How are voltage disturbances dealt with in a power amplifier?

[roog]

Triggered by the discussion breaking out in the recent Nordost cable review thread which seems to have drifted into the perceived impact of 'impure mains' on audio equipment.

@amirm It might be helpful to demonstrate, by measurement, the impact of both mains disturbances but also the power amplifier stages themselves on the DC rails in a couple of amplifiers and ultimately the impact on the speaker output.

Whilst I have used Spice simulations to predict such things in a simplistic way, I don't think that this is as effective as a practical demonstration such as measuring what happens in an actual item of equipment. I expect the result to be amount to a negligible effect, but it might be good to demonstrate how the various stages of an amplifier, PSU, voltage regulators, negative feed back etc handle the various sources of undesirable artifacts on the internal DC supply rails serving the amplifier.

I appreciate that this is straying from the core business of measuring audio equipment performance, but much like your other educational videos it might help to clear a few things up.

 

[voodooless]

Power supply rejection ratio - Wikipedia

en.wikipedia.org

 

[solderdude]

Triggered by the discussion breaking out in the recent Nordost cable review thread which seems to have drifted into the perceived impact of 'impure mains' on audio equipment.

@amirm It might be helpful to demonstrate, by measurement, the impact of both mains disturbances but also the power amplifier stages themselves on the DC rails in a couple of amplifiers and ultimately the impact on the speaker output.

Whilst I have used Spice simulations to predict such things in a simplistic way, I don't think that this is as effective as a practical demonstration such as measuring what happens in an actual item of equipment. I expect the result to be amount to a negligible effect, but it might be good to demonstrate how the various stages of an amplifier, PSU, voltage regulators, negative feed back etc handle the various sources of undesirable artifacts on the internal DC supply rails serving the amplifier.

I appreciate that this is straying from the core business of measuring audio equipment performance, but much like your other educational videos it might help to clear a few things up.

This is highly device dependent so there is no 'single' demonstration possible.

 

[sergeauckland]

This is highly device dependent so there is no 'single' demonstration possible.

In general terms, the power supply reservoir capacitors are what smooth out any noise and other fluctuations in the incoming mains. The mains transformer removes any DC on the mains, (although the transformer may hum more if there is any DC to remove) whilst the reservoir capacitors store the charge obtained in every half-cycle of the mains and deliver smooth DC albeit with some 100Hz or 120Hz ripple on it.

The device's Power Supply Rejection Ratio then removes the effect of the 100/120Hz ripple, so the output from the device is the audio only, plus any circuit-generated noise.

It's possible that with very polluted mains there may be a need for external filtering, but in most cases it's unnecessary, and this is why fancy mains cables, connectors, mains regenerators and the like do nothing for audio quality.

S.

 

[solderdude]

For small variations this is certainly the case.
Some devices with linear power supplies and regulators in it may well rely on the minimum required voltage drop. When that is exceeded the regulator action is gone and that may lead to hum becoming audible.

Cables will do nohing here but the OP mentioned disturbances on the power lines. When these become substantial (voltage drop or over voltage) this could become an issue.
It would be highly device dependent on the device in question. Some are very wide range, others less so.
Some devices can handle short power interruptions others may not.

So basically what the point is where devices start to operate outside of their intended mains input voltage range and what the audible or operational effects can be as opposed to operating within normal operating conditions.

 

[roog]

Oh I agree that it would be device dependent, and whilst I understand your explanations, there are a large number of people who do not believe this. If they did they wouldn't be considering or buying all these 'mains treatment' products.

I suppose I had in mind a practical method of demonstrating for instance that noise does not get past particular measures which are incorporated in a well designed 'blameless?' amplifier to any meaningful extent.

One example I recall was the use of the DC power rails in a basic amplifier to serve the input stages without the use of additional active regulation and what the effect of the perturbations caused by the output devices swinging the load had on the high sensitivity input stages. How much does this matter? and what does it do to the output?

I totally agree cables have no influence and it wasn't my intention that this thread should discuss this.

 

[fpitas]

There's a nice tutorial around here called The Truth About Power Supplies, or something like that.

Aha! https://www.audiosciencereview.com/...th-about-hifi-amplifier-power-supplies.36904/

 

[fpitas]

One example I recall was the use of the DC power rails in a basic amplifier to serve the input stages without the use of additional active regulation and what the effect of the perturbations caused by the output devices swinging the load had on the high sensitivity input stages. How much does this matter? and what does it do to the output?

Well..unless you're feeling every penny, that's just dumb or ignorant. Even a resistor and zener diode, costing a whole 10 cents, will clean things up. If they goofed that up, I would expect the rest of the amp to be a disaster.

 

[MaxwellsEq]

I agree with @sergeauckland and @solderdude If there are gross errors on the mains supply such as brownouts (e.g. caused by lightning or HV switching faults). There is no predictable behaviour amingst different amplifier designs. Linear PSUs with large reservoir capacitors may "paper over" minor brownouts better than smaller reservoirs. If the rails are well regulated (with a significant V drop across the regulator), there may be less sensitivity to gross power supply issues.

If the mains is nominally normal, then the amplifier should be designed with sufficient PSRR, to make it measurably insensitive to noise in the mains.

 

[roog]

Well..unless you're feeling every penny, that's just dumb or ignorant. Even a resistor and zener diode, costing a whole 10 cents, will clean things up. If they goofed that up, I would expect the rest of the amp to be a disaster.

I agree, but I have seen this. Given the trivial cost of regulation I would not expect this to be an issue in a competent design.

I think the point I am trying make, perhaps not very well is that, such things aren't an issue because of the way equipment is designed, this is spoken about quite a bit on this forum, I thought it might be helpful to show how these measures work in action.

I am not trying to create argument, just to promote understanding.

 

[solderdude]

I think the point I am trying make, perhaps not very well is that, such things aren't an issue because of the way equipment is designed and that we can show this in action.

That too is highly design dependent so you could only demonstrate such for a specific design and the result could be very different on another design.

Key point here is 'design' and that is not even limited to component choice but also PCB routing, wire routing etc. in said device.
Testing amp-A or DAC-A will not say much about other devices.

Another point is that internal DC rail variations will differ from AC mains variations and may not have a direct relation.

 

[Speedskater]

Typically the problem children are from the spikes of refrigerators, air conditioners or pumps starting or stopping.

 

[MaxwellsEq]

I think the point I am trying make, perhaps not very well is that, such things aren't an issue because of the way equipment is designed, this is spoken about quite a bit on this forum, I thought it might be helpful to show how these measures work in action.

There are some tests that may help clarify your point. Every so often, Amir reviews mains conditioners, power cables etc. For power cables he uses the AP in loopback mode then inserts the cable into that loop - and demonstrates there are no differences. For power conditioners, he treats real mains power as a very loud 60Hz signal and looks at how distorted it is (very) he puts it through the conditioner and discovers minimal benefit in the audio range and sometimes things are worse.

But here's the useful bit - he often also measures a real world piece of kit both with and without the mains cord, conditioner etc. He uses sensitive line level kit, because that's more likely to demonstrate perturbation. What he's shown is that competent kit measures the same with or without different mains cords and conditioners. Recently he created a very distorted mains signal and fed it to a piece of kit, whose measurements were totally unaffected, even though the mains signal was worse than any of us should experience. The lesson is that competent design filters mains problems...