It seems the shorting ring is not a very effective method as many JBL drivers with Symmtrical Field Geometry (SFG) have used it since the 1980s. E.g. The 2226H has 1.75mH of inductance, 2235H has 1.2mH, 2231H has 1.4mH and so on. This is a very old method and also needs to remove the heat coming from the shorting ring.
The 18Sound paper cited by OllBoll is a fairly recent (2005) and novel method, though the drive requirements for its extra fixed coil can be complicated.
The 18Sound paper cited by OllBoll is a fairly recent (2005) and novel method, though the drive requirements for its extra fixed coil can be complicated.
Last edited:
Well - the conclusion that shorting rings are not effective might be wrong. While the drivers you mentioned do indeeed have quite high inductance there are two things to be said:
1.) It is not the inductance per se that is causing distortion but its variance with excursion. While your examples may have quite high inductance it does probably vary only very little with excursion.
2.) Shorting rings also help to reduce flux-modulation of the magnetic field, which is another source of nonlinear distortion.
18 sound's AIC method will also cause losses and heat that has to be removed. But it seems, that its effectiveness regarding distortion reduction can be increased significantly by the implementation of a clever amp configuration.
And one may not forget that with most drivers the negative effects of coil inductance can be lowered by the use of current drive or amps with high source resistance.
Regards
Charles
1.) It is not the inductance per se that is causing distortion but its variance with excursion. While your examples may have quite high inductance it does probably vary only very little with excursion.
2.) Shorting rings also help to reduce flux-modulation of the magnetic field, which is another source of nonlinear distortion.
18 sound's AIC method will also cause losses and heat that has to be removed. But it seems, that its effectiveness regarding distortion reduction can be increased significantly by the implementation of a clever amp configuration.
And one may not forget that with most drivers the negative effects of coil inductance can be lowered by the use of current drive or amps with high source resistance.
Regards
Charles
Maybe you're right. The Le of their newer driver 2216Nd used in the JBL M2 (attached) shows much weaker variations (w.r.t X) when compared to the generic variations shown in post #2. However, I'm not sure if this one has the ring in it (it probably has).
Any other methods for Le control ? Really interesting thread, lots of points.
Any other methods for Le control ? Really interesting thread, lots of points.
Last edited:
A method I like is to use a current source amplifier (high output impedance relative to speaker). With such an amp then the impedance variations does not matter.
* Doesn't work as well within ~ 2 octaves of fs. Can be amazing when crossed higher though. I have measured distortion with a pro midrange 6ND430 with a pure voltage source (Hypex NC400) and a pure current source (Firstwatt F2J Clone). The current source F2J drops third order distortion by ~ 10 dB.. Note that this effect is smaller on drivers with more fancy motors that already have shorting rings. The cool thing is that I feel is underused is that if I've understood correctly then shorting rings reduce the efficiency, so if we threw them away and used a current source amp we could have higher overall efficiency without sacrificing distortion performance.
Last edited:
However, driving with current will make any impedance peaks show up as gain, isn't it? I understand that we are not interested in the region near fs, but still.
Last edited:
How about an amp with an impedance that is rising with frequency ? Low around fs and higher further up. No problem to build something like this with frequency dependant current feedback. It would call for some EQing of course.
Regards
Charles
Regards
Charles
True, but that can be fixed with some filters. The problem is rather that the distoriton isn't reduced in that region, in some cases it can even be amplified. Unless you go into all in into fancy territory like the Purifi Audio sub where they use a current source amp with negative feedback from an accelerometer to eliminate a lot of distortion.
Well that would leave the impedance resonance with current control only in the fairly constant impedance region beyond fs. Since current drive is meant for the purpose of eliminating impedance variations (due to say excursion), does that really give any advantage, as it selectively operates where the excursion is lower (higher freq)?
I think pure current drive is OK for midranges that are often not used near fs. It may also be suitable for compression drivers in which case we clearly tend to stay away from resonance.
Somehow, I am getting the feeling that the inductance by itself is not the villain as long as it remains constant. Also, methods for managing inductance often end up solving other problems like distortion etc (shorting ring, differential drive, current drive ...)
Last edited:
It works! Thats what Bruno Putzeys does in the Grimm LS1 on the midwoofer (pure current source on the tweeter)
OK, so current control shifts the driver impedance compensation to the amplifier input, which is then applied as a filter (EQ).