Balanced Volume Controller / Line Stage

Member
Joined 2009
Paid Member
What I want to know is, for instance, if I have a DAC outputting 1.2V RMS, I assume at minimum volume that will be 0V, at some point, it will be no attenuation so 1.2V, and from some point we get some gain, idk how much but let's imagine 2V?
+15.5dB is about x6 V/V gain, so at the maximum volume setting, 1.2Vrms at the input would give you ~7.2Vrms (~20Vpp) at the output.
...do these digital pots "have a sound"?
While "having a sound" is subjective, it is totally possible to make a sonically transparent opamp stage - hope I won't get too many rotten tomatoes for saying this. As for digital pots proper, they are used in all kinds of hi-end consumer equipment, so I figure they must be ok sonically. The secret sauce is usually the implementation.
 
+15.5dB is about x6 V/V gain, so at the maximum volume setting, 1.2Vrms at the input would give you ~7.2Vrms (~20Vpp) at the output.
While "having a sound" is subjective, it is totally possible to make a sonically transparent opamp stage - hope I won't get too many rotten tomatoes for saying this. As for digital pots proper, they are used in all kinds of hi-end consumer equipment, so I figure they must be ok sonically. The secret sauce is usually the implementation.

Thank you so much again for your patience, I will build one of these
 
hey Alex,

hopefully last stupid question ;)

does the gain of the power amp after the digital volume controller change the “effective” dB step size? let me explain. if I placed the controller just before the speakers (let’s forget about impedances for a sec), the 0.5dB steps are clear, but if I place it in the preamp, shouldn’t those 0.5dB steps size get factored by the power amp gain?

Thank you
 
You can not place volume controller at the power amp output, only at the power amp input. Volume controller only attenuates signal that the power amp receives. Power amp gain is usually fixed unless power amp has input pot.

of course, i was just trying to describe the gain question, which still remains the same.

the digital pot doesn’t know the gain stage(s) after it, how does it(them) affect the effective 0.5dB steps?
 
Member
Joined 2009
Paid Member
dB is a ratio, whatever gain after the pot will keep the ratio of the change in volume between two steps.
That’s absolutely correct. The power amp’s gain won’t affct the ratio, although the absolute level in e.g. volts will change.

Note that many volume controls, even digital ones, do not have a dB scale, and often have no scale at all. Look at the volume control in a typical smartphone, for example, or on a TV. While having dB reading gives a sense of precision and may even seem cool, it may also be too much information for a typical use case.
 
Member
Joined 2009
Paid Member
PGA2310 board schematic and BOM

I was asked to publish the schematic and the BOM for the PGA2310 balanced volume control board from post #3. The schematic is attached. The BOM is available as a shared project at Mouser. Note that many parts can be substituted, e.g. the opamps included in the BOM are LM4562, but other dual opamps can be used, e.g. NE5532 or even TL072 (that one with a lowered performance); other brands of electrolytic caps can be used as long as they have the same diameter and lead spacing; voltage regulators and transistors can be replaced with those you may have, etc.
 

Attachments

  • bvc_gs_2310_sch-Lch,PSU.png
    bvc_gs_2310_sch-Lch,PSU.png
    442.1 KB · Views: 919
  • bvc_gs_2310_sch-Rch,SPI.png
    bvc_gs_2310_sch-Rch,SPI.png
    428.2 KB · Views: 899
Last edited:
I was asked to publish the schematic and the BOM for the PGA2310 balanced volume control board from post #3. The schematic is attached. The BOM is available as a shared project at Mouser. Note that many parts can be substituted, e.g. the opamps included in the BOM are LM4562, but other dual opamps can be used, e.g. NE5532 or even TL072 (that one with a lowered performance); other brands of electrolytic caps can be used as long as they have the same diameter and lead spacing; voltage regulators and transistors can be replaced with those you may have, etc.
Thank you!

Would it be possible to get the schematics in the .sch type, not an image?

Also, where do you buy the WM8816 chips? I think they're not available anywhere :(
 
Member
Joined 2009
Paid Member
Why so many opamps

The schematics posted above raise questions on why so many opamps for a simple volume control, and why just not limit yourself to one, or better no opamps at all?

The question is valid, and of course, alternatives exist - see, for example, post #27 above.

In the attached schematic, however, each opamp has a role:
  • U3 together with R5,R6,R7 provides high (2.2Meg) common mode input impedance and reasonably high (~100k) differential mode impedance. Besides lighter load for the signal source, this helps to mitigate the imbalance in the source's output impedances on the "cold" and "hot" wires.
  • U4 compensates for the DC component of the input signal, as well as for the DC offset that is created by U3's input currents flowing though R5.
  • U5A is the difference amplifier, which calculates the input signal as the difference between the "cold" and the "hot" inputs.
  • U6 is the volume control proper; its gain is controlled by the WM8816.
  • U5B is the ground sense output stage that allows the following stage (say the power amp) to be referenced to its local ground (that is, to have a single ended input) without creating e.g. ground loops.
 

Attachments

  • bvc+gs_8816_sch-Lch+PSU.png
    bvc+gs_8816_sch-Lch+PSU.png
    185.8 KB · Views: 1,098
Hello again,

Thank you so much for all the details. I really need to digest all the info, too complex for me at the moment. Just a pair of questions though.

