Going on, it might be possible to more fully understand why it appears that power transformer type and size can actually make a sonic difference, even if it is difficult to measure why. The return path of the amps output has to go somewhere as we do not usually plant a stake in the ground for it to go there. By the way the word 'electricity' is a key word to focus on what I was trying to show on the WIKI link: http://en.wikipedia.org/wiki/ground
john curl said:
What are you saying here, John? Are you saying that the "return path" of the amp's output goes back to the power line? Is that what you meant when you said "Unfortunately, many here think only of the forward filtering and not so much about the return path back to the power line"?
se
Also, the large power supply caps are also an alternate path for the AC part of the audio output that has already gone through the loudspeaker. This makes the caps potentially sonically important, because the audio must go them as well as the loudspeaker itself.
Of course, there is RFI getting into the 'ground' and acting as another input to the loudspeaker.
Of course, there is RFI getting into the 'ground' and acting as another input to the loudspeaker.
We have as well found from experience, that power supply diodes are also very important. We now use high speed/soft recovery diodes in all Vendetta, CTC, and Parasound products. They cost more, and are not very convenient, as they are harder to get in full wave bridge form, but they make less 'hash' and take some of the 'electronic' sound out of solid state and tube products. Actually for tubes, I like tube rectifiers best. For much the same reason as why we use high speed diodes for solid state, but rectifier tubes appear to do it even more gently.
This is about all that I can report on about power supplies at this time.
While this is not directly applicable, I received an E-mail today from an electronic engineer residing in mainland China. He purchased one of my 20+year old Vendetta designs in HK and wanted to upgrade it, if possible, even though he thought that it sounded great stock.
I gave him much the same advice as I have given here. Bigger caps, bigger transformer, and high speed diodes. So far, the caps have been updated, he tells me. I hope to hear again from him when the caps break in (his words).
This is about all that I can report on about power supplies at this time.
While this is not directly applicable, I received an E-mail today from an electronic engineer residing in mainland China. He purchased one of my 20+year old Vendetta designs in HK and wanted to upgrade it, if possible, even though he thought that it sounded great stock.
I gave him much the same advice as I have given here. Bigger caps, bigger transformer, and high speed diodes. So far, the caps have been updated, he tells me. I hope to hear again from him when the caps break in (his words).
this "break in" thing got espoused by many so easily....i wonder what is inside those caps that need breaking in? or for that matter "burning-in" , many seems to use these words rather loosely.....cassually interchanging them that theys now sound to be the same....😱
Although I did not record the references I remember reading quite reasonable accounts that noted that the following change their characteristics during the first few hours of use:
* Electrolytic capacitors
* Metal film resistors
* Vacuum tube cathodes
I seem to remember that the note about the metal film resistors was in an Analog Devices publication, so if my memory is correct then it is likely to have been quite well researched before publication.
* Electrolytic capacitors
* Metal film resistors
* Vacuum tube cathodes
I seem to remember that the note about the metal film resistors was in an Analog Devices publication, so if my memory is correct then it is likely to have been quite well researched before publication.
Gordy said:
aren't these supposed to happen at the manufacturer's? as part of their testing/ quality control/assurance programs?
aren't we entitled to get those "good-to-go" the moment we purchase those?
why ais this burn-in thing being passed on to consumers?
Regarding the resistors, this seems to be the AppNote, from ADI:
AN348 Avoiding Passive-Component Pitfalls, page 4, Resistors: Voltage, Failure and Aging.
IMHO it doesn't automatically induce that they change their sonic character, even if the resistance value will need some time and/or thermal stress to settle.
Aging (included the pre-aging at manufacture) for tubes is well documented. The same seems to be the case with el-caps, where initially an electrode forming process must take place when new or after long periods without applied voltage.
- Klaus
AN348 Avoiding Passive-Component Pitfalls, page 4, Resistors: Voltage, Failure and Aging.
IMHO it doesn't automatically induce that they change their sonic character, even if the resistance value will need some time and/or thermal stress to settle.
Aging (included the pre-aging at manufacture) for tubes is well documented. The same seems to be the case with el-caps, where initially an electrode forming process must take place when new or after long periods without applied voltage.
- Klaus
john curl said:
Not a lot of people realize this. I've found that the schematic of a Circlotron can sometimes help, as it makes the signal-through-the power-supply-cap link somewhat more explicit. The problem then is that they get fixated on Circlotrons as having this "design flaw" without being able to generalize it to all amps. The problem is simply that "normal" amp schematics show the power supply off to one side as though it's a separate and distinct thing. Not so.
One of the things I've said at intervals is that to even begin to understand power supplies and why they matter you have to imagine yourself standing at the node between the power supply and the active circuitry. It's a two way street, both at the rails and at ground. It's not just a steady stream of electrons flowing from the power supply to meet their destiny creating music in your listening room. You not only have to be able to look from that node towards the circuit --which seems to be intuitively obvious to a lot of people--but you also have to be able to look back "upstream" to the power supply. That isn't as obvious.
If you really want to mess up some peoples' minds, tell them there's no such thing as ground. Smoke comes out of their ears. Their eyes roll in circles. Their brains short circuit. Even setting aside the arbitrary nature of ground (just for fun, consider what happens during lightning...is "ground" still "ground" during a lightning strike?), once you go through a transformer, you can just as easily attach the positive or negative rail to an external ground (meaning the earth sort) as the center tap. No, I'm not suggesting that anyone do this, as that one piece of equipment will refuse to play nicely with other equipment...but taken on its own, the circuit will still function more or less normally. If you've got +-30V rails and choose to attach the +30V rail to an external "real" ground, you'll no longer have +-30V rails. You'll have ground, -30V and -60V. But the circuit doesn't care as long as you're not trying to hook it to something else. All the parts know is that they see the same relative voltages that they've come to know and love and the circuit will be quite content.
Tony said:
Electrolytics are the most easily seen to break in because their imperfections are whole orders of magnitude worse than film caps, for instance. Take an electrolytic that's been sitting on a shelf for a long time, charge it, disconnect the power supply, and watch it discharge. Now leave it hooked up to a power supply for a couple of days, then retest.
And for anyone who has never done so...charge an electrolytic cap to full rated voltage. Discharge safely but quickly through a resistor. Then lay a screwdriver across the terminals so you know the confounded thing is discharged. Dead. Zero. No charge left in it at all. Then remove the screwdriver, but leave a meter hooked to it.
What's this?
Voltage appearing where none was before?!
What?
And people think that caps can't have any effect on sound quality. Not so. The ghost of everything you've played for the last five or ten minutes is coming back to haunt you. Perhaps that's overly cinematic...let's just say that the stored charge can blur the signal coming through the cap, and as John is pointing out, everything goes through the cap, even if it's in the power supply as opposed to something like a coupling cap, where it's more obviously in the signal path.
Needless to say, film caps are far, far better for this sort of thing, but then you bump up against the cost and physical size problems.
Grey
Tony said:
The thing to grasp here is that as a part sits--and this is fairly obvious with electrolytics, although perhaps less of a problem with other parts--they un-break in. To the extent that a piece of equipment is manufactured, tested, boxed, shipped, sits in a back room at the retailer, finally gets sold, gets taken home by the customer, unboxed, and finally...at long last...gets hooked up again, you're talking weeks easily and often months. When I was in retail, we had items that sat for as much as a year and sometimes longer.
Even with best intentions, you're fighting a losing battle.
Grey
john curl said:
i am an electrical engineer by education....at the moment i work as a QC/QA inspector for an LNG plant being constructed for SHELL here in the Sahkalin islands...
I didn't read this one thoroughly the first time through:
Alternate path?
For all intents and purposes, it's the path.
Most all of the current being delivered to the loudspeakers is drawn from the plates of the supply caps which are tied to the voltage rails. In order to draw that current from the supply caps, you have to have the same amount of current flowing back into the opposing plates of the capacitors, i.e. those plates which are tied to the internal reference ground.
The only exception to this is what relatively little current may be supplied to the load by the power transformer's secondary during the relatively short refresh cycle, and most of the current being delivered by the transformer's secondary is going into topping up the power supply caps.
The amplifier's output does not return to the power line as you imply. It returns for the most part to the power supply caps and in very small part to the power transformer's secondary via the transformer's center tap.
se
john curl said:
Alternate path?
For all intents and purposes, it's the path.
Most all of the current being delivered to the loudspeakers is drawn from the plates of the supply caps which are tied to the voltage rails. In order to draw that current from the supply caps, you have to have the same amount of current flowing back into the opposing plates of the capacitors, i.e. those plates which are tied to the internal reference ground.
The only exception to this is what relatively little current may be supplied to the load by the power transformer's secondary during the relatively short refresh cycle, and most of the current being delivered by the transformer's secondary is going into topping up the power supply caps.
The amplifier's output does not return to the power line as you imply. It returns for the most part to the power supply caps and in very small part to the power transformer's secondary via the transformer's center tap.
se
- Status
- Not open for further replies.