I can't find the notice from TI about end-of-life and I've not run into the radio problem (I'm driving the inputs at milliamps and have a quiet enclosure environment so the stray signals don't get a look in)

Re: Super Low Distortion Ultra Pure Audio Oscillators Revisited

Posted: Wed Jan 17, 2018 4:56 pm
by mediatechnology
Thanks for dropping by and following us!

I would try it with 5534s first.
Next question (a bit off topic) is about the current move away the LME49720? Is this an end-of-life issue or the problem with HF rectification? I can't find the notice from TI about end-of-life and I've not run into the radio problem (I'm driving the inputs at milliamps and have a quiet enclosure environment so the stray signals don't get a look in)
I think the end of life problem has been resolved. Victor has had issues getting quiet LME49720s. If shielded they are not DECT EMI susceptable. I still use them in low-gain circuits because nothing matches their low distortion. But I don't think I want to be locked into them for the oscillator.

Re: Super Low Distortion Ultra Pure Audio Oscillators Revisited

Posted: Wed Jan 17, 2018 7:34 pm
by JR.
psupine wrote: Wed Jan 17, 2018 4:41 pm Hi all,

I'm heading down a Cordell THD analyser build. There's a perpetual arms race to make test gear quieter than the gear it's testing, so naturally I am looking at every option I can find that might make it better.

My own personal biases are that I really prefer working with DIPs rather than SOICs which limits what opamps are available. Another bias is that I already have a stock of THAT4301s, so if there is anything to do with VGAs or RMS sensing they will be my first call.

So my first question to the group is this. If I build Bob's design as is, except that I replace the original 5534s with LME49720s, what sort of residual am I likely to see in that oscillator? Is there a point in changing what ain't broke?

Next question (a bit off topic) is about the current move away the LME49720? Is this an end-of-life issue or the problem with HF rectification? I can't find the notice from TI about end-of-life and I've not run into the radio problem (I'm driving the inputs at milliamps and have a quiet enclosure environment so the stray signals don't get a look in)
I will resist my temptation to ask why... :lol:

I would be careful about nixing SMD, because many newer parts are not even sold in thru hole.

That said I don't expect performance of Cordell's design to be compromised by the devices. You may be able to make small incremental improvements in the margin, but I am not sure how to even measure that.

I am not familiar with the THAT part numbers but a latest generation VCA might be nice in the AGC loop... Of course degenerate it so it isn't working very hard. The clever JFET gain control Cordell did does not suck... low noise and low distortion but there are quieter JFET available since then.

I had some experience designing (crude) test equipment back in the 80s and it seems like chasing your tail to make better THD analyser resolution and then making cleaner oscillators... When are you finished? :lol:

JR