Here is an idea for a voltage-controlled volume.
It is based on PWM, and has the potential to be more linear than conventional VCA's.
The PWM signal is generated by by comparing the control voltage against an exponential waveform, yielding an inherently logarithmic function.
The circuit shown in example is just a (working) proof of concept: to make it usable in the real world, the swtching frequency would need to be increased, by a factor 5 or 10 at least.
In this prototype, the frequency is only 133KHz and is limited by the LM339.
The switch would also need to be a better one than the CD4053: at least a CD40HC4053, but preferably a modern, low charge injection one like the LTC1043.
With quality components, there would be no sources of non-linearity in the signal path, except for the intermodulation products between the signal and the sampling frequency, and they fall very rapidly if the sampling frequency is pushed higher.
In this circuit, the control voltage varies between -6 and +6V, giving a nice and smooth 0 to 100% logarithmic variation.
The particular arrangement of ground, control voltages and supplies is just an example, one could adapt it to obtain a 0 to 5V control range for instance.
It is based on PWM, and has the potential to be more linear than conventional VCA's.
The PWM signal is generated by by comparing the control voltage against an exponential waveform, yielding an inherently logarithmic function.
The circuit shown in example is just a (working) proof of concept: to make it usable in the real world, the swtching frequency would need to be increased, by a factor 5 or 10 at least.
In this prototype, the frequency is only 133KHz and is limited by the LM339.
The switch would also need to be a better one than the CD4053: at least a CD40HC4053, but preferably a modern, low charge injection one like the LTC1043.
With quality components, there would be no sources of non-linearity in the signal path, except for the intermodulation products between the signal and the sampling frequency, and they fall very rapidly if the sampling frequency is pushed higher.
In this circuit, the control voltage varies between -6 and +6V, giving a nice and smooth 0 to 100% logarithmic variation.
The particular arrangement of ground, control voltages and supplies is just an example, one could adapt it to obtain a 0 to 5V control range for instance.
Attachments
Yes, that would be interesting: knowing the take of a great mind on a problem is always enlightening.
Note that I certainly do not claim any originality in the concept: it is part of the common knowledge body of electronics.
I think that the use of an exponential reference waveform is particularly relevant in this case (perhaps not original either, I didn't research the subject).
Here are some pics: the first waveform is the exponential signal.
The second is the output signal with a LF triangle input, and the volume set at 50%. Note the incomplete filtering of the carrier.
Following pics show the prototype board.
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.