[Steve Eddy]

Quote:

Originally Posted by Currawong /img/forum/go_quote.gif You've been commissioned to build a balanced B22 for iPodPJ, so you'll have enough boards to test.

Yeah, but I'm not all that keen on testing it. iPodPJ's amp will have dual power supplies so that rules out using ground channels anyway. I was simply curious to see the measurements if there were some that had already been made.

k

 

[IPodPJ]

Quote:

Originally Posted by Koyaan I. Sqatsi /img/forum/go_quote.gif Yeah, but I'm not all that keen on testing it. iPodPJ's amp will have dual power supplies so that rules out using ground channels anyway. I was simply curious to see the measurements if there were some that had already been made. k

But why not, Steve? Your case would much easier be made to those of us who don't have the technical knowledge if you can show there is no measurable difference. I certainly don't mind if you test it.

Since I wanted a 5-board originally (which I don't think we will do anyway now because it's not necessary) I'd like to see the measurements and I'm sure everyone else would too.

Not using perception or placebo as a reason (of course they should never be ruled out), there are a lot of things in this hobby that do make an audible difference where they likely shouldn't -- especially relating to power.

 

[linuxworks]

one test case is 3 boards on a single PSU. then remove the 3rd board (from power) and test the 2ch with input-ground as 'the ground'. don't even need a 2nd PSU to do this test.

 

[sochee]

guys, I think there is a reason his avatar looks like a troll..maybe he's hinting something to us..

 

[Steve Eddy]

Quote:

Originally Posted by IPodPJ /img/forum/go_quote.gif But why not, Steve? Your case would much easier be made to those of us who don't have the technical knowledge if you can show there is no measurable difference. I certainly don't mind if you test it.

Which "case" are you referring to?

I've already said that an active ground could have some benefit with regard to interchannel crosstalk for an unbalanced amp with a single power supply due to it maintaining constant current draw on the supply.

As for the claim that the load current doesn't go to ground when using an active ground, there's no need to do any measurements. The only way for the load current to not go to ground, it would have to either spill out onto the floor, or evaporate into the air.

And that just ain't happening.

k

 

[Steve Eddy]

Quote:

Originally Posted by linuxworks /img/forum/go_quote.gif one test case is 3 boards on a single PSU. then remove the 3rd board (from power) and test the 2ch with input-ground as 'the ground'. don't even need a 2nd PSU to do this test.

It wouldn't work with a second PSU anyway. In order for the active ground to keep supply current constant, it has to powered by the same PSU that's powering the amp channel.

k

 

[Steve Eddy]

Quote:

Originally Posted by sochee /img/forum/go_quote.gif guys, I think there is a reason his avatar looks like a troll..maybe he's hinting something to us..

And maybe my avatar isn't an avatar at all. Maybe it's really a mirror.

k

 

[sochee]

Quote:

Originally Posted by Koyaan I. Sqatsi /img/forum/go_quote.gif And maybe my avatar isn't an avatar at all. Maybe it's really a mirror. k

look who's getting all defensive now

calm down guy, i was just having a bit of fun.

 

[Steve Eddy]

Quote:

Originally Posted by sochee /img/forum/go_quote.gif look who's getting all defensive now calm down guy, i was just having a bit of fun.

Fair 'nuff.

k

 

[tomb]

This discussion started out about an inverting ground channel, but turned into an involved discussion about ground channels in general. Perhaps it sheds some light on the conversation - quotes from CMoy, Tangent, Amb, Jan Meier, PRR, Snoopy and Morsel (she claims credit for the active ground idea) - she indicates at one point that early testing of the M3 indicated a big performance change going from 2-channel to 3, but not as much going from 3 to 4. There's also quite a discussion and modeling on a balanced ground:
HeadWize: DIY Workshop > Does it matter if active ground channel inverts?

 

[Steve Eddy]

Thanks, TomB.

This is the sticking point for me:

Quote:

Originally Posted by morsel Cmoy, I came up with the idea for using a headphone amplifier ground channel to shift responsibility for the high current reactive load of the headphones from signal ground to the supply rails, thus removing the primary source of signal ground contamination in 3 wire headphone systems.

This notion that with a ground channel the load current doesn't return to ground, but instead goes into the supply rails has been repeated elsewhere.

There's no doubt that the load current does indeed flow through the supply rails.

However that begs the rather obvious question, where does it go from there if it doesn't go to ground?

I asked this question at least several times and finally amb answered by saying it goes into the power transformer.

But that can't be the case because for the majority of the time, the power transformer isn't in the circuit for that load current to flow into it. And the load current doesn't just accumulate someplace and sit waiting for the next refresh cycle to come along so it can flow into the power transformer.

So where does it go?

Contrary to the popular notion, it does indeed go to ground, or perhaps more precisely, through the ground node.

It must.

The current began as the excess electrons on one plate of one or the other reservoir capacitors. And the same number of electrons must be deposited on that capacitor's opposite plate. And since each of the reservoir capacitors has one of its plates tied to ground, then every last smidge of current drawn from those capacitors must ultimately flow through the ground node. The same node from which the amplifier takes its reference from.

k

 

[digger945]

I really want to try and wrap my head around what your saying here so bear with me please. I really think we could all learn something here as we move from a place where one tries to prove one's self correct or another wrong to a place of thinking about what's actually going on here.

