[Folsom]

How do they respond? Does it pass? dump to ground? Burn up first in heat dissipation? I'm not sure what is inside of the linear regulators; I just see them used in DC applications. Perhaps they are just the same?

[JohnR]

Hi, it's not really clear what the question is, but typically a linear power supply (for e.g. a preamp) has two stages, rectification and filtering, and regulation. The first converts the AC waveform from the power transformer into a rectified sinewave, and the filter capacitors "smooth" this out so it's a DC (constant) voltage but with some ripple (remnants of AC). The regulation stage takes that DC voltage and uses active circuitry to generate a lower DC voltage with much lower ripple.

Does that help?

[G Georgopoulos]

How do they respond? Does it pass? dump to ground? Burn up first in heat dissipation? I'm not sure what is inside of the linear regulators; I just see them used in DC applications. Perhaps they are just the same?

A linear regulator smooths out any ac ripple from the input by keeping the output at constant voltage,for example as ac ripple goes up and down the regulator action smooths it out at the output,JohnR has already
mentioned the various stages and how they work,one last thing linear regulators dissipate more energy than
say switchmode regulators...

[Folsom]

Hm perhaps I should be curious about AC and DC at the same time.

It has occurred to me that some very good sounding circuits have regulators, but only ones that also use ground. The thing about them is that voltage regulation doesn't tend to offer the kinds of improvements that are gained when the specific type of regulation is used. Ones that are simple and don't reference to ground don't sound as good but the voltage is the same.

It just seems like some sort noise reduction is occurring. I'd like to understand why the behavior at a regulator would be conducive to this.

[avahifi]

Any typical 3-pin regulator is referenced to ground.  The three pins are input, output, and ground.  Their normal limits are input voltage (too high and the regulator overheats) and output current (too much demand and the regulator overheats).  These are easy conditions to take into consideration in proper design.

What often is not considered adequately is insuring that the circuit conditions can never cause the regulator and the active circuit after the regulator to oscillate.

An interesting test is to drive the audio circuit with high frequency square waves and move the scope probes to the output of the circuit's regulators and watch the ringing and garbage at that point.  Then you realize that the circuit needs a lot of engineering help, which by way, better sounding capacitors won't fix.

Regards,

Frank Van Alstine

[mboxler]

Any typical 3-pin regulator is referenced to ground.  The three pins are input, output, and ground.  Their normal limits are input voltage (too high and the regulator overheats) and output current (too much demand and the regulator overheats).  These are easy conditions to take into consideration in proper design.

Is there a "simple" rule of thumb as to the ratio of the input voltage to the output voltage?  For a small voltage regulator, say 5 volts, would the number of diodes used enter the equation as well?

I'm building a PS for my Squeezebox touch, using an LD1085V regulator, and find it hard to determine the optimum transformer output voltage, taking into account the 1.414 rule.

Thanks for the thread, D of S.  Hope it's okay to expand it a little.

Mike

[avahifi]

In general you want the lowest possible voltage unregulated power supply ahead of your rated regulator voltage that is certain to regulate even if your Ac line voltage droops sometimes.

For a 5V regulator, I would suggest 10V DC in should be safe, solid, and reliable.  This means you have to back out the numbers to decided on the AC output voltage of the selected power transformer to provide the needed DC voltage after the diode rectifiers.

Assuming a power transformer with a 7V secondary and adequate current capacity for the application, your rectified DC voltage after a full wave bridge would be about 10V DC (AC voltage times 1.414).  With an adequate filter capacitor after the bridge, maybe 1000uF at 16V and another 1000 at 16V after the regulator, along with 1uF and .01uF film caps on the regulator output as close to that part as possible, and another 1uF film cap at the input to the regulator, again as close to the regulator as possible, and you should end up with a quiet and useful power supply.  Keep the ground side wiring heavy.

Frank Van Alstine

Frank Van Alstine

[Folsom]

Thanks for the thoughts Frank.

Frank how much could you explain how noise works its way from the transformer primary side, into the output of an amplifier? I understand how to prevent noise in a system, but once it is in the field becomes a bit bigger than my understanding as I'm not at a level like you or Pass or someone who designs circuits. I can understand them, but the non-theoretical side of what shouldn't be there being there, is something a bit beyond me.

Essentially I'd like to look at filtering within a component, eventually. Power conditioning changes any system I've heard from good (at any price range) to joy. However I have a moderate understanding of how the noise moves about. It isn't hard to understand how many bad things noise can do to say a DAC, but how and what arrives at it is... fascinating.

[avahifi]

Several easy ways for electromagnetic fields from the power transformer or other sources can end up audible in your speakers.

The first way is if an unshielded transformer is located too close to sensitive low level audio signal paths.  This is the reason we specify shield rings on all our power transformers.

Another way is with ground paths in the circuit boards or in circuit board connections are not done very carefully. If there is more than one ground path often this will cause ground loop hum issues.  This can happen between components if there are different ground schemes in some of them or if some components have three wire AC power cords and plus and other units have two conductor power cords.

Running signal cables close to and parallel with AC power lines can induct AC hum into the signal cables.

External devices, such as dimmer switches, make an absolute mess of the AC line powering them and also may broadcast hum fields.  This often is a source of hum.  I have seen aquarium pumps and water bed heaters do this too, along with high intensity lamps.

A near by radio or TV station can broadcast right into your speaker cables.  Twisting the cables about 3-4 turns per foot by chucking them into an electrical drill often will fix this.  And of course unshielded interconnect cables are a terrible idea.  They always will be picking up hum and its there screwing up the fidelity of your system even if you can't hear much of the hum.

Of course a broken ground connection in an interconnect cable or jack will cause a really LOUD hum.

Finally if you can't find and fix the issue, consider that possibly that the component just does not know the words of the song.  :)that

Frank Van Alstine

[Folsom]

I don't believe in power conditioning to fix "wrong" things. To me it is all about the incredible improvement to sound in ways that are, well, hard to describe even. The ones I've been making do such incredible things that basic theory of how audio components work are not explaining how a power conditioner can change playback so damn much. You don't even have to A B test to know it.

The one clear thing to me is common mode simple can make diodes act more often, and causes more noise themselves. Hence why fast, quiet, diodes are so coveted. I'm unclear on what this noise looks like if you were to analyze it though (Is it very small AC?)

I've been thinking about AC filters post rectification.

[mjock3]

Frank,
I would like to thank you for your input. Coming from someone who is in the business it is very valuable. I wonder if you could tell me what a shield ring is. Where I could look at one on the internet. I have tried a seach like at Mouser but did not find it.

Thank you,
Mark

[brj]

There are a few decent pictures of one in this thread:

http://www.polkaudio.com/forums/showthread.php?131537-Plitron-1000VA-Transformer-Evaluation

[mjock3]

Thanks brj, I have never seen one of those before. 

[avahifi]

All our amplifiers and tube based preamp power transformers are custom special order units with shield rings specified and built in and with the input and output voltages and current capacity needed.  No short cuts here.

Frank Van Alstine