Jump to content
Sign in to follow this  
Ricci

Focusworks Audio GUJ Driver Discussion

Recommended Posts

Parameters / pictures and testing of the 18" version in a large sealed cab have just been posted.

Please discuss the drivers and/or the measurements for these new drivers here.  :)

Share this post


Link to post
Share on other sites

Those look like some nice drivers, but I find myself more interested in the driver used in the Funk Audio FW21.0, assuming it has more motor strength and maybe better Le control.  Any idea if that driver will be made available to the public?  Either way, these are some nice subs as reflected in their price.

Share this post


Link to post
Share on other sites

The TSAD21v2 driver is a beast. Probably one of the better all around drivers on the market. That's IMHO of course. I don't think Nathan plans to sell those to the public but I'll let him comment on that if he wants. I'll tell you this they are not cheap by any means. That is a LOT of neo in those motors. 

Share this post


Link to post
Share on other sites

The TSAD21v2 has a BL^2/Re of 422 n^2/w, that is a scary amount of force. Nathan told me it can apply as much as 365 lbs of force to the cone. That is a hell of a motor, and give it 4800 watts of power, its like the Bugatti Veyron of subwoofers. 

Share this post


Link to post
Share on other sites

Those look like some nice drivers, but I find myself more interested in the driver used in the Funk Audio FW21.0, assuming it has more motor strength and maybe better Le control.  Any idea if that driver will be made available to the public?  Either way, these are some nice subs as reflected in their price.

 

No plans to sell the TSAD21v2 on its own. However we do have some other possibilities we are working on to sell to the public. Imagine; GUJV2 :ph34r: and all that that implies....

Share this post


Link to post
Share on other sites

Holy crap!  I wonder how the cost would compare to the RF T3S2-19?

 

Probably in the same ballpark of what the RF's real price is  ;). Perhaps a bit cheaper. I'd put the TSAD21v2 in my top 5 lusty subwoofers for sure. I'd be looking at Nathan's stuff or working with him on something custom if those RF's weren't around. I just don't dig ferrite motors anymore. When you have to reinforce the frame and baffle or support the motor because it's 100lb's and it still isn't getting the amount of efficiency you can get from a well designed neo motor that's less than half the weight it's probably time to look at a different approach.

 

About the GUJ platform. It has a lot of things working for it. It's got a lot of clean stroke. More than I had thought it would actually. They also address one of the #1 pet peeves I have with underhung drivers and that is they just don't handle power well. These do thanks to a big old 4.5" coil. I've got the 21" version on hand and I'll be running tests on it soon. Based on how the 18" did the 21 should be a first class air mover. 

  • Like 1

Share this post


Link to post
Share on other sites

These are some good points.  Assuming you are building a multiple-sealed sub system and aiming to maximize efficiency below resonance while keeping total box volume as low as possible, I believe you want to maximize BL^2/Re/Vd^2.  The assumption here is that you have a fixed amount of total internal volume to work with, regardless of the number of drivers you choose to use, and that you want to keep the power requirement as low as possible.  Of course, the number of drivers you actually choose to use will be determined by the required displacement, but note that with total internal volume fixed, adding more drivers doesn't help efficiency or help it much.  As the internal volume divided by the number of drivers becomes significantly less than Vas, adding drivers to the same volume doesn't affect overall efficiency much at all.

 

The implication here is that bigger drivers with more excursion need more BL^2/Re to maintain the same low-end efficiency, or else they'll need more amp power to push them to their limits.  This typically means they need bigger, stronger magnets to keep up with the smaller drivers.  So big drivers with high efficiency in the low-end end up needing to use Nd or they end up being very heavy.  There's really no way around it.  The trend I often see with drivers is for BL^2/Re/Vd^2 to decrease as the drivers get larger in a particular product line.  Often a vendor will use the same motor for models of several different sizes.  Even though naive intuition would suggest that the larger drivers will have more low frequency efficiency, the reality is often opposite.  If we're talking about the weight of each driver and the kind of construction that is needed to support that driver, multiple small drivers may be advantageous.