Your saying that the current from the rails in a bipolar power supply actually passes directly through the caps tieing them to ground. I really think this needs to be established here if I(we) are to gain an understanding of how the current is traveling full circle here.

 

[amb]

Quote:

Originally Posted by digger945 /img/forum/go_quote.gif Your saying that the current from the rails in a bipolar power supply actually passes directly through the caps tieing them to ground.

Quiescent DC current cannot "pass through" the PSU's bulk caps (and by definition cannot cause ground pollution). What we're interested in is the AC current as a result of the amplifier(s) driving signal into the load. That does pass through the bulk caps and has the potential of causing ground pollution. And indeed it must to some degree in a passive ground system, because ground impedance is never zero in real life, and that AC current is simply the sum of the current from the two stereo channels being dumped through ground.

I have shown (in the other linked thread, in particular this post), that in a 3-channel active ground system operating in class A, due to rail current cancellation (the current flow through the ground channel is the same as the sum of the two stereo channels, but in anti-phase due to the source/sink relationship), the net AC rail current sums to zero.

Note that my example in that post does not use a virtual ground. It has a standard bipolar power supply just like the β22.

Zero current change = Zero ground pollution.

This is also true in a 4-channel balanced configuration, but instead of having the ground channel taking in the sum of the currents from the two stereo channels, the "hot" amp and "cold" amp cancel each other in each balanced channel. You reap the same benefit by using a single PSU for all four channels, or one PSU per hot/cold pair of amps.

It really is as simple as that.

This will be my last post in this thread, because I don't need to have the last word on the subject. Heck, active ground isn't even my invention. I defend it because I've done the analysis, and enough testing/measurements, to conclude that 3-channel active ground topology is a superior way to drive a 3-wire TRS headphone system.

I hope that this post is enough to address your question, and is the very heart of the matter. If this doesn't convince you, then nothing will.

 

[Lifthanger]

my money is on nothing will

.
but this thread provided some nice insights into the design. also made me think about some more which is nice.

 

[Steve Eddy]

Quote:

Originally Posted by digger945 /img/forum/go_quote.gif I really want to try and wrap my head around what your saying here so bear with me please.

No problem. As I said previously, I'm nothing if not patient.

Quote:

I really think we could all learn something here as we move from a place where one tries to prove one's self correct or another wrong to a place of thinking about what's actually going on here.

Exactly!

It's NOT a personal issue. It's not about who's right or who's wrong. It's about what's going on and what's not going on.

Quote:

Your saying that the current from the rails in a bipolar power supply actually passes directly through the caps tieing them to ground. I really think this needs to be established here if I(we) are to gain an understanding of how the current is traveling full circle here.

It depends on what you mean by passing "directly through" the caps.

I didn't mean, or intend to imply in anything I've said that current was passing through the dielectric of the capacitors. It's not. So the current can never travel completely full circle. It will always come up short by the thickness of the dielectric.

But let me elaborate a bit here on what I've been trying to say.

When a capacitor is fully discharged, and there's no voltage across it, each plate has a more or less equal amount of positive and negative charges on it which cancel each other out and the charge on each plate is "neutral."

When we charge a capacitor, we do so by pulling some electrons (which have a negative charge) off one plate, and putting them on the other.

Now one plate has a positive charge (the one which the electrons were removed from), and the other plate has a negative charge (the one which the electrons were moved to). And there is now a potential difference, or voltage if you will, across the capacitor. And in the context of what we're talking about here, that voltage will be one or the other rail voltages.

Of course it required some energy to displace those electrons. But that energy didn't go to waste. It's now stored in the charged capacitor.

In order for that energy to be used, such as the energy stored in the reservoir capacitors of a power supply, it must discharge to one degree or another.

As it discharges, as in the case of powering an amplifier circuit, electrons are drawn from the negatively charged plate. But those electrons can't be drawn off the negatively charged plate without an equal number of electrons flowing into the positively charged plate. And that means current flow.

Now let me relate this to what I've been trying to say here.

Here's the original illustration, however I've flipped the direction of the red arrows to reflect electron current flow rather than conventional current flow.

Consider the reservoir cap on the positive side of the power supply in A, the passive ground example.

Load current flows from the capacitor's negative plate, through the ground node, and then to the driver's negative terminal, and then on around the loop, where there is an equal amount of load current flowing into the capacitor's positive plate.

Now let's consider B, the active ground example.

Again, let's start with the reservoir cap on the positive side of the power supply.

Load current flows from the capacitor's negative plate, through the ground node, just as it does in the passive ground example. The only difference here is that instead of flowing next into the driver's negative terminal, it flows into the positive plate of the reservoir cap on the negative side of the power supply.

Of course that current can't flow into that cap's positive plate without an equal amount of current flowing out of it's negative plate. And this current flows through the negative rail, the output stage of the ground amp, finally to the load itself, and ultimately on back to the positive plate of the positive rail's reservoir cap.

And this illustrates that a ground channel does not divert load current from ground as has been claimed. Instead, load current flows through ground just as it does in the passive ground scenario. The only difference is that the load current takes a more circuitous route in the active ground scenario.

Does this help?

k