I think it might help to explain how the current sampling (small R value in series with speaker) causes the speaker acoustic output to have lower distortion.
I believe this is the indirect subject Joe is talking about.
THx-RNMarsh
TAA, 1985 -- Motional Feedback.
View attachment AA Low feed back amp design 3.pdf
I believe this is the indirect subject Joe is talking about.
THx-RNMarsh
TAA, 1985 -- Motional Feedback.
View attachment AA Low feed back amp design 3.pdf
Last edited:
Joe also seems to be trying to equate voltage as force vs mechanical force, ignoring that electromagnetic transduction is an intermediary mechanism in a speaker.But distorted current gets fed back right through the voice coil, just as F = BLI tells us.
Interestingly, that equation can be reversed. It is bi-directional and it can be both at the same time (current device and voltage source, since F is force or voltage). It takes mental effort.
In other words, he seems to be suggesting that amplifier or speaker back EMF voltage is proportional to BLI, something like that. But EMF depends on velocity, not force.
Last edited:
Motional feedback is about as far from what he is talking about as possible.
No contradiction at all.
But can you restate it in different words, like what do you mean by "both cases."
I will then take it one step at a time and hope to satisfy you.
As I understand it Joe is advocating impedance eq across loudspeaker input terminals causing loudspeaker to appear as 'clean/flat' resistive load. This eliminates harmonic/circulating currents which means amplifier is not having to output harmonic currents in order to satisfy constant output voltage condition. In the case of 'perfect' amplifier reactive load of no importance but in the case of typical 'consumer mid-fi' reactive load is of strong importance and this is what Joe is addressing.
Dan.
Last edited:
Actually, I understand very well, only you don't understand me. OK, so be it.
I am not sure by what you mean:
Please explain?
Huh? Once again you have completely misunderstood my position. Alas I am getting used to that.
I have not proposed current drive, I keep saying that. What you just said could have come out of my own mouth. In fact I have said the same thing.
But please consider this, it is easy to measure that the dBSPL of a driver is totally proportional to the current (and that is true with any kind of drive). I am surprised you don't know that? It can even be computer modelled and it actually confirms what happens in the real world. Additionally, the heat dissipated in the voice coil, seeing that the average driver has 0.5% efficiency, that this heat is also proportionally (within 1%) to the dBSPL of the driver. All of this is easy to prove to yourself, as I am sure you have the capability. To understand that, you have to put some interesting pieces together.
And I never said it was:
And:
ABSOLUTELY!
Maybe if you tried to understand my real position, maybe we could find common ground on which to agree.
Re the measurement you asked about earlier, in time I will provide one, probably only one, see how you respond and see if you take it seriously and if you do, then maybe we can make progress. I would like that. But if not, then I will leave it at that. OK with you?
Here Joe the I in the speaker is the same no matter what impedance Z has. It's known as Ohm's law. We understand you are not using current drive, but you are not controlling the current in speaker either.
Attachments
Last edited:
Correct, but amplifier output current is controlled, ie amplifier output current is in phase with/matches amplifier output voltage. Eliminating amplifier/cable/loudspeaker harmonic currents removes a bunch of system dependencies and this is just standard simple electrics 101, Joe is correct except for his terminology. Another way to describe Joe's approach is that circulating currents are between loudspeaker input impedance compensating network and crossover/drivers and not between amplifier and loudspeaker, same deal this is electrics 101. As per electricity distribution systems, elimination of circulating currents reduces resistive losses and effectively improves system efficiency.
Dan.
Last edited:
Exactly the diagram we need! Anyone take bets we can actually get agreement that this is how the world actually is?
Please show the lecture notes for this is its 101
But it doesn't, all the speaker current still flows in the amplifier. Sorry introducing the garbage amplifier argument is a misdirection. The claims are a class A/B amplifier can't be made that works, this is wrong.
Come on, the shunt networks increase current and degrade efficiency.
Last edited:
- which makes no benefit if the amplifier is not ill-designed. Additional load only increases distortion, though this may be unimportant. Additional load concept is a nonsense. I will probably be very strict for the reason that people like you and Joe are spreading superstitions and and technical nonsense which the beginners may take as a fact. I am not sure why the nonsense is supported, one reason is the ignorance and the second one are commercial reasons.
You said it’s easily seen in datasheets. It’s not even slightly visible in the few I’ve looked at. How can AK4499 be improving on something that you can’t even show exists? Why didn’t you comment on the plots from the AK4497 that I attached? Because it doesn’t confirm your pet theory?
I am not claiming that a Class A/B amp can't work with reactive load. I am saying that the output stage drive voltage is quite different for reactive load compared to resistive load and will contain harmonics in order to maintain constant output voltage condition.
In my experience the result is counter, ie with speakers measuring flat impedance out past 40kHz amplifiers run cooler at high output power.
Dan.
Pavel, not everybody in the world has your perfect amplifier, in fact most of the world runs typical Yamaha, Sony, Pioneer etc mid-fi amplifiers and receivers that are in my experience quite strongly cable/load dependent.
This is exactly where Joe's Elsinore speakers are aimed, ie the music enthusiast who can afford to diy build a pair of good speakers but can't afford the likes of a pair of Pass monoblocks for $80,000.
This Elsinore impedance compensation approach goes a long way to achieving good sound out of affordable and relatively mediocre electronics.
In other words $100 worth of passive components wired across the loudspeaker input terminals allows the diyer to achieve much better sound than he can afford otherwise, ie thousands of dollars more for amplification.
This is a win for the diy constructor and a win for Joe and both can be confident that his design will behave as predicted on pretty much any amplifier.
The opposite condition is the likes of Wilson loudspeakers and no doubt others with extreme impedance dips ensuring misoperation/distortion/limiting with most common/affordable amplifiers.
I say $100 worth of impedance eq is good engineering and economically solves problems that are far more costly to solve otherwise.
Dan.
Pavel, are your amplifiers reviewed in the likes of Stereophile ?....I am sure they are decent but how good are they really when compared to other amplifiers of renown.
Last edited:
- Status
- Not open for further replies.