Linear circuit design was a lost art in 1985? Seems people keep finding it an arguing about it to this day. Just like Swift foretold in 1728!
Maybe in the eyes of the audio community, or should I say The Hermetic Order of Audio Craft.
Last edited:
🙂
Tonight I am killing a valve preamp to replace its guts with many many 8 legs and switched mode supplies and reed relays. Any more heresies I should add?
Tonight I am killing a valve preamp to replace its guts with many many 8 legs and switched mode supplies and reed relays. Any more heresies I should add?
Hi John, I kinda gathered that - the irony is that being even slightly controversial is being frowned upon and I am hardly an audio rebel and yet treated as if. Just inching past what they deem acceptable has to be 'shut down' (and I am quoting you here) or should that be 'shouted down' - if only we were allowed to be slightly controversial, it would actually make the thread better. We might also get some fresh blood when it comes to participants. Every great river needs a flush every now and then.
But free speech is only deserved by those who claim it, even if there is a price to pay!
Being controversial is fine. Spouting gobbledygook will get peanuts thrown. Has always been the way. You do keep quoting this 'shut down' as if that is your goal with every thread you start.
Why does the voltage change? You said the voltage is directly controlled by the amplifier, the speaker impedance changes, so the current through the load changes. I think we can all agree with that, but why does the voltage also change? I thought the whole idea was that the amplifier is a voltage source and the current varies iinversely as the load impedance varies (both above and below Re). Where are you deriving the voltage change? You make it sound like the current has spontaneously changed through a fixed load impedance, which changes the voltage, which you said was constant but the impedance changed. I am obviously missing something in that paragraph, and would really appreciate some clarification.
This is too silly to answer in detail - so I will only give one comprehensive answer:
The Elsinores can be driven by any amplifier of any output impedance, the alignment stays in place (other than a minor blip just under 20 Hertz) and also the crossovers are locked into place and this is what explained by just one thing - the amplifier sees a load so that the current is the same at all frequencies - this is the essense of current EQ, the current is equalised to be the same (or as close as can be) at all frequencies.
No mention of current drive, here or elsewhere, as being my preferred amplifier - and an intelligent person should be able to grasp that, OK?
Constant current (load) is not the same as constant current drive (amplifier)!
Hi Joe,
Just so I understand you correctly, you're saying that you could basically take a high current amplifier, like a Krell for example, and insert some reasonable values of resistance in series with the speaker and the response wouldn't change? By reasonable values of resistance, I mean that the sound level can be maintained without any trouble by increasing the drive voltage from the amplifier without getting into clipping.
So you are basically saying that your speaker design presents as a resistive load at all frequencies in the audio spectrum.
-Chris
Just so I understand you correctly, you're saying that you could basically take a high current amplifier, like a Krell for example, and insert some reasonable values of resistance in series with the speaker and the response wouldn't change? By reasonable values of resistance, I mean that the sound level can be maintained without any trouble by increasing the drive voltage from the amplifier without getting into clipping.
So you are basically saying that your speaker design presents as a resistive load at all frequencies in the audio spectrum.
-Chris
Only in the minds of a few, again a feel good hypothesis with no real substance. Joe is irritated that folks are not flocking to his speaker solution.
Last edited:
Go to #3778 and look at the amp mod and the resulting speaker's acoustic output THD reduction.
Tell me why this happens... reduction in THD from the loudspeaker.
First one with the correct answere wins the Grand prize. OK Waly is included.
THx-RNMarsh
Tell me why this happens... reduction in THD from the loudspeaker.
First one with the correct answere wins the Grand prize. OK Waly is included.
THx-RNMarsh
Last edited:
That's fair enough, and the answer is surprisingly simple - but it takes a little effort to realise how that is the case.
I have a Peerless Nomex driver, it has a DC resistance of 6 Ohm, the Re of the driver. At around 1500Hz the total impedance has doubled to 12 Ohm. If we were driving this from an amplifier with 6V RMS output, the load will draw 500mA. If there had been no additional inductive back-EMF, then the current would have been a full 1 Amp. We have reduced the current by half.
Now the Re will see 3V and the back-EMF part of the impedance also 3V, we have a split because the two are in series.
It should be apparent that a high source impedance equals current drive, so let us use that as a reference - and now our driver would now see 12V across its terminals. Just follow what the current does. Now the back-EMF will see 6V and the Re the other 6V, being in series, the driver terminals will see the full 12V.
But the back-EMF contributes nothing to the dB-SPL of the driver, it is the voltage across Re that does. But even more, and this applies with both current and voltage drive, since the presence of back-EMF impedance, the voltage across the Peerless driver will never be the same as that which can be calculated as arriving at Re, which actually produces the dB-SPL (you will need to put a microphone in front of the speaker with a baffle). In the presence of additional series impedance, that voltage will always be lower, lower than what the terminals sees.
In our voltage example above, 500mA versus 1 amp, the dB-SPL will change by 6dB, but the back-EMF does not contribut to the dB-SPL at Re where the current times voltage (and the voltage is formed by the current) affects the force that produces dB-SPL.
The back-EMF is basically a series impedance (being back-EMF it is in itself a voltage source that impedes (and hence an impedance) the current.
That back-EMF causes current that is less than stable and all those funny wrinkles that show up in the impedance plot, that should not be there, affects and changes the back-EMF impedance - the Re is rock steady (but of course it has thermal dependency issues, so there is no free lunch).
Perhaps the below attachment should help. Cheers, Joe.
PS: I added the last double graph to show that it works at all frequencies, even at 42 Hertz and the motional impedance (which is also motional back-EMF). Hope you find it interesting.
Attachments
Last edited:
Richard, I am pretty sure that your circuit change adds a small amount of positive current feedback that tends to partially cancel the RESISTANCE (what is measured with a DC driven ohmmeter) that limits the damping of the bass driver at resonance. This should reduce the peak excursion of the driver and therefore lowers the distortion at the fundamental resonance. I think that it is called 'negative damping factor'. This technique is also used by Audio Precision in its oscillators, and it is a very powerful tool that most here have probably never heard of.
Last edited:
I thought I was overhung, but have been informed I am underhung. No coupling, no flux. Now I'm just hungover.
I could use the valves if you have no use for them 🙂
After reading the Bob Katz article that was recently posted, I decided I need a preamp with a strong 2nd harmonic character. Let´s say 0.3% to 0.4% THD. I have to look up the formula and see what the corresponding level in dB is.
The one shown below has a gain of 36 times. My usual sources are multibit DACs with a small resistor for current to voltage conversion. With an input of 50mV peak the output is 1.8V peak, and the 2nd harmonic is -60dB down. 3rd harmonic is -85dB down.
Folded cascode is nice, doesn´t require much supply voltage and the PSRR is good. Bandwidth should be good because of the cascoding.
Using a PNP for the cascode looks much worse. I hope the 20 or so 2SJ74 I´ve got are genuine, I believe they are. I bought them many years ago here at diyaudio.
Wish me luck 🙂
-Alex
Attachments
Last edited:
- Status
- Not open for further replies.