@markbakk
ok, lets sum it up. crossover, enclosure and so on make it impossible to define it before buying cause it can be equal.
or from the other side, all are equal good. (in case of a finished system)
postings above where ppl are are telling me that Bl, Re, Le, η0 and so on are playing a part in this case
but, didnt tell how this values should be.
what should i believe ?
ok, lets sum it up. crossover, enclosure and so on make it impossible to define it before buying cause it can be equal.
or from the other side, all are equal good. (in case of a finished system)
postings above where ppl are are telling me that Bl, Re, Le, η0 and so on are playing a part in this case
but, didnt tell how this values should be.
what should i believe ?
I'd choose a larger mid and cross it lower to a 10" or double 8" woofer. The ScanSpeak 18M8631T is better than any other mid they make, even the illuminator. Cross it first order HP and LP to maintain the transient response. You can easily do a 1st order HP with a larger dome ie. Seas T35C002 and theoretically have a transient perfect design.
You can also do a FAST design with a peerless TC9 cone fullrange and 8" woofer of your choice, but it won't handle as much power or play as loud as a 3 way.
As above: transient response is linked to bandwidth so all other things being equal, the driver with the wider linear operating range has superior transient response. The simplistic way of looking at it is if a driver can move quickly enough to reproduce, say, 2KHz accurately, then by definition it is fast enough to reproduce all lower (i.e. longer, aka 'slower') frequencies, down to its mass-corner Fhm (2Fs/Qts') below which it's defined by the box / baffle system alignment. There's more to it than that of course, but it serves as a basic general guideline.
T/S parameters as such only have relevance up to the mass corner, arguably with the exception of Le (which is a fundamental parameter that gets lumped into the generic 'T/S' parameter handle) and even that is questionable since Le itself isn't necessarily fixed -it depends on how it's derived and at what frequency: not all drivers will track the nominal curve it, Re (and system alignment in the LF) describes. In fact, most won't, hence the reason there are more complex methods to try to mathematically generate a representative curve, which itself won't account for mechanical resonances / storage etc. In reality, the impedance curve is what you want rather than a static number, as it's far more revealing / useful.
The driver examples you link to are not really a meaningful comparison since they're completely different types -an alloy cone woofer and a hard-pulp cone midrange driver which are also in totally different size brackets -the Scan is less than half the size and they are not meant to be used for the same purpose.
Last edited:
And still all this is moot. One’s never going to use any of these drivers full range, except for maybe the TC9 that was suggested. And since no one defined good transient reproduction, let me throw in that the range from 30Hz to 100Hz might be of equal importance as the range above 1k for good transient response. So you end up with a full range system comprising several transducers or a planar that has several sections plus crossover and/or some kind of level vs frequency correction that changes the transient response.
I wouldn't call it moot as such, since the OP wasn't initially aware that transient response above a driver's mass-corner goes hand in hand with bandwidth. I'd say getting that clarified, and that in the LF alignment dominates makes it worthwhile, if nothing else. Additional filters & their potential effects are another issue, but one thing at a time. 😉
Ok true. But it’s really a matter of changing the applied current if you don’t like the transient response, isn’t it?
Using a very high output impedance amplifier, a current drive or transconductance amp, mitigates the effects of the coil inductance causing high frequency roll off, as acceleration is proportional to current. At the low end of the frequency range an asymmetric shelf filter as popularized by Linkwitz can be used to cancel the natural low frequency resonant peak and produce the desired F3 and Q at the low end. The limits with this approach comes down to the dispersion pattern due to cone diameter and cone breakup and Xmax * cone area, or total linear displacement for limiting the SPL max at the low end. With active crossover and EQ and a transconductance amplifier other parameters pretty much do not matter anymore.
You can't do anything about the effects of VC inductance regarding IMD distortion. If its too high it will always result in modulatied Le ruining the midband. A high current amp only deals with the back EMF of the motor, not the effects of modulated Le. This is what limits drivers such as the Seas A26 woofer used in the newer 2 way inspired by the Dynaco A25s..
In response to profiguy, I didn't see anyone in the thread reference using a "high current amp", so I'm assuming you are referring to my post #28? I mentioned using a high output impedance amplifier. A transconductance amplifier (An amplifier that converts a voltage to a current) with an output impedance of 1000 ohms for instance. This sort of amplifier puts out a current that is independent of the load impedance. So if the impedance of the driver rises with high frequency it puts out more voltage to compensate and keep the same output current. This is done specifically to avoid the IMD due to varying inductance with motion of the voice coil. It is well documented on this website: https://www.current-drive.info/
Last edited:
Lars Risbo posted on something alike recently here, using a passive approach. True current drive would mitigate the mentioned distortion indeed, essentially because you need a current to drive a moving coil loudspeaker. No current, no drive. Nothing new here though, I think this has been investigated since the ‘80s.
