Balanced Preamp with discrete solid state volume control.

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi All,

I'm thinking about designing and building a pre-amp based on the attached skeleton schematic.

I would like the pre-amp to have the following features,

- Fully Balanced from input to output.
- Discrete Solid State Volume Control.
- Output able to drive headphones.
- Dual Mono construction.
- Input Selection with relays.
- Full remote control functionality.
- Controller with ATMEL AVR uC for Remote decoding, display driving etc.

From the schematic you can see it is based on a BJT shunt type volume control. I've shown only a 6 bit design, I haven't done the maths to figure out how many would be required to get decent enough resolution in practice. I guess 8 bits would probably give fine enough volume resolution.

Just a quick overview of the operation. The upper 6 PNP BJTs are turned on/off by input control signals. The PNP devices drive current through the Bases of the NPN shunt devices turning them hard on. When the PNP devices are off the bases of the shunt devices are pulled to the negative supply rail by pull down resistors. Note that it does not work if the bases are only pulled to ground.

I think that using BJTs offers a couple of important advantages over FET type analog switches.
1 - The BJTs have a lower on resistance.
2 - BJTs have a lower capacitance giving the circuit a wider bandwidth.

I know I've shown op-amps at the input and output, this may not please some of you but I think that there are some pretty good devices out there so I'll probably stick with the op-amp approach but with suitable devices for driving headphones on the output. Please though lets not turn this thread into a debate about op-amps Vs. discrete or which op-amp is best. There are plenty of those sort of threads around.

I've played around with some simulations of the BJT shunt design at the results look promising, very low distortion especially with a fully balanced approached.

Cheers,

Dave
 

Attachments

  • balpre.zip
    4.8 KB · Views: 685
AX tech editor
Joined 2002
Paid Member
Hi Dave,

One issue to think about is that BJT's in their off-state have a capacitance that is non-linear with the instanteneous signal level. That is one of the reasons that they should not be used as mute switches in DACs and other equipment: many people claim that they can hear the effects.

Jan Didden
 
Hi Jan,

Thanks for the response.
I'm aware that this circuit does have a close similarity to those mute circuits which are not really considered to be an audiophile solution.
However comparing it to other solid state volume control solutions such as MDACs or CS3310 I'm hoping the BJTs will still out perform the FET based switches used in these devices.

This is an interesting solution from Mark Levinson,

http://www.marklevinson.com/image_library/32AB_lo.jpg

it uses as far as I can tell from the high res image these
switching devices -
http://www.vishay.com/analog-switches/list/product-70662/

Again I think the BJTs will perform better given a much lower on resistance to ground, capacitance in the BJTs should be similar or less.

I'm really keen to get further feedback from people here on this design.
 
AX tech editor
Joined 2002
Paid Member
Sorry, but I think BJT's are about the worst choice you can make here. The integrated solutions are FAR better. BJT switches always have an offset voltage when on, and a varying cap when off. A good FET switch can have sub-1 ohm on resistance, which becomes important at high attenuation steps, and extremely low Ron modulation when used intelligently.
You may want to use a specific solution for your own reasons, but the technical reasons are not good.

Jan Didden
 
If you are trying to use the BJT's to attenuate the signal, I would have to aggree with Jan... It will be easier to just change the DC voltage on the gate of a JFET & use as a DC voltage controled voltage divider circuit than with BJT's. You can graph out the operation and adjust the DC voltage on the potentiometer for the gate to attenuate on a logrithmic scale as is needed for linear audio volume control. As for the OP Amps, I don't see any problem there, but you may want to use them as just unity gain voltage followers to limit the typically poor slew rate affects. Just connect the output to the inverting input.
 
Hi Cunningham,

The BJTs are operated strictly as switches either fully on or fully off - so the attenuation is in discrete steps. This is certianly not supposed to be a voltage controlled amplifier circuit. I've never seen one of those used in a audio and probably for good reason.

I think I'd keep the op-amp setup as is, they serve as differential amplifiers keeping in line with the balanced topology. This gives improved rejection of noise and distortion both at the input signal and at the output of the attenuator circuit.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.