advertisement


The effect of volume control

philiphifi

pfm Member
hi - I noticed something very odd yesterday. On my Audion monoblocs, there is a volume control at the front panel for each device. If i turned that to max (to bypass the resistor) and use the preamp to control the volume adjusted to the same loudness, the sound is very different. (the preamp is used in both cases). More bass and treble. How is this possible? A resistor acts as a filter or is it that the preamp is sounding different at different volume settings?
 
Sounds like your amp has a significant capacitance, etc, right after the controls you mention. The output resistance of the control varies with the setting.

A shunt capacitance will combine with the output resistance of the control to produce an HF filter. A series capacitance means the LF may change. Thus the frequency response can vary with the control setting. The result should be flattest at the max and min settings, and least flat when (approx) halving the input voltage.
 
thank you Jim. Just so i understand properly, is the best way then to bypass the power amp resistor and use the preamp to control the volume? I don't understand the halving the input voltage part. I would need to halve the preamp volume but would this lead to the same problem?
 
If you have an 'active' preamp it will have a gain/buffer section after it's volume control. So your power amp won't know anything about the change in output resistance of the preamp's volume control. Only the change in level.

'Passive' pre-amps can produced the same effect, though. Sometimes made worse by the capacitance of the connecting cable.

The snag is that your *preamp* might do the same sort of thing! The behaviour is almost impossible to totally prevent, but good design should ensure it is too small to be audible.

So it is worth setting the power amp control to 'max' and trying just using the preamp volume control. That should be OK. But can't be sure without various details or measurements.
 
This is an advantage of digital volume control, another one is perfect channel balance. (There are disadvantages, but if done properly, nothing I worry about.)
 
This is an advantage of digital volume control, another one is perfect channel balance. (There are disadvantages, but if done properly, nothing I worry about.)

There are no advantages I can think of and a variety of disadvantages, depending hugely on how exactly a digital vol control is designed. A switched attenuator has perfect channel balance.
 
I had a similar thing happen. I bought a couple of cheap class D amplifiers (to power some exciters for vibration experiments) and decided to try them on a loudspeaker I am developing. The sound was much better than expected. Then, as I was assessing things with a view to TV sound, I wanted remote control, so built one using an Alps motorized pot. Connected to the class D amplifier, I bypassed the onboard pot (turned it to max volume) and used the Alps pot to control volume. There was a definite improvement, which I put down to better component quality.
 
The larger (physically) ALPs pots tend to be good. However this also depends on the choice of pot end-to-end resistance and the cable capacitances, etc.
 
The Alps Blue Velvet became pretty much the industry standard for vol pots, and for good reasons, such as first and foremost excellent channel balance, also it lasts a good while before any issues of crackles etc and it has a good "feel" to it in use.
However, I was surprised as to how much better a good switched attenuator can sound! I have no technical explanation for this but the difference was very obvious to me... greater transparency.
 
good to know that a good switched attenuator can sound so good. I have one (double gang) to put in my unity gain line stage and single gangs in two stereo (current feedback) amplifiers I am building, the 'pots' will adjust levels in one of the two amplifiers that feed different mid range units. [Sorry, very difficult to describe]
 
The main advantage is avoiding a preamp entirely, if you have just a digital source.

And usually have all sorts of other problems from the digital volume control itself, which will be far from the near perfection of a stepped attenuator.... and probably worse than an active pre/vol control also. There are digital vol controls.. and digital vol controls...
 
The process of simply rescaling LPCM vlues should be pretty easy and accurate if you use 32 bit (or bigger) values and dither the calculations. Far more accurate that any analogue pot. The problem may arise when values that are well scaled down go though the DAC and output low analogue values. Thus exposing them to the noise levels of the output side of the conversion, DAC word size, etc.

Ideally, you'd put the volume adjustment right up against the power amp to shorten the 'low level' path.

What I like(d) about the old 40mm Alps pots with detent was they could provide very accurate level settings. In particular maintaining very accurate channel balance as you altered the volume. This avoided changes in balance affecting the sound. I really regret they ceased being made. But I assume the close-matched ones were too expensive for most makers to choose.
 
I disagree. Controlling volume in the analogue domain is vastly better and more accurate than doing it digitally. By accurate I do not mean just channel balance but everything.. THD, noise the lot. A stepped attenuator is for all intents and purposes perfect.
Digital vol controls are just a way of saving money and making remote control easy... code costs nothing to manufacture!
 
A stepped attenuator is for all intents and purposes perfect.

I have to disagree. It will have contacts that corrode and cause variable (and often non-linear) resistance, and the mechanical construction exposes the signal to potential electrical noise.
 
I have to disagree. It will have contacts that corrode and cause variable (and often non-linear) resistance, and the mechanical construction exposes the signal to potential electrical noise.

It's not an issue AFAIC. I expect to have to maintain equipment correctly.
 
I disagree. Controlling volume in the analogue domain is vastly better and more accurate than doing it digitally. By accurate I do not mean just channel balance but everything.. THD, noise the lot. A stepped attenuator is for all intents and purposes perfect.
Digital vol controls are just a way of saving money and making remote control easy... code costs nothing to manufacture!

I doubt any engineered item is 'perfect' TBH. You'd need to define in engineering terms some of the words you use like 'accurate'. I also doubt what you say about THD and noise.

Digital controls are like analogue ones. They *can* be a way to save money if you go for cheap examples for the sake of savings. And it isn't the 'code' that does the work, its the physical electronics.

So in sum, I doubt all the absolutist sweeping claims you just made. :)
 
I have to disagree. It will have contacts that corrode and cause variable (and often non-linear) resistance, and the mechanical construction exposes the signal to potential electrical noise.

I also tend to find with analogue controls that the balance is never perfect across all attenuation settings. OK, the best Alps pots I've used to get to around 0.1dB or 0.2dB tracking balance over a lot of their range, but digital would do orders of mag better if sensibly designed and built into a DAC unit. They also dodge the problem raised earlier in this thread of the varying output impedance with setting of an analogue system tending to change the frequency response.

Of course, such tight specs might not matter. But that isn't a case of 'perfect' just real-world 'adequate'. :)
 
I doubt any engineered item is 'perfect' TBH. You'd need to define in engineering terms some of the words you use like 'accurate'. I also doubt what you say about THD and noise.

Digital controls are like analogue ones. They *can* be a way to save money if you go for cheap examples for the sake of savings. And it isn't the 'code' that does the work, its the physical electronics.

So in sum, I doubt all the absolutist sweeping claims you just made. :)

Vol control carried out in DSP or similar will need code to program it yes?

Worrying about dither, ENOB, jitter etc all goes away with a wholly analogue vol control. I'm not saying a digital vol control cannot be very good, just that they vary hugely in implementation and are certainly not all equal and that therefore it is usually better to leave the digital giving full output and use an analogue solution. IMHO. I'm an "analogue only" engineer who never took the digital side of things beyond designing and using hard wired logic with TTL and CMOS and don't deny an anti digital bias!:rolleyes:

Whilst of course nothing is perfect in engineering a switched attenuator with say 0.1% resistors and gold plated contacts on a good quality switch is about as perfect a vol control as you will get in terms of channel matching, distortion, bandwidth (bandwidth dependent on how it's used etc) and dynamic range. Obviously mechanical wear will eventually end its life but this could be many years down the line :)
 


advertisement


Back
Top