1. Looks like providing balanced in for each channel is compulsory as it is explained above, right?

2. If that is not the case, if we can input unbalanced V+ for each channel, and if (big if) we already have passed the signal through an active filter (opamp based), impedances would already be OK, so wouldn't it be possible to start at U5 position?

Sorry again, I know these are very newbie and maybe stupid questions, but I'd like to learn a bit more.

Also, again, not trying to criticise the design at all.
 
Member
Joined 2009
Paid Member
The article that started this whole project can be found at edn.com. In it, Bruno Putzeys comprehensively and thoroughly explores the balanced/unbalanced/grounding topic and, I believe, answers your questions better than I can. Spoiler: he suggests that "the RCA connector and all it stands for should be banned by law".
 
Last edited:
The article that started this whole project can be found at edn.com. In it, Bruno Putzeys comprehensively and thoroughly explores the balanced/unbalanced/grounding topic and, I believe, answers your questions better than I can. Spoiler: he suggests that "the RCA connector and all it stands for should be banned by law".

Yeah I know about the excellent article. I think there’s a misunderstanding tho, in question 1 I was referring to the balanced signal before the volume controller, not after. I will create the balanced out at the exit of the preamp, not in the DAC, as it seems to be the case here, if I understood correctly.
 
Member
Joined 2009
Paid Member
Yeah I know about the excellent article. I think there’s a misunderstanding tho, in question 1 I was referring to the balanced signal before the volume controller, not after. I will create the balanced out at the exit of the preamp, not in the DAC, as it seems to be the case here, if I understood correctly.
Sorry I did not understand your context. If you want a volume control to be connected to the output of the active filter following your DAC and follow the logic of Bruno Putzey's article, then:
  • You can probably omit U3 and U4 and the associates circuitry if your active filter can drive 1kohm load directly.
  • If you can have your active filter use the volume control's local "ground" (the point where R19 and R21 connect) as the active filter's output reference point, the way U5A uses it as its output reference, then the active filter can replace U5A and everything before it.
  • If you active filter uses some other reference point for its output, you'd need to keep U5A.
  • You can omit U5B if you do not need the ground sensing feature.
  • When you take stages out, keep in mind that the absolute phase of the output signal should be preserved.
It would be helpful if you posted the schematic of your DAC's analog output.
 
Last edited:
Member
Joined 2009
Paid Member
How to control a digital volume control

Resurrecting my own thread.

Having built a bunch of various digitally controlled volume controls (with PGA2310, WM8816, MUSES72320 and DS1882 above, plus an unpublished relay switched attenuator I made a while ago for my implementation of Mauro Penasa's MyRef C), I am struggling with finding a good way to provide a good "universal" front panel control.

In terms of firmware, it is relatively simple. I already have a rotary encoder controlled hardware/firmware prototype (post #25 above) that:
  • Controls any of the above mentioned chips (separate firmware for each chip)
  • Can be controlled with an Apple remote (aluminum or the original white, not the latest black touchpad type) and supports remote pairing
  • Keeps volume setting when the power is down
  • Supports an optional I2C connected 16x2 display (any with Adafruit's backpack or the one on the photos below) to show volume setting
  • Can be powered by any of the balanced volume controls discussed above
  • Fits, together with the display, into a 40mm/1U high enclosure
  • With a little more coding, a second board can easily control input selector relays and use the same display
What I am struggling with are the form factor and mounting options.

I want to get away both from prescribing the mutual positions and sizes of front panel controls and from designing and fabricating a separate control PCB for every project - I don't want my (or your) amp to look like someone else's!

My prototype is designed to hold on to a rotary encoder, which attaches to the front panel with its bushing and a nut. It is a flexible solution, it can be placed anywhere on the front panel, and it only needs one hole. It works if you connect it with a flat cable to one of my balanced volume controls. It even resembles a regular pot if you piggyback on it my DS1882 board (although it requires power supply unlike a regular pot). It is fairly compact at 1.1x1.1 inch, close to the size of a (better) pot.

However, using the encoder's bushing is not the only way. In fact, cheap encoders often have no bushing, and those having it may have no anti-rotation lugs. They are designed to be mounted on a PCB, not to have a PCB mounted on them. Adding mounting holes requires bigger boards (1.35x1.35 inch or 35x35mm seems to be the smallest size that can mount around a 30mm/1.2inch knob), as well as provisions (studs of tapped blind holes) on the front panel. The latter can be provided by e.g Schaeffer/FrontPanelExpress, but at a cost.

Also, using a rotary encoder is not the only way. A simple (linear) pot is another. It dispenses with remote control but provides a more "natural" feeling with stops at the minimum and maximum volume.

I have been thinking and prototyping but have not settled on a better solution.

In your view, what's a good universal digital front panel control that would work well in your project?
 

Attachments

  • hifiocean - 47.jpg
    hifiocean - 47.jpg
    159.8 KB · Views: 834
  • hifiocean - 35.jpg
    hifiocean - 35.jpg
    382.4 KB · Views: 814
Last edited: