Steve Eddy said:
Fair 'nuff. In the strict sense of current loops, there is no direct path. I guess my point was to illustrate that, despite the lack of a direct current carrying path, there is nonetheless a way for signal-related current to end up on the AC mains, and it happens because power is transferred back through the transformer, not current. Only in certain cases would this not be true (such as the balanced class-A example).
That is a valid way of looking at it, but I wouldn't characterize it as 'more accurate', by any stretch. It's different. Each perspective can be useful for describing the circuit in a way that is suited to a particular kind of analysis. It is certainly more common to think of PSU caps from the perspective of their role as an energy reservoir, as indeed this is the most basic function they perform, and the one most commonly paid attention - what's the ripple I want, the peak load current, the ESR, etc., and thus calculate what cap size is needed to fit the requirements. The 'reservoir' perspective greatly simplifies this calculation, but is not conducive to understanding what happens to signals of various frequencies in the circuit.
It is equally valid, however, and much more useful in this context (analyzing the flow of AC signals through the power supply), to think about the PSU from the perspective of it's role as a filter. A nonlinear, time-varying filter, perhaps, but a filter nonetheless. In the old days of tubes (and again in the modern era of SMPS), the filtering function of PSU caps was more obvious, due to the common presence of other filter components such as inductors and resistors. But even without those components, there are such elements present, usually as parasitics not shown on the schematic, and the electrical function is effectively the same.
One can simply consider the load currents and the charging pulses, as merely AC signals feeding into the filter. The DC components can of course be ignored for the purposes of such modeling. The caps, combined with the surrounding impedances, function as a filter, attenuating those AC signals, usually such that the resulting signal amplitude (aka 'ripple') is within some desired range.
Or you might want to be a bit more sophisticated... For example, you have the impedance seen looking back into the transformer secondary, in series with the variable impedance of the rectifier diodes, forming a variable low-pass with the capacitors. You might choose to analyze the behaviour of this by simplifying ... perhaps by examining a time slice when the diodes are in peak forward conduction. Or perhaps when the diodes are reverse biased, and they have a very high series impedance, plus their parasitic capacitance. Now you can quite easily model what is happening to, say, an undesired 100MHz signal in the circuit.
Steve,
Please...give it a rest. You're determined to spin in semantic circles and it's getting annoying. Everyone but you seems to get it, which leaves two possible interpretations:
1) You really don't get it.
2) You do get it, but you're determined to be a thorn in John's side.
The first is beyond anyone else's power to change and the second has become tiresome.
As for no-power-transformer amps--wasn't the Monster Amp supposed to be sans power transformer? Not that a functioning specimen exists that I know of, but in principle it can be done. In that case they bought vast quantities of capacitors (there are pictures here at DIY showing small mountains of caps), but I don't see any reason you couldn't use LRLRL... to do the trick. Granted, the thing would weigh a ton and cost a young fortune, but it would work. At this point, it's more economical to use caps than inductors.
You can DIY inductors to save money, in which case the cost will come down considerably, but the high price of copper at the moment will still be a burden. DIY caps aren't going to be competitive on a uF/$ basis, but I confess that I still get itchy fingers when contemplating rolls of aluminum foil and Saran Wrap side-by-side. So far I've resisted the temptation. In terms of reliability, waxed paper would probably be a better choice, although it's thicker and any capacitors you build would have less capacitance than ones built using Saran Wrap.
Actually, come to think of it, you could build a foil inductor out of aluminum foil...
Grey
Please...give it a rest. You're determined to spin in semantic circles and it's getting annoying. Everyone but you seems to get it, which leaves two possible interpretations:
1) You really don't get it.
2) You do get it, but you're determined to be a thorn in John's side.
The first is beyond anyone else's power to change and the second has become tiresome.
As for no-power-transformer amps--wasn't the Monster Amp supposed to be sans power transformer? Not that a functioning specimen exists that I know of, but in principle it can be done. In that case they bought vast quantities of capacitors (there are pictures here at DIY showing small mountains of caps), but I don't see any reason you couldn't use LRLRL... to do the trick. Granted, the thing would weigh a ton and cost a young fortune, but it would work. At this point, it's more economical to use caps than inductors.
You can DIY inductors to save money, in which case the cost will come down considerably, but the high price of copper at the moment will still be a burden. DIY caps aren't going to be competitive on a uF/$ basis, but I confess that I still get itchy fingers when contemplating rolls of aluminum foil and Saran Wrap side-by-side. So far I've resisted the temptation. In terms of reliability, waxed paper would probably be a better choice, although it's thicker and any capacitors you build would have less capacitance than ones built using Saran Wrap.
Actually, come to think of it, you could build a foil inductor out of aluminum foil...
Grey
Steve Eddy said:
>Well, the amp's output current doesn't return to the AC line's neutral.
It should be easy enough to test.
Cut the neutral wire, leaving only the 'hot' wire.
If you're right, everything will continue to work just fine.
(test on a 2-conductor amp circuit. With a 3-wire: hot,neutral,ground the neutral and ground are usually wired in parallel back to the breaker panel)
Warning. This test can create a shock hazard or even electrocution hazard.
GRollins said:
What "Monster Amp" do you mean? You don't mean Hiraga's Le Monstre do you?
Well of course it can be done. Why would there be any question that it could? Ultimately you're still using a power transformer. Just that in this case it's up on the utility pole and not in the amplifier. Using a power transformer in the amp is really only for safety (isolating from the mains) and as a convenient way to step the mains voltage up or down in order to get the voltage you want.
Well, yeah. Provided whatever you had it hitched up to was drawing a constant current as it would have a rather high impedance.
But at the end of the day I'm left wondering what's the point? You can build all sorts of crazy things. How does any of this relate to building better amplifiers?
se
I once participated in a design that did not, in principle, use batteries, caps, inductors, or transformers. Instead, we used photovoltic cells for power. It is now in a time capsule, somewhere. What this shows, is that we are not reliant on energy storage, so long as energy is continually replenished, even by the sun, directly. That is not an essential part of how an amp works. It has been implied here, that it is most important, but it is not really essential after all.
AndrewT said:
Hi,
My point is that it is in the circuit, therefore capable of effecting the sound. Especially with poor design practices.
The power supply caps are in parallel and try to filter out any garbage on the hot or neutral.
Any garbage on the line is still there, just diminished by the filter caps.
There will be something on the neutral.
myhrrhleine said:
If you cut the neutral wire, nothing will work. Not because I'm wrong, but because you don't understand how it works in the first place.
The AC line's neutral is the return path for the AC line's hot (again, half the time as it's AC so half the time hot is the send and neutral is the return and the other half of the time neutral is the send and hot is the return).
Current flows from the AC's hot into the top of the power transformer's primary winding, through the primary, and out through the bottom of the primary where it meets up with neutral, forming the current loop for the primary side of the transformer.
Cut the neutral lead, and you get no current in the transformer's primary. No current in the transformer's primary, no induced voltage on its secondary. No induced voltage on the secondary, no supply voltage. No supply voltage, amplifier no workee.
Again, current only flows in loops. And there are three distinct current loops in a typical amplifier. The loop on the mains side of the power transformer, the loop between the transformer's secondary and the power supply caps (this is the loop which charges the caps), and the loop between the supply caps and the loudspeaker (this is the loop which discharges the caps).
The current flowing in the first loop is not the same current that flows in the last loop, and hence the current flowing in the last loop does not return to the first.
se
john curl said:
I don't know whose implication you're referring to here, but it wasn't mine.
But when your circuit requires DC and your power source is AC, you do indeed need some form of energy storage in order to achieve a reasonably stable DC voltage, whether it be energy stored in capacitors or energy stored in inductors.
se
Steve Eddy said:
It's not that I don't understand, it's that I'm failing to communicate well with you.
If you are defining current flow as movement of individual, specific electrons, I agree with you.
If current flow in general, I disagree.
But I don't want to derail this fine thread, so I'll go back to lurking for now.
<lurk>
myhrrhleine said:
This seems to be a distinction without a difference.
Even if we consider "current flow in general," the amp's output current is still not "flowing" back to neutral. Current must still "flow" through some path, and there's no path between the amp's output and the AC line's neutral for it to flow through.
se
- Status
- Not open for further replies.