As in how quickly you can slew it through the coil? Sure. Hard for a driver to move quicker than the current change.
Hi, Le gets into acoustic domain only if it affects current. Current through voice coil is what makes force and cone move, acoustic output.
When there is high impedance in series with the voice coil, the Le gets miniscule in comparison and has only very little effect on circuit current, which also flows through the voice coil, hence very little effect on acoustic output = reduced distortion. This also means increased HF output which you might have to EQ somehow.
With voltage amplifier the voice coil impedamce dominates circuit impedance and has macimal effect on circuit current, iow Le(x) gets fully emitted acoustically.
Slew rate is hardly an issue with modern amps. But I meant also the current level vs frequency corrections you need to get a linear acoustic response. That leads in turn to an 'optimal' transient response. 'Optimal' as in the driver or system has it's limits of course.
Which all in all could lead to yet another discussion about the probable phase shifts that will lead to deformation of the transient response. Let me be clear that I follow common knowledge that moderate phase shifts aren't audible but in very exceptional listening conditions (reflection free). Everyone is entitled to his or her own belief or experience of course.
Which all in all could lead to yet another discussion about the probable phase shifts that will lead to deformation of the transient response. Let me be clear that I follow common knowledge that moderate phase shifts aren't audible but in very exceptional listening conditions (reflection free). Everyone is entitled to his or her own belief or experience of course.
I think "slew" referred more to the possible (or impossible rapid) current change in a high inductance coil.
@tmuikku That does make sense in respect to current VS voltage thing, but I've noticed it still matters how much Le varies on drivers with high inductance, even when current drive is used. It does appear to affect lower mids alot if there is more excursion even when its still within mathematical xmax. I do notice the amp does make a big difference on these mote difficult drivers with higher Le. This may explain why some drivers that have a sketchy reputation sound so different, especially on tube amplification. Perhaps the type of passive crossover used in conjunction with a high Le driver has more impact on SQ than on a lower Le driver. Any series resistance with the VC makes things worse, so its an argument in favor of using active crossovers. I've noticed on very low crossover points the inductor DC resistance really affects driver performance.
Shorting rings do wonders for most drivers if they're used in the right place. The problem is the drop in efficiency when the VC gap is enlarged to accommodate the shorting rings. There are losses at play here as well not having the inductance control in the proper location in respect to the most linear portion of magnetic flux. That's always a compromise. I've added copper rings right on top of rhe pole piece on some drivers which didn't always have predictable benefits. Often times you can cut the self inductance almost in half.
Shorting rings do wonders for most drivers if they're used in the right place. The problem is the drop in efficiency when the VC gap is enlarged to accommodate the shorting rings. There are losses at play here as well not having the inductance control in the proper location in respect to the most linear portion of magnetic flux. That's always a compromise. I've added copper rings right on top of rhe pole piece on some drivers which didn't always have predictable benefits. Often times you can cut the self inductance almost in half.
For my part, was just pointing out that your acceleration factor calculations, based purely on BL/MMS was flawed, not that 'acceleration factor' gives much meaning in the first place.🙄
And the discussion on current and voltage drive always drags up a host of incorrect electrical terms.
'Current drive' is simply a circuit that produces a constant value of current regardless of source voltage or load resistance/impedance.
hardly a very advanced circuit in it's basic principle.
Last edited:
And that is exactly why the voltage capacity of a 'current drive amplifier' becomes important. In other words, it is the applied voltage that decides the rate of rise / fall of current (since VC is largely inductive). Since this relation is non-linear (due to varying inductance), a higher voltage capacity (than calculated) is often necessary. Also, the current loop needs to have enough control bandwidth in order to allow reasonably good correction into the highest frequencies.
Now, one may argue that the rise / fall in current still happens even at a lower voltage, but the resulting slew rate maybe insufficient for achieving the targeted current THD.
so what happened if a impedance correction (flattening the impedance line- not the peak) is established ?
how does this circuit (RC) impact le over frequency and transient response ?
how does this circuit (RC) impact le over frequency and transient response ?