Has anybody ever tried to figure them out? I recently had an opportunity to hear an Accuphase C-2800 and I really liked the volume control: Feels as precise as a pot, but no mechanical wear, and the sound quality obviously was much better than all those Crystal / TI volume chips or the DAC approach.
What they do is V/I-convert the signal in 16 binary weighted currents, switch those currents, add them up and I/V-convert them back.
The C-2800 seems to be an all discrete device, but the lower models like C-2000 have all opamps except the output stage.
What do you think of this kind of volume control? Worth reverse-engineering it?
What they do is V/I-convert the signal in 16 binary weighted currents, switch those currents, add them up and I/V-convert them back.
The C-2800 seems to be an all discrete device, but the lower models like C-2000 have all opamps except the output stage.
What do you think of this kind of volume control? Worth reverse-engineering it?
sample schematic
Hi,
I tought about it a while and came to this solution. It is my guess and what I would start with. I have no clue how similar my solution is to the original.
I attached 2 files: a picture and a zip file containing a simulation file for LTspice and the model fot the jfets.
The parts used and the values are not especially chosen. So it is up to you to choose optimal sounding parts and values. The trick is, that the current summing stage U5 has an input potential of zero volts. Thus the V-I-converts are simple buffers with a resistor connected to the input of U5. For example the output current of branch 1 is voltage V1/R1 and the output current of branch 2 is voltage V1/R2. Both currents flow into the current summing point at the minus input at U5 and then they all together flow into R5, giving the output voltage, which hence is the I-V-converter. The JFETs are used as the signal switches. If their gates are 0V, they are on, if tey are negative, they are off.
Only 4 instead of 16 branches are shown but you can add the missing 12 easily.
Each branch has got it's own switch. Only switch 4 is switched on.
If you want to change the volume, change the settings of the switches.
Have fun.
Peter
Hi,
I tought about it a while and came to this solution. It is my guess and what I would start with. I have no clue how similar my solution is to the original.
I attached 2 files: a picture and a zip file containing a simulation file for LTspice and the model fot the jfets.
The parts used and the values are not especially chosen. So it is up to you to choose optimal sounding parts and values. The trick is, that the current summing stage U5 has an input potential of zero volts. Thus the V-I-converts are simple buffers with a resistor connected to the input of U5. For example the output current of branch 1 is voltage V1/R1 and the output current of branch 2 is voltage V1/R2. Both currents flow into the current summing point at the minus input at U5 and then they all together flow into R5, giving the output voltage, which hence is the I-V-converter. The JFETs are used as the signal switches. If their gates are 0V, they are on, if tey are negative, they are off.
Only 4 instead of 16 branches are shown but you can add the missing 12 easily.
Each branch has got it's own switch. Only switch 4 is switched on.
If you want to change the volume, change the settings of the switches.
Have fun.
Peter
Attachments
Hi!
This seems to be very old topic but just for curiosity, has anyone tried this? Or anything similar?
Based on pr's simulation model, I would use T-switch structure to replace the single switch for better off-isolation. And preferably go for fully balanced version. In addition it would require some sort of logic to recognize zero crossing from input signal, otherwise there will be annoying noises when switches operate.
I've been thinking about building pre amp with "an exotic" volume control and found Accuphase's AAVA documentation for a some time ago. Something like this combined with input selector and AV-loop would solve problems with two sets of equipment (HT and stereo) but only one pair of speakers.
This seems to be very old topic but just for curiosity, has anyone tried this? Or anything similar?
Based on pr's simulation model, I would use T-switch structure to replace the single switch for better off-isolation. And preferably go for fully balanced version. In addition it would require some sort of logic to recognize zero crossing from input signal, otherwise there will be annoying noises when switches operate.
I've been thinking about building pre amp with "an exotic" volume control and found Accuphase's AAVA documentation for a some time ago. Something like this combined with input selector and AV-loop would solve problems with two sets of equipment (HT and stereo) but only one pair of speakers.
Well, I wouldn't say it's a DAC nor ADC. Operating principle of course is somewhat digital, but signal is not converted from one domain to another.
I have been looking for solution that could be used also with DAC without compromising the signal quality. Digital multiplier inside the DAC chip or in the DSP is not the optimal solution and neither is the traditional analog methods. I've been using those PGA chips for some time and looking for something new to experiment with.
I have been looking for solution that could be used also with DAC without compromising the signal quality. Digital multiplier inside the DAC chip or in the DSP is not the optimal solution and neither is the traditional analog methods. I've been using those PGA chips for some time and looking for something new to experiment with.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.