Demian, what happened with the one oscillator you purchased from Germany if I recall correctly? It was already in an enclosure and had either multiple outputs or multiple frequencies? Sometime within the last 6 months...one only knows how far back in this thread 😉
It worked well but there was leakage from the other oscillators. He was going to do a redesing to resolve that issue. Ill check with him on his progress. It was quite low distortion.
Sent from my LG-H811 using Tapatalk
Sent from my LG-H811 using Tapatalk
Let me know what you find. I would like to upgrade to a better oscillator. Victor's is great, but my time is so limited right now...one with an enclosure and ready to go is very appealing.
Have you guys seen this:
Ultrapure reference 1kHz sine oscillator bare circuit THD 0.000015% (-135dB) | eBay
Opinions/experiences?
Jan
Ultrapure reference 1kHz sine oscillator bare circuit THD 0.000015% (-135dB) | eBay
Opinions/experiences?
Jan
nice find.
looks impressive and reasonably priced pcb for the level of performance claimed. unfortunately, I have no gear to evaluate it, but I want one anyway.
🙂
mlloyd1
looks impressive and reasonably priced pcb for the level of performance claimed. unfortunately, I have no gear to evaluate it, but I want one anyway.
🙂
mlloyd1
Hi Frex,
Yes, that is the same as the board I got from you.
Curious, but I haven't asked yet. The 10 KHz version was built as a 10 KHz oscillator, but it is running at 5 KHz. Thoughts? The 1 KHz version runs at the proper 1 KHz frequency.
-Chris
Yes, that is the same as the board I got from you.
Curious, but I haven't asked yet. The 10 KHz version was built as a 10 KHz oscillator, but it is running at 5 KHz. Thoughts? The 1 KHz version runs at the proper 1 KHz frequency.
-Chris
Hello Chris,
Very curious, i built both for me, and they run fine at 1kHz and 10kHz.
There is probably a mistake in RC network value.
I will check on my side soon if there is no problem in the 1kHz bom.
Frex
Very curious, i built both for me, and they run fine at 1kHz and 10kHz.
There is probably a mistake in RC network value.
I will check on my side soon if there is no problem in the 1kHz bom.
Frex
What passive notch does not need gain correction (second picture 2nd and 3rd harmonics are read directly off of screen). I might have missed one of Samuel's musings.
The 2nd may be a few dB optimistic due to the passive notch, but still 1st class performance I think.
Frex, if I would build this I would make it switchable for a couple of spot frequencies, make sure the level control is a good quality pot with a knob rather than a trimpot, and also include a balanced output connector, an XLR or better yet a combination XLR/TRS.
And of course power it with a SilentSwitcher ;-)
Any chance you will make an alternative packaging for some standard box with panel drawings for such a 'deluxe' version?
Jan
Frex, if I would build this I would make it switchable for a couple of spot frequencies, make sure the level control is a good quality pot with a knob rather than a trimpot, and also include a balanced output connector, an XLR or better yet a combination XLR/TRS.
And of course power it with a SilentSwitcher ;-)
Any chance you will make an alternative packaging for some standard box with panel drawings for such a 'deluxe' version?
Jan
What I find amazing is that you can get this performance with opamps that are not exactly very low distortion.
BTW, on the graphs, it says 'passive notch filter (10dB attenuation)'. What does that 10dB attenuation mean?
Jan
BTW, on the graphs, it says 'passive notch filter (10dB attenuation)'. What does that 10dB attenuation mean?
Jan
Composite op-amps, Jim Williams and I exchanged a few mails on this and the original "distortionless" claim. B. Oliver described the problem in wonderful maths when the original HP oscillators sometimes failed to stabilize when the distortion was too LOW. The trade off is difficulty in trim as the distortion is reduced, indeed getting the parts per billion in AN67 would be difficult to hold.
EDIT - Dave and I crossed. I might add that few people realize that the light bulb oscillator has two mechanisms working and it's stability is due to the frequency independent voltage coefficient of resistance not the obvious thermal one. That's why a couple of years ago I showed the Christmas tree bulb oscillator which gets 20dB better than the common instrument bulb usually used.
Last edited:
That's LT's super gain block oscillator. Nothing new here. Someone just packaged it up.
Victor's does better at a lower cost. This oscillator can be rather unstable.
Hi Scott,
I have a number of #382 lamp type oscillators. Since they already exist, I'd love to play with them to see how good they could be made. What x-mas tree lamp did you use?
-Chris
I have a number of #382 lamp type oscillators. Since they already exist, I'd love to play with them to see how good they could be made. What x-mas tree lamp did you use?
-Chris
Is this to say the higher voltage lamps have a lower voltage coefficient? I never looked so don't know.
Hi Frex,
I was very careful to build it following your BOM. Mostly Digikey parts. Needless to say, I was surprised to see 5 KHz. That's not a welcome frequency since I use an HP 339A, and switching a decade in frequency is easy. 5 KHz isn't.
They both work, but the 2nd harmonic is higher than it should be as Jan noted.
Jan, you could do that with a normal board. Just run the wires to your switching assy from the PCB, same for the level control.
-Chris
I was very careful to build it following your BOM. Mostly Digikey parts. Needless to say, I was surprised to see 5 KHz. That's not a welcome frequency since I use an HP 339A, and switching a decade in frequency is easy. 5 KHz isn't.
They both work, but the 2nd harmonic is higher than it should be as Jan noted.
Jan, you could do that with a normal board. Just run the wires to your switching assy from the PCB, same for the level control.
-Chris
Red 😀
I have to check the number, but it's the standard 2" or so high one shaped like a candle flame.
Hi Chris.
If you go back a zillion posts ago you can read about my lamp multiplier. I don't know if you where involved with this thread that long ago.
It might be possible to use higher voltage lamps. The key is to keep the temperature of the filament down to avoid the sensitivity of a hot lamp.
It does require a descent R delta to work though
Last edited: