In the past I've build several speakers with the little Tang Band 5" wideband W5-1611SAF, which has a ferrite magnet.
I want to explore a new idea and found this driver is now also available, lower cost, with a neodynium magnet, as W5-1611SA.
When I look at the data sheets, they are very similar but there's a large difference in Levc, which I assume is the voice coil inductance.
The ferrite version is 0.023mH, the neodynium is 0.11mH, a factor of 5 difference.
That seems a very large difference; the DC voice resistance, voice coil dimensions etc are the same.
BL is also similar (5.53 vs. 5.18).
Where does the large induction difference come from, or maybe it is a typo?
Jan
I want to explore a new idea and found this driver is now also available, lower cost, with a neodynium magnet, as W5-1611SA.
When I look at the data sheets, they are very similar but there's a large difference in Levc, which I assume is the voice coil inductance.
The ferrite version is 0.023mH, the neodynium is 0.11mH, a factor of 5 difference.
That seems a very large difference; the DC voice resistance, voice coil dimensions etc are the same.
BL is also similar (5.53 vs. 5.18).
Where does the large induction difference come from, or maybe it is a typo?
Jan
Hello Jan
This is an important item. My experience is,nedium is more powerful and less steering reserve,lets more hear at small level and ferrite is less powerful but more steering reserve and lets minder hear.
Grusse Wollie 😀
This is an important item. My experience is,nedium is more powerful and less steering reserve,lets more hear at small level and ferrite is less powerful but more steering reserve and lets minder hear.
Grusse Wollie 😀
Unfortunately my English is not good, I have to make do with a translation program, and I am dyslexic.
Regards Wollie
Regards Wollie
Hello Jan
0.023 mH is usually tweeter territory. Maybe it should be 0.23 mH for the ferrite version.
Regards
Charles
0.023 mH is usually tweeter territory. Maybe it should be 0.23 mH for the ferrite version.
Regards
Charles
For Le it doesn't matter whether it is ferrite or Neo. What matter is extent of saturation of magnetic circuit. With fully saturated circuit, iron permeability becomes unity and VC behaves as air core inductor. With iron away from saturation, the VC is like iron core inductor. So, this means that in your Neo driver there is less iron saturation.
In my books, low Le is a sign of speaker quality. And that's why I like field coil drivers.
In my books, low Le is a sign of speaker quality. And that's why I like field coil drivers.
The lower BL accounts for the lower moving mass, but not quite so the neodymium driver has higher output in the upper range. A lighter voicecoil means the neodymium driver can excite the cone more above it's breakup point so that also increases the treble response slightly. The slight upturn at the end of the ferrite impedance tells me that this driver may be less underhung and so might have more flux modulation. I'm just LARPing as an expert though.
Most markaudio driver have very low Le (e.g alpair 7p 0.016mH), so not impossible
Because of the underhung voicecoil these drivers experience a strong "semiductance" characteristic, so a simple value for the voicecoil inductance is going to create a lot of errors. They are more similar than different, so the values shouldn't be so different. Just look at the impedance chart and estimate the inductance.
I think it's likely the inductance value is measured at 1KHz and since there are resonances in that region, the machine spits out a silly inductance value.
I think it's likely the inductance value is measured at 1KHz and since there are resonances in that region, the machine spits out a silly inductance value.
Keantoken replied my post. My stab is this is the typical TB sloppyness: some specs are spot on, others too good to be true.
What I was really wondering is whether that factor 5 of difference in Le could be explained with the differences shown in the data sheets.
Or maybe there is a hidden difference, for instance in degree of underhung, as Keantoken hinted at.
Edit: just saw @lasercut's post. That explains it largely. The 'official' data sheet shows 0.011mH, not 0.11mH on the datasheet I downloaded from the seller. Down to a factor of 2 'only' .
Thanks!
Jan
Or maybe there is a hidden difference, for instance in degree of underhung, as Keantoken hinted at.
Edit: just saw @lasercut's post. That explains it largely. The 'official' data sheet shows 0.011mH, not 0.11mH on the datasheet I downloaded from the seller. Down to a factor of 2 'only' .
Thanks!
Jan
Last edited:
Yes. I guess this is measured in non-saturated situation. So the difference could be explained by a difference in core permeability.
But a factor of 2? Wouldn't that also impact efficiency big time?
Jan
But a factor of 2? Wouldn't that also impact efficiency big time?
Jan
The inductance value is just a single point on the impedance curve, sloppily chosen by a machine or underpaid worker, so why would you take it for granted when you have the full complex impedance in front of you?
Also, the inductance at 1khz is heavily defined by eddy currents in the pole piece, so gap geometry or silicon content of the pole piece can change it drastically.
Also, the inductance at 1khz is heavily defined by eddy currents in the pole piece, so gap geometry or silicon content of the pole piece can change it drastically.
Notice that the slope of the impedance is steeper below 1KHz than above it. This is because the skin depth of the eddy currents is becoming larger than the pole piece. The pole piece is so magnetic that if it weren't for the eddy currents shunting it's inductance, the inductance would be 10-100x larger.
So you can't determine the voicecoil inductance just by counting turns. You have to know the properties of the pole piece as well.
It is also true that without the semiductance characteristic of the pole piece, driver manufacturers might not even be able to make speakers with flat response. Since output is a function of voicecoil current, the speaker impedance has to be a close inverse of the cone's acoustic gain otherwise you do not get a flat response. So while the gap geometry can be optimized for distortion, if the resulting voicecoil impedance cannot match well to any available cones, you do not get a winning speaker.
I have often wondered which comes first, a well behaved cone or a good motor. How often is a motor designed and then a suitable cone chosen, rather than a cone designed and a suitable motor chosen for it?
So you can't determine the voicecoil inductance just by counting turns. You have to know the properties of the pole piece as well.
It is also true that without the semiductance characteristic of the pole piece, driver manufacturers might not even be able to make speakers with flat response. Since output is a function of voicecoil current, the speaker impedance has to be a close inverse of the cone's acoustic gain otherwise you do not get a flat response. So while the gap geometry can be optimized for distortion, if the resulting voicecoil impedance cannot match well to any available cones, you do not get a winning speaker.
I have often wondered which comes first, a well behaved cone or a good motor. How often is a motor designed and then a suitable cone chosen, rather than a cone designed and a suitable motor chosen for it?
This guy was not impressed at all with the performance of the ferrite version:
https://mapoulin.wixsite.com/audiobymartin/tan
https://mapoulin.wixsite.com/audiobymartin/tan