Gang of Four
now the rest can know http://www.arrl.org/files/file/Technology/tis/info/pdf/9811qex026.pdf
now the rest can know http://www.arrl.org/files/file/Technology/tis/info/pdf/9811qex026.pdf
That's his Billerica MA basement I posted.
Attachments
Last edited:
Brief update on my oscillator project: The front panel PCB is fully assembled and, as far as already tested, working as expected (yes, the shafts of the rotary switches will get a cut!):
Samuel
Precision only found in Swiss watchmaking
! Wow! Absolutely amazing work. Very much interested in making it when available.
When you can find them, this one has a lot gooing for it:
Ultra low distortion (<0.00001%) 1kHz sine generator assembled and tested PCB
Victor's osc schematic is here:
http://www.diyaudio.com/forums/equipment-tools/205304-low-distortion-audio-range-oscillator-21.html#post3116055
http://www.diyaudio.com/forums/equipment-tools/205304-low-distortion-audio-range-oscillator-21.html#post3116055
Thanks Arnold. It is hard to believe I posted this question 4 years ago and this thread is still active. I have learned a lot since then and I thank all members for their contributions.
I do have one of Victor's oscillators in a box that runs off a bank of 9 volt cells.
I primarily use my HP339 or my HP239 at my bench.
I do have one of Victor's oscillators in a box that runs off a bank of 9 volt cells.
I primarily use my HP339 or my HP239 at my bench.
Samuel,
Would it be possible for your oscillator to have an external input that would allow to be driven by another source, so that could do sweeps? I am thinking that way, some analyzers still could do automated measurements, like THD +N vs frequency, but benefitting by driving this higher quality oscillator? Or maybe you already have that feature?
Best
AR2
Would it be possible for your oscillator to have an external input that would allow to be driven by another source, so that could do sweeps? I am thinking that way, some analyzers still could do automated measurements, like THD +N vs frequency, but benefitting by driving this higher quality oscillator? Or maybe you already have that feature?
Best
AR2
Precision only found in Swiss watchmaking
Thanks!
The oscillator is not continuously tunable. True continuous, steppless tuning would be very difficult to implement at these distortion/noise levels as it requires some sort of VCA (motor-driven pots would perhaps be the best bet).
However it is ready for computer control, as all control parameters are in the form of digital, opto-coupled signals. So if anyone is up to it (not me), a simple USB interface plus some host-software would do the trick.
Samuel
As others have said, very impressive work Samuel!
I would be happy to help with computer interface when time comes.
Chris
I was thinking that might be the case. Certainly not a big deal, but it might be nice.
Thank you.
Digital multiplier
I've taken the digitally controlled multiplier from my state variable generator design and placed on it own board. I intended some time ago to try the multiplier on other oscillator types like the Bridged T and Wien Bridge types.
I installed the multiplier in s 339A oscillator replacing the Jfet. The multiplier runs off the original controller of the 339A. below is some spectrum of the results.
The first pic is a measurement on a Shibasoku 725D off the oscillator op amp bypassing the buffer amp. The measurement is really quite good considering the oscillator is operating at 10Vpp (7Vrms). The 725D measured -129.5dBV THD and -114dBV THD+N
The second pic is the output of the 339A oscillator on the buffered side. The measurement is good at the same relative levels but not nearly as good as the oscillator itself. We just just get the distortion of the buffer op amp, The 725D measured -117dBV THD and -110dBV THD+N
I'll work on the buffer more to get the distortion down.
The Shibasoku 724D is in the -90dBV range. The spectrum is scaled such that 0dBV is -90dBV so add 10dB to the spectrum measure to get the results.
I've taken the digitally controlled multiplier from my state variable generator design and placed on it own board. I intended some time ago to try the multiplier on other oscillator types like the Bridged T and Wien Bridge types.
I installed the multiplier in s 339A oscillator replacing the Jfet. The multiplier runs off the original controller of the 339A. below is some spectrum of the results.
The first pic is a measurement on a Shibasoku 725D off the oscillator op amp bypassing the buffer amp. The measurement is really quite good considering the oscillator is operating at 10Vpp (7Vrms). The 725D measured -129.5dBV THD and -114dBV THD+N
The second pic is the output of the 339A oscillator on the buffered side. The measurement is good at the same relative levels but not nearly as good as the oscillator itself. We just just get the distortion of the buffer op amp, The 725D measured -117dBV THD and -110dBV THD+N
I'll work on the buffer more to get the distortion down.
The Shibasoku 724D is in the -90dBV range. The spectrum is scaled such that 0dBV is -90dBV so add 10dB to the spectrum measure to get the results.
Attachments
Last edited:
David- Are you able to share the PCB design? I have a Shibasoku waiting for this I think.
Right now it is something of a sample +hold + gain control. Would it be possible to add a digital value for a reference voltage so it can be come most of the AGC block? Theoretically using two it would be possible to tune the oscillator as well.
Right now it is something of a sample +hold + gain control. Would it be possible to add a digital value for a reference voltage so it can be come most of the AGC block? Theoretically using two it would be possible to tune the oscillator as well.
I'm prototyping the board right now and a second generation will be available soon.
I'll send you a board with the ADC mounted. I got a bunch of ADC at a reduced cost.
How many board do you want? I made a few minor changes.
It take about three weeks for the boards to deliver.
I'll try the multiplier next with a Wien bridge. The only one I have is Victor's.
Should be interesting.
I have a schematic done already and a theory of operation written up.
For tuning, unless you mean auto fine adjust, I use a micro controller and Mdacs.
The ADC and pulse generator is not necessary. I guess for manual tuning control from a control voltage it would work but this would be a much more costly way of doing it.
I would use a much cheaper ADC for this and multiplex the ADC for multiple Mdacs.
A different design would make it feasible.