Since a compression driver is so sensitive, and generally needs to be padded down relative to a woofer, there's almost always power to burn in a series resistor. It's true that's not exactly current drive, but is pretty close if the resistance is much larger than the rated impedance.
I would agree with you completely, although some people just don't like to use passive comps in series with drivers. Don't understand it myself.
I am no electric engineer, so if these questions are boneheaded, I apologize in advance.
As I understand it, we are getting a 'taste' of the advantages of current driver with a series resistor. But I have a bunch of questions:
1) First, why not use an L-Pad? An L-Pad would not affect the frequency response of the loudspeaker as much as a series resistor does.
If I understand this thread correctly, it seems that the reason you're experiencing an improvement is that the impedance curve is flatter. Series resistance will flatten the curve, but so will an L-Pad. So why not use the latter?
2) My second question is related to shorting rings. As I understand it, shorting rings reduce distortion by making the inductance of the loudspeaker more consistent. A shorting ring can change the BL curve so that it looks "U shaped" instead of looking like a bell curve.
Would series resistance provide some of the same benefits?
I know that resistance and inductance are two different things. But a series resistor should reduce the variance right?
3) Does anyone think that some of the sonic benefits of low order xovers may be related to the fact that loudspeakers with first order xovers generally have flatter impedance curves than speakers with high order xovers?
As I understand it, we are getting a 'taste' of the advantages of current driver with a series resistor. But I have a bunch of questions:
1) First, why not use an L-Pad? An L-Pad would not affect the frequency response of the loudspeaker as much as a series resistor does.
If I understand this thread correctly, it seems that the reason you're experiencing an improvement is that the impedance curve is flatter. Series resistance will flatten the curve, but so will an L-Pad. So why not use the latter?
2) My second question is related to shorting rings. As I understand it, shorting rings reduce distortion by making the inductance of the loudspeaker more consistent. A shorting ring can change the BL curve so that it looks "U shaped" instead of looking like a bell curve.
Would series resistance provide some of the same benefits?
I know that resistance and inductance are two different things. But a series resistor should reduce the variance right?
3) Does anyone think that some of the sonic benefits of low order xovers may be related to the fact that loudspeakers with first order xovers generally have flatter impedance curves than speakers with high order xovers?
I tried current drive a while ago, on a gainclone configured for it. AFAIK, Qms will dominate Qts and the enclosure needs to be design with this in mind, not under voltage drive paradigm. I did not have any speaker I deemed suitable, but the small fullrange drivers I tried it on performed terribly. Xover design is a different game as well. You want more treble? Add a series coil. Yes. 
IG
IG
Sorry if that's offtopic, but I'm rather curious as a long-time lurker and a noob in this hobby. If I have a woofer with low Fs, say 50 Hz, and use it with current-drive amplifier, there will be a peak in FR in 50 Hz. But what if a really steep filter is applied at 100 Hz, say an 8th order IIR or an even steeper FIR? Will the woofer be reasonably linear above and below 100 Hz?
Most of the references shown deal with frequency response changes from high impedance drive. Certainly a lot of full range units look better, mid and top, with constant current, but you need to deal with the LF problem.
But response improvement or the self equalizing aspect are fairly trivial benefits. My only interest would be if distortion were significantly lowered, and the results for that are a mixed bag.
I don't think any of you have tried to design a multiway network driven by a constant current amp. Have a go at that first and then we can talk about the overall pros and cons of constant current.
David S
But response improvement or the self equalizing aspect are fairly trivial benefits. My only interest would be if distortion were significantly lowered, and the results for that are a mixed bag.
I don't think any of you have tried to design a multiway network driven by a constant current amp. Have a go at that first and then we can talk about the overall pros and cons of constant current.
David S
There doesn't seem to be a point in doing so (even with the NP reference), because some/most of the potential advantage of current drive is lost once you parallel something to a driver, notably a treble kill cap on the woofer. Advanced drive impedance techniques are best suited to active systems (or fullrangers)
My 2ct.
For the impedance the driver sees, there is no systematic difference. Lpad equivalents to a lower input level but the impedance is just the paralleled value of the sections. For the impedance the XO is loaded with it makes a difference, with series you can make it only higher, with Lpad you can make it lower and flatter (at the cost of some wasted energy).
There are two kinds of shorting rings, ones that wrap around the pole core below (or above) the gap, and others which are in the gap, copper tubes actually. Only the latter act as a shorted secondary turn of an air-core transformer, helping to stabilize and to lower the voice coil self-inductance. Both types stabilize flux modulation (prevent change of force factor). Series R will reduce effect of any present inductance modulation as you say.
Yes, I believe there is something to it, but I think it has to do with part of the frequency range where the drive impedance approaches high levels, that is the high end of the woofer range (series coil) and the low end of tweeter range (series cap).
Most woofers do not profit from pure current drive around the system resonance frequency, especially high Qms ones with non-conductive formers and low-friction supensions. Those run into "jump-resonance" phenomena (a chaotic behaviour) once you have significant excursion. A reasonable amount of damping is needed (conductive formers work best as their effect is not suffering from a Xmax limit, unlike electrical damping).
Those are interesting observations. In any event, since I have designed the woofer section with voltage drive in mind, it would require some serious redesign (probably much lower Q woofers) to reap any benefits from current drive.
I think for a woofer, one would want low Qms, as spec'd Qes/Qts can be disgregarded. I would not want a multi-hump impedance profile, so most venting is out of the question. A single broad and shallow bump might be OK, such as what might be given in a large aperiodic box.
Just for kicks, using a Fostex FF225WK, Qms=2.31. A 70L sealed box with an aperiodic leak to equate Ql=2 is ~26ohm at 60Hz versus ~6 ohm in the midrange. Voltage will be up ~13dB at 60Hz versus midrange. That will still result in gain above the driver reference sensitivity and Ql=1 (which is extreme) would be better yet, with only a couple dB's above reference to fight-off baffle-step.
IG
Just for kicks, using a Fostex FF225WK, Qms=2.31. A 70L sealed box with an aperiodic leak to equate Ql=2 is ~26ohm at 60Hz versus ~6 ohm in the midrange. Voltage will be up ~13dB at 60Hz versus midrange. That will still result in gain above the driver reference sensitivity and Ql=1 (which is extreme) would be better yet, with only a couple dB's above reference to fight-off baffle-step.
IG
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.