You said you wanted 1kohm+, so as a consequence when the voice coil resistance goes up during power peaks, the voltage will tend to spiral up.
Let's just say that -- in the usual spirit of amplifier designers blaming everything on terribly imperfect speakers -- it would be the speaker's own fault for getting itself blown up despite the amplifier's perfect control over the current.
this will work. but class-A is not a requirement
don't get it. i see time invariance only as long term effect which is not relevant in our discussion. for short term the impulse response will be the same now or minute later especially with current driveActually for a speaker, it is if its "Output/Input function does not depend on the absolute time." You can use Prof. Angelo Farina's sweep to check Output/Input quickly and easily in many conditions including at different times while music is being played ... or even during the sweep itself
because voice coil temperature variation influence is eliminated"Time Invariance
The characteristics of a loudspeaker can vary over time due to changes in the environment. For instance, a paper cone might increase in mass as it takes on moisture when the humidity is high. Unusually high or low temperatures can affect both the compliance and damping of suspensions, the voice coil resistance, and the strength of permanent magnets.
Characteristics can also vary over time when a loudspeaker is exposed to damaging signals. The stiffness of compliant elements may change due to incidents of high excursion. Cones and diaphragms can become weakened from repeated exposure to high mechanical force or air pressure.
"
Memory effects could be very quick. These could be magnetic, or also mechanical, where rubber or plastics yield a little when they bend, so they have to be pulled in reverse to get back to their starting position.
But you did say the amp latched up
abstract, are you a current amp aficianado? I'm after a small 'true' current drive amp. My practical definition of this is > 1k Ro over 20 - 20k Hz. Mills & Hawksford is one such beast but too complicated for my small brain.
... and no one seems interested in how to design a speaker to suit current drive
Last edited:
I've tried this, and as you say, it's not quite current drive at 20kHz
Certainly current drive makes a speaker more TI as it reduces compression due to VC heating. it's one of the benefits of current drive. But the AUDIBLE effects I'm referring too are evident within a 1min timescale or even less. But why are we arguing?. i see time invariance only as long term effect which is not relevant in our discussion. for short term the impulse response will be the same now or minute later especially with current drive
In da 21st century, Angelo's method allows us to TEST if the IR varies even during a 14s B&K 2307 sweep
And yes. I've tested this.Thanks for the welcome George. The weather in Cooktown makes it difficult for dis beach bum to go fishing and diving for the next few days. But I've had a few HD crashes so some of my pontificating might be from the wrong orifice as I've lost a lot of info
No, the conjugate circuits have been designed to defeat current drive so that the response remains the same regardless of the source impedance.
cos current drive helps make speakers more TI. It is one of the actually audible benefits. Yes it also reduces THD etc but some of these 'improvements' aren't audible ... in a DBLT
Someone claimed speakers weren't Linear but were TI
Only people who don't even have one good speaker need stereo (to loosely paraphrase Floyd Toole)
It is for pretty much universal for all drivers. Adjust RC on output if you want higher Zo. For the tweeter in exampleI've tried this, and as you say, it's not quite current drive at 20kHz
This is Angelo's method
https://www.aes.org/e-lib/browse.cfm?elib=10211
I came up with the same method in the 90s when I was trying to measure response & THD in the theoretically shortest possible time but I'm quite happy for Prof. Farina to take the credit as he popularised it. In da 90s, the computing power, and particularly good A/Ds, were $$$ and I wanted a lot of them for Factory Test.
I disappeared into the bush circa 1997/8. When I emerged, I found the cheapest PC had more than enough computing power to do this and some even had good DACs/ADCs built in
(Sadly no longer the case ) Angelo was surprised when I showed him my Jurassic code that did exactly what he was doing It's used in the latest AP, Clio and probably all good 21st century acoustic test gear worthy of the name. You can use sweeps of any length to measure the IR (and hence the Freq Resp) even with loadsa noise. So you can do a 0.3sec sweep while music is playing, and see the Freq Resp within that 0.3sec.
Probably the simplest hook on the TI of speakers is to measure the IR with a 0.3sec sweep and compare with a 14sec sweep. That will clearly show the advantage of current drive on many speakers ... but not the non TI behaviour that current drive doesn't address.
I'll assume you are a DSP guru and know to get the Freq Resp the same for current or voltage drive, deal with room reflections bla bla
Answers to " what level/time should I match Freq Resp?" on a postcard I'm after a 'universal' current drive amp. Looks like you buy it from Ye Olde Unobtainium Shoppe
You don't have to adjust the usual voltage amp to suit your speaker
Anything with > 5x DUT impedance at 20kHz is good enough, at least that's what I've found in my experiments over the years.
Which means ~200Ohms @ 20kHz covers all sane choices of drivers.
A main trick to increase bandwidth is to increase the sense resistor, at the cost of increased losses. One can make the sense resistor a frequency-dependent impedance, though...
Nice, step-up transformer (or autoformer) seems to be a good way to keep sense impedance losses as low as possible but one has to deal with the band-pass nature of it (install DC and HF feedback paths to keep the amp happy).
I have a couple of long-suffering wide-band drivers that could do with higher output impedance at high frequencies. So keep an eye out in case I post something (no guarantees), but it'll most likely be mixed mode.But you did say the amp latched up
abstract, are you a current amp aficianado? I'm after a small 'true' current drive amp. My practical definition of this is > 1k Ro over 20 - 20k Hz. Mills & Hawksford is one such beast but too complicated for my small brain.
... and no one seems interested in how to design a speaker to suit current drive
So far I've found that the bass impedance peaks are probably best dealt with using voltage control, so as to avoid big distortion peaks and the need for active notch filters, which also add their own distortion.
What would be new and interesting, would be a higher order slope, rather than the usual 6dB per octave gradual change from voltage to current control.