Demain & David,
thanks for the info, I'll have a look at it.
-----:-----
Returning to basic fixed freq super-oscillators, anybody have experience with the design of the 10kHz osc from LT (LT app note 67, page 62ff)? The clever composite opamp building blocks based on high speed CFB opamps used there look promising as does the ALC control element based on a transconductance amp, haven't seen that elsewhere.
A unity-gain stable composite with 120dB gain at 10kHz is something not seen every day, maybe it is of use for notch filters, lowpasses and IMD signal mixers etc, too. And for real applications like preamps etc...
For sure construction and adjustment/verification of this osc is an extremely daunting task. How is one supposed to check residual distortion of a claimed "parts-per-billion"-grade osc?....
thanks for the info, I'll have a look at it.
-----:-----
Returning to basic fixed freq super-oscillators, anybody have experience with the design of the 10kHz osc from LT (LT app note 67, page 62ff)? The clever composite opamp building blocks based on high speed CFB opamps used there look promising as does the ALC control element based on a transconductance amp, haven't seen that elsewhere.
A unity-gain stable composite with 120dB gain at 10kHz is something not seen every day, maybe it is of use for notch filters, lowpasses and IMD signal mixers etc, too. And for real applications like preamps etc...
For sure construction and adjustment/verification of this osc is an extremely daunting task. How is one supposed to check residual distortion of a claimed "parts-per-billion"-grade osc?....
I've never been a big fan of injection locking, always preferring a PLL where possible. If you need to lock your oscillator to your VCO, you can adapt some of the auto-tune circuitry in my THD analyzer to make a PLL that locks the SVF oscillator.
Cheers,
Bob
Hi Bob,
I was thinking of your auto tune circuits when I wrote about this.
How much ripple would you say comes from those circuits and does it get in the way?
Is there a reason? PLL's tend to be complicated and an injection locking scheme is pretty simple, essentially dropping out of the circuit when not in use. What is the benefit of the PLL? If you are using like the Boonton for precise tuning across a broad range I can see it but for a fixed oscillator it may be overkill.
I thought someone had built this and posted info on it. Some of the LT app stuff has proven really hard for me to duplicate and this was so off the wall I did not try. I would like to see the layout they used to make it work.
Ultra low distortion audio oszillator the never ending search for a pure sine is one effort with unbelievable breadboarding. If someone has a layout that worked I would be interested.
I think it's more like this.
"Super Gain Block Oscillator Circuitry
When A1, as described above is connected with U1, as
shown in Figure 89, the resulting circuit is not only unitygain
stable but has open-loop gain of 180dB at 10kHz".
There is at least one thread on the forum and if you google you will find a bit more.
It will require careful layout and ground planes.
I think the 1ppb was more a theoretical number. I haven't heard of anyone actually measuring this. What analyzer goes down this low?
That one IS crazy, even the great Jim Williams would have been impressed I guess.
A double-sided single PCB for this will be tricky, but I could imagine a "hybrid" approach using a main board and vertical sub-PCBs, and the occasional shield and p2p wiring. Although I'm tempted to try I foresee a lot of effort going down the river until this plane will really fly...
For the time being, Victor's plug&play builds shall do for me.
That one IS crazy, even the great Jim Williams would have been impressed I guess.
A double-sided single PCB for this will be tricky, but I could imagine a "hybrid" approach using a main board and vertical sub-PCBs, and the occasional shield and p2p wiring. Although I'm tempted to try I foresee a lot of effort going down the river until this plane will really fly...
For the time being, Victor's plug&play builds shall do for me.
I've done a couple of double sided boards with ground planes recently.
One turned out great and the other was a flop. I have to fine tune the process.
What mades it work was a 1/16 inch acrylic sheet placed between the the upper and lower
transparencies. A piece of scrap 1/16 FR4 at the top to have something to tape to, all on sheet of glass. Replace the acrylic sheet with the board when everything is lined up.
This is off topic so I'll leave it at that.
Definitely limited by your sound card at this point, but the response looks impressive (the card, not the oscillator) in comparison to my M-Audio ProFire 610.
The only way to really see what Victor's oscillator can do is to notch it.
I have had some difficulty getting ideal driving capability out of Victor's oscillator after some testing this weekend. Tried driving my balanced input preamp, followed by a notch filter, to see how my preamp measures up. Fought for a long time with noise, grounding, and the annoying reality that Victor's oscillator really wants circuit common to be referenced to earth/shield.
I have Pete Millet's soundcard interface, which helped a bit to at least get me some decent measurements (happy to say the preamp measures really well), but the interface adds a bunch of 60Hz and its harmonics to the spectrum, as well as adding quite a bit (-116dB) of 2nd distortion to Victor's output. So there was no reliable way of measuring 2nd harmonic at this point, but at least the 2nd is fairly benign.
I will look into a Jensen output transformer to better balance out the oscillator output. In my experience a good transformer having to only run at 1 kHz will beat the soundcard interface.
The only way to really see what Victor's oscillator can do is to notch it.
I have had some difficulty getting ideal driving capability out of Victor's oscillator after some testing this weekend. Tried driving my balanced input preamp, followed by a notch filter, to see how my preamp measures up. Fought for a long time with noise, grounding, and the annoying reality that Victor's oscillator really wants circuit common to be referenced to earth/shield.
I have Pete Millet's soundcard interface, which helped a bit to at least get me some decent measurements (happy to say the preamp measures really well), but the interface adds a bunch of 60Hz and its harmonics to the spectrum, as well as adding quite a bit (-116dB) of 2nd distortion to Victor's output. So there was no reliable way of measuring 2nd harmonic at this point, but at least the 2nd is fairly benign.
I will look into a Jensen output transformer to better balance out the oscillator output. In my experience a good transformer having to only run at 1 kHz will beat the soundcard interface.
Even an amatuer like myself can see that the layout of this board is very poor. I am interested in working on this in spite of my lack of experience.
Based on the schematic, it looks like the output is balanced. Is that correct?
I think I will start by building the super gain block first and see how that goes.
Last edited:
Using the advanced search function, put in username "frex" in topic "oscillator"
Frex has built it and sells the boards. I'm about half way through building it.
It would be useful for folks to read Bernie Oliver's article in the HP Journal on how bridged T oscillators really work.
Do you mean this one?
The Effect of u-Circuit Non-Linearity
on the Amplitude Stability of RC Oscillators
http://www.hpl.hp.com/hpjournal/pdfs/IssuePDFs/1960-04.pdf
It's a very nice looking board:
[3rd round] AN67 Ultra-low THD 10kHz sine oscillator PCB Group buy. - Discussion Page 10 - diyAudio
He did a great job. It's designed to fit into a specific chassis. It's certainly the highest performance Wein bridge in the known world.
My idea is to have a set of fixed frequency oscillators all in one box with a selector switch to power up and connect the one to be used. Each oscillator could be optimized for it's particular frequency, though I don't know how to do that. I want all oscillators to be on separate boards, each board having a mounting hole in each corner of the board to maximize flexibility in installing it in a chassis. This one could certainly be the basis for the 10kHz oscillator I would use. I'm thinking of: 1kHz, 10kHz, and a twin tone oscillator of some kind for IM testing.
Dirk