Having done the same the Wien bridge is a nice way to do it. I seem to remember I used a 28V 40mA lamp and 680R. Distortion was about 0.05%.. A small resistance can be added to the lamp to get slightly better distortion ( 100 R ? ). The 680R increased. Eventually stable oscillation is lost if going too far.
I always wondered if a power amplifier could be made into a Wien oscillator. I imagine a simple class A amplifier would be able to do this. The Linn Valhalla seems to be class A. 320VDC that finds 90Vrms for the motor. The Valhalla is let down by it's oscillator power supply. It uses a CD4060 and filter with 3.2768 MHz crystal. The circuit is a masterpiece of simplicity. Valhalla THD is 0.05%. Later Valhalla used 80 Vrms. Rumour has it Ariston had a similar unit further back in time. Ray Collins of Ariston told me. I doubt it was untrue. He said like many things they did not pursue it when others did.
I always wondered if a power amplifier could be made into a Wien oscillator. I imagine a simple class A amplifier would be able to do this. The Linn Valhalla seems to be class A. 320VDC that finds 90Vrms for the motor. The Valhalla is let down by it's oscillator power supply. It uses a CD4060 and filter with 3.2768 MHz crystal. The circuit is a masterpiece of simplicity. Valhalla THD is 0.05%. Later Valhalla used 80 Vrms. Rumour has it Ariston had a similar unit further back in time. Ray Collins of Ariston told me. I doubt it was untrue. He said like many things they did not pursue it when others did.
Thank you tiefbassuebertr
Good food for the mind.
Hi Nigel
The wien bridge can be good but It must be very well calculated to be stable in the frequencies of interest.
In my case I experimented with severall jfets and corresponding biasing circuit until I got stability from 35Hz to 2khz.
Good food for the mind.
Hi Nigel
The wien bridge can be good but It must be very well calculated to be stable in the frequencies of interest.
In my case I experimented with severall jfets and corresponding biasing circuit until I got stability from 35Hz to 2khz.
Last edited:
The main thing with the Wien is it is very stable once running. 50 ppm resistors like MRS25 and COG capacitors almost perfectly match each other.
The Linn way is a tiny bit easier if wanting correct frequency without adjusting.
The State Variable Filter Oscillator is much like a Wien. If using 3rd harmonic cancellation simple diode limiters are enough. I remember ESP Audio having one. Didn't find it. One advantage of the SVF is that a sine and cosine section is available. JLH showed a sine cosine Wien. Back in the day that got the most from op amps of the time ( common mode improvement ). .
The Linn way is a tiny bit easier if wanting correct frequency without adjusting.
The State Variable Filter Oscillator is much like a Wien. If using 3rd harmonic cancellation simple diode limiters are enough. I remember ESP Audio having one. Didn't find it. One advantage of the SVF is that a sine and cosine section is available. JLH showed a sine cosine Wien. Back in the day that got the most from op amps of the time ( common mode improvement ). .
Wien oscillators are sensitive to temp variation. VPI used a Wien bridge osc in their ADS power supply and the frequency drifted up ~0.25Hz with temp over 10-15 minutes:
ADS temp drift
The stability was approximately the same as an average wall socket, but the accuracy was many times worse.
ADS temp drift
The stability was approximately the same as an average wall socket, but the accuracy was many times worse.
That shouldn't happen. I had a clock correct to 7 seconds a day on a Wien. The trick is NPO caps that are 30 ppm. Folklore has it they are best suited to 1% resistors. Crystals of the type Linn used are 30 ppm. Polyester are often 200 ppm and polystyrene between. A 10 nF COG/NPO costs pennies. TL072 a good op amp at 50 Hz. Silver mica are posh NPO.
The difference I believe, is the clock probably didn't generate much heat. The ADS amps had heat sinks that were internal to the case and the air temp inside the enclosure could exceed 60°C when powering an 7.5W motor; I could watch the frequency change as the temp of the supply increased. At room temp (or with the top cover off) this did not occur.
No. I have had that problem and it affected voltage more. If you take a remarkably small amount of trouble the problem vanishes.
What is noticeable is stability takes a few moments to go from good to excellent. To power the oscillator 24/7 is a low cost option. If using diode stabilisers LED are a little better.
I used a Racal frequency meter to check. It's crystal in a small oven. For 50 Hz is says 20 mS . Nixie tube display.
What is noticeable is stability takes a few moments to go from good to excellent. To power the oscillator 24/7 is a low cost option. If using diode stabilisers LED are a little better.
I used a Racal frequency meter to check. It's crystal in a small oven. For 50 Hz is says 20 mS . Nixie tube display.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.