 

I'm not saying small drivers always beat large drivers, of course.  Real-life optimization is a lot more complicated than this.  Small drivers will almost always cost more for the same amount of Vd because you have to buy more of them.  For some people including you, Josh, I suspect total system weight also comes into play.  For a HT user, total system weight may be a minor consideration because the system doesn't get moved around once it's setup, but if you have to transport the system regularly to shows and whatnot, you definitely want to minimize total system weight.  For this application, there's no replacement with high Vd drivers with big Nd magnets.

 

I, on the other hand, am actually leaning toward 15" ferrite-based vs. 18" or larger as a good sweet spot for efficiency, unit weight, and cost.

 

Edited for clarity

 

Edit2: And then I saw the 21" IPAL.

Edited by SME

Share this post


Link to post
Share on other sites

The implication here is that bigger drivers with more excursion need more BL^2/Re to maintain the same low-end efficiency, or else they'll need more amp power to push them to their limits.  This typically means they need bigger, stronger magnets to keep up with the smaller drivers.  So big drivers with high efficiency in the low-end end up needing to use Nd or they end up being very heavy.  There's really no way around it.  The trend I often see with drivers is for BL^2/Re/Vd^2 to decrease as the drivers get larger in a particular product line.  Often a vendor will use the same motor for models of several different sizes.  Even though naive intuition would suggest that the larger drivers will have more low frequency efficiency, the reality is often opposite.  If we're talking about the weight of each driver and the kind of construction that is needed to support that driver, multiple small drivers may be advantageous.

 

I'm not saying small drivers always beat large drivers, of course.  Real-life optimization is a lot more complicated than this.  Small drivers will almost always cost more for the same amount of Vd because you have to buy more of them.  For some people including you, Josh, I suspect total system weight also comes into play.  For a HT user, total system weight may be a minor consideration because the system doesn't get moved around once it's setup, but if you have to transport the system regularly to shows and whatnot, you definitely want to minimize total system weight.  For this application, there's no replacement with high Vd drivers with big Nd magnets.

 

Good observations...Yes weight matters to me. The amount of drivers and cabs that I move/pickup/mount/unmount/repackage/ship etc is absurd. I'm not even counting the band equipment here. If I can save 20 or 30lbs per cab I'm all about it.

 

Yes most mfg's use the same motor on a 15 or 18 that was used on a 12 or 10. By default the smaller drivers are more efficient once you buy enough to equal the cone area of the larger driver. Not to mention you have half the thermal stress per driver. The pitfalls are it is more expensive, more complicated to wire and build and usually much heavier. the other issue is that 12" and smaller drivers typically have limitations on the suspensions just due to their size. Try designing a 4" voice coil 10" driver while fitting the spider for truly high excursion. 15's are the sweet spot where you have enough size to fit the bigger components and clearances that are needed to keep pushing the envelope.

Share this post


Link to post
Share on other sites

I just posted a ton of measurements and notes for the GUJ18v1 driver in the new vented / large sealed cab that I'll be using. Check em out.  :)

  • Like 3

Share this post


Link to post
Share on other sites

Just posted the data for the GUJ18v1 driver in the normal 4.2ft sealed cab so that can be compared to the larger sealed test from earlier.

 

Also posted the measurements from the GUJ21v1 in the new large DBV21 cab with all of the vents plugged. Should be about 320L sealed. 

 

Haven't gotten the notes finished yet but I should have them done in the next 24hrs. 

 

Enjoy... ;)

  • Like 1

Share this post


Link to post
Share on other sites

Any DIY newbies lurking should definitely use the measurements of the larger sealed GUJ18v1 system and compare them to the new measurements in the smaller enclosure.

 

It's one thing to talk about WinISD or Unibox simulations but right here with real objective data you can compare the same driver in two alignments to see the real world effects of enclosure size/alignment.

 

Really good demonstrative data.

Share this post


Link to post
Share on other sites

This is very interesting.  Comparing the ULF performance between the 18" in each of the two boxes and 21" yields some interesting patterns.  Among the 18s, the larger box gets the higher CEA score, but the maximum output is only slightly higher.  Box size has a big impact on distortion but not so much on ultimate max output, provided you have the amp power and thermal headroom to achieve it.

 

Comparing the 21" in 320L with the 18" in 280L is especially interesting.  The 18" wins or ties the CEA tests at 12.5 Hz and below, but the 21" wins on max output.  The 21" has more maximum displacement, but it suffers more from distortion than the 18" does, even though it uses the same motor and an even bigger box.  Is it because of efficiency?  We can estimate efficiency at 12.5 Hz using sensitivity and impedance data:  For the 18": 89.5 dB @ 28.3V and 9.27 ohm and for the 21": 89.0 dB @ 28.3V and 8.09 ohm.  Yes, the 18" is about 1 dB more efficient there with 70 dBSPL/W @ 2 m vs. 69 dBSPL/W @ 2 m but that doesn't explain the lower distortion.  The 21" taps out on the CEA test with fewer watts (400W vs 600W) and amps (7 A vs. 8 A) than the 18", too.

 

I think it really comes down to response shape.  Distortion harmonics generated in the motor or suspension are amplified depending on the response difference between the harmonic and fundamental.  The 21" has a higher Fb and higher Qtc, and the gap in response between 12.5 Hz and 37.5 Hz (third harmonic) is greater, so the 21" system is amplifying that harmonic from the driver so much more than the 18" system is that it fails CEA with less displacement and even less excursion.

Share this post


Link to post
Share on other sites

I do not believe it is response shape so much that is to account for this, although it could have an affect as well but what I have found is that lowish distortion passing numbers in the very low end are related to the lowish BL, at least for the box size and displacement at work, vs Xmax vs box size. In my experience given the same cone size in the same airspace lower BL will allow the driver to pass CEA with output further past its Xmax around resonance, in many cases it passes CEA even when it hits Xmech, but much closer to Xmax away from resonance, and in some cases like the GUJ21V1 tested in that box it can only get to coil overhang amount of travel and no more before getting too much distortion. Higher BL gives you more past Xmax away from resonance, but less around resonance. Many databass tests show this trend, if you look at any of the super high BL drivers like the IPAL, they maintain output corresponding to nearly identical cone travel regardless of frequency, wheras most other drivers have more around their resonant frequency.

  • Like 3

Share this post


Link to post
Share on other sites

Agreed with the above. This trend is clearly seen in almost all of the tests.

 

A lot of things are linked together but what should be looked at is the driver Qts, the system Qtc, driver Fs, system resonance or impedance curve, normalized motor force, etc...

 

The simple single parameters that seem to capture this most completely are the "real" system Qtc (not some model on fictional small signal params, but the real deal) and the "real" driver Qts which drives the Qtc in large part. Higher efficiency results in less distortion. Current induced distortion is a real problem in some drivers. 

 

In a nutshell what appears to happen with extremely strong drivers like the LMS, RF 19, Ipal, TSAD21v2, etc...Is that they are effectively wider bandwidth devices. ;) 

 

Comparing the 18 GUJ vs the 21. The 21 has about 3dB higher maximum output as expected. However the motor is driving a larger cone area in a cab that is effectively almost the same size as the 18" at 290L vs 320L so the system "Q" for the 21 is higher and distortion below the resonance goes up a bit comparatively. 

Share this post


Link to post
Share on other sites

This is very interesting.  Comparing the ULF performance between the 18" in each of the two boxes and 21" yields some interesting patterns.  Among the 18s, the larger box gets the higher CEA score, but the maximum output is only slightly higher.  Box size has a big impact on distortion but not so much on ultimate max output, provided you have the amp power and thermal headroom to achieve it.

 

Comparing the 21" in 320L with the 18" in 280L is especially interesting.  The 18" wins or ties the CEA tests at 12.5 Hz and below, but the 21" wins on max output.  The 21" has more maximum displacement, but it suffers more from distortion than the 18" does, even though it uses the same motor and an even bigger box.  Is it because of efficiency?  We can estimate efficiency at 12.5 Hz using sensitivity and impedance data:  For the 18": 89.5 dB @ 28.3V and 9.27 ohm and for the 21": 89.0 dB @ 28.3V and 8.09 ohm.  Yes, the 18" is about 1 dB more efficient there with 70 dBSPL/W @ 2 m vs. 69 dBSPL/W @ 2 m but that doesn't explain the lower distortion.  The 21" taps out on the CEA test with fewer watts (400W vs 600W) and amps (7 A vs. 8 A) than the 18", too.

 

I think it really comes down to response shape.  Distortion harmonics generated in the motor or suspension are amplified depending on the response difference between the harmonic and fundamental.  The 21" has a higher Fb and higher Qtc, and the gap in response between 12.5 Hz and 37.5 Hz (third harmonic) is greater, so the 21" system is amplifying that harmonic from the driver so much more than the 18" system is that it fails CEA with less displacement and even less excursion.

That's a very interesting observation. I think it would help to take a look at klippel data in this case to see what might be going on. There could be something with the surround -- thats probably the biggest difference between the two drivers. I don't think the cone would make much of a difference esp. down in the lower freq.

Share this post


Link to post
Share on other sites

If you think BL is a factor, you could rule out all the sofparts and just make two 18's, one with less magnets and see what happens to distortion -- kind  of an expensive experiment, but could be cool :) 

Share this post


Link to post
Share on other sites

If you think BL is a factor, you could rule out all the sofparts and just make two 18's, one with less magnets and see what happens to distortion -- kind  of an expensive experiment, but could be cool :)

Basically done this. But with the 21.

Share this post


Link to post
Share on other sites

I do not believe it is response shape so much that is to account for this, although it could have an affect as well but what I have found is that lowish distortion passing numbers in the very low end are related to the lowish BL, at least for the box size and displacement at work, vs Xmax vs box size. In my experience given the same cone size in the same airspace lower BL will allow the driver to pass CEA with output further past its Xmax around resonance, in many cases it passes CEA even when it hits Xmech, but much closer to Xmax away from resonance, and in some cases like the GUJ21V1 tested in that box it can only get to coil overhang amount of travel and no more before getting too much distortion. Higher BL gives you more past Xmax away from resonance, but less around resonance. Many databass tests show this trend, if you look at any of the super high BL drivers like the IPAL, they maintain output corresponding to nearly identical cone travel regardless of frequency, wheras most other drivers have more around their resonant frequency.

 

I generally agree with this, but I also want to point out that, all else the same, the sensitivity (response) *at resonance* is inversely proportional to BL.  So in your low BL example, response actually is higher at resonance too.

 

The simple single parameters that seem to capture this most completely are the "real" system Qtc (not some model on fictional small signal params, but the real deal) and the "real" driver Qts which drives the Qtc in large part. Higher efficiency results in less distortion. Current induced distortion is a real problem in some drivers. 

 

In a nutshell what appears to happen with extremely strong drivers like the LMS, RF 19, Ipal, TSAD21v2, etc...Is that they are effectively wider bandwidth devices. ;)

 

Comparing the 18 GUJ vs the 21. The 21 has about 3dB higher maximum output as expected. However the motor is driving a larger cone area in a cab that is effectively almost the same size as the 18" at 290L vs 320L so the system "Q" for the 21 is higher and distortion below the resonance goes up a bit comparatively. 

 

While I don't doubt that current induced distortion can be a problem sometimes, I don't believe it's the primary factor.  I believe it is correct that high efficiency correlates with lower distortion, but I don't believe this is a cause and effect relationship.  It just so happens that more efficient systems with lower Qtc have a much more broad transition region where the response is rolling off at less than 12 dB/octave.  With less roll-off, the harmonics won't be amplified as much vs. the fundamental in the low Qtc system.

 

I would argue that Fb is also important, but the situation is similar to efficiency in that it depends on what you change to lower Fb.  If you increase box size or suspension compliance to lower Fb, you get a lot more output in the low-end relative to the rest of the spectrum, so distortion improves a lot.  (Qtc goes down quite a bit too.)  If you add mass instead, the increase in low end output is modest (and vanishes once you go low enough); whereas, the added mass primarily reduces the top end response.  Qtc goes up while Fb goes down, and the effect on ULF distortion is to improve it only slightly.

 

That's a very interesting observation. I think it would help to take a look at klippel data in this case to see what might be going on. There could be something with the surround -- thats probably the biggest difference between the two drivers. I don't think the cone would make much of a difference esp. down in the lower freq.

 

Klippel