Safe AC for exercises.

Wendy

Joined Mar 24, 2008
23,415
A light bulb has negative resistance, and it is relatively linear. The more current you put through it the higher its resistance.

In the Wien Bridge oscillator, if the signal increases the resistance increases, and drops the gain of the circuit, lowering the signal. It is a form of AGC loop. The light bulb does not even have to glow, or be near its rated voltage for the effect to manifest.

The down side is if the frequency of the oscillator gets too low, where the filament begins to warm and cool with the waveform, severe distortion sets in. Otherwise a very pure sine wave is the result, as the filament achieves an RMS temperature.

You could use simple diodes to get much the same effect, or just set the gain of the circuit without an AGC, though it not nearly as good. Just adjust the sine wave with an Oscope.
 

Thread Starter

hp1729

Joined Nov 23, 2015
2,304
A light bulb has negative resistance, and it is relatively linear. The more current you put through it the higher its resistance.

In the Wien Bridge oscillator, if the signal increases the resistance increases, and drops the gain of the circuit, lowering the signal. It is a form of AGC loop. The light bulb does not even have to glow, or be near its rated voltage for the effect to manifest.

The down side is if the frequency of the oscillator gets too low, where the filament begins to warm and cool with the waveform, severe distortion sets in. Otherwise a very pure sine wave is the result, as the filament achieves an RMS temperature.

You could use simple diodes to get much the same effect, or just set the gain of the circuit without an AGC, though it not nearly as good. Just adjust the sine wave with an Oscope.
Thank you for that.
 

Thread Starter

hp1729

Joined Nov 23, 2015
2,304
A light bulb has negative resistance, and it is relatively linear. The more current you put through it the higher its resistance.

In the Wien Bridge oscillator, if the signal increases the resistance increases, and drops the gain of the circuit, lowering the signal. It is a form of AGC loop. The light bulb does not even have to glow, or be near its rated voltage for the effect to manifest.

The down side is if the frequency of the oscillator gets too low, where the filament begins to warm and cool with the waveform, severe distortion sets in. Otherwise a very pure sine wave is the result, as the filament achieves an RMS temperature.

You could use simple diodes to get much the same effect, or just set the gain of the circuit without an AGC, though it not nearly as good. Just adjust the sine wave with an Oscope.
Got the bulbs. Built the circuit per attached drawing. Adjusted the pot for a nice sine wave. Amplitude was lower than the J-FET design, but it does work. No, the lamps don't light up. :)
 

Attachments

Wendy

Joined Mar 24, 2008
23,415
Looks good. If you were to use an oscilloscope on the power drivers you will see some crossover distortion, which is why I went through my contortions. It is the bugaboo of a lot of circuits. Even some op amps, such as the quad LM324, will show this glitch with no other circuitry. You can think of this op amp as a low voltage quad 741.

For your application is will be pure enough. If you were using it as a audio amp you might be able to hear it. I am not an audiophile, so I don't hear what a lot of folk claim to. I do believe what my instrumentation shows me though.

You can measure R30 and replace it with a fixed resistance. You would be surprised how close you can get to any resistance with two 5% resistances in parallel.

It is funny, I did not realize you were using the LM324 until I finished the above comments. I am using a very high speed audio preamp op amp, the NJM4565. It is a dual op amp, same layout as the 1458.
 

Thread Starter

hp1729

Joined Nov 23, 2015
2,304
Looks good. If you were to use an oscilloscope on the power drivers you will see some crossover distortion, which is why I went through my contortions. It is the bugaboo of a lot of circuits. Even some op amps, such as the quad LM324, will show this glitch with no other circuitry. You can think of this op amp as a low voltage quad 741.

For your application is will be pure enough. If you were using it as a audio amp you might be able to hear it. I am not an audiophile, so I don't hear what a lot of folk claim to. I do believe what my instrumentation shows me though.

You can measure R30 and replace it with a fixed resistance. You would be surprised how close you can get to any resistance with two 5% resistances in parallel.

It is funny, I did not realize you were using the LM324 until I finished the above comments. I am using a very high speed audio preamp op amp, the NJM4565. It is a dual op amp, same layout as the 1458.
Re: NJM4565
Better chip at a good price!
Just not something I have on hand at this time. Is there some reason you didn't go with a FET input?
 

Wendy

Joined Mar 24, 2008
23,415
FET op amps have an odd property. If you get near the power supply voltages on the inputs they go through a phase inversion, which is a very funky and weird severe distortion unique to these families. Since I tend to push the boundaries quite regularly I just don't mess with them, but for most uses they are fine.

The LM358 (a dual op amp) will work all the way down to 0V on the inputs. Most op amps don't work well when their inputs are near the power supply voltages.

I have mentally cataloged the idiosyncrasies of various op amps over the decades. As I have gotten older I have found it is not what you remember that is important, but what you forgot.

My blog points to a chart I have made about op amps.

High Speed Op Amp Query
 

Thread Starter

hp1729

Joined Nov 23, 2015
2,304
FET op amps have an odd property. If you get near the power supply voltages on the inputs they go through a phase inversion, which is a very funky and weird severe distortion unique to these families. Since I tend to push the boundaries quite regularly I just don't mess with them, but for most uses they are fine.

The LM358 (a dual op amp) will work all the way down to 0V on the inputs. Most op amps don't work well when their inputs are near the power supply voltages.

I have mentally cataloged the idiosyncrasies of various op amps over the decades. As I have gotten older I have found it is not what you remember that is important, but what you forgot.

My blog points to a chart I have made about op amps.

High Speed Op Amp Query
Thanks, nice chart.
 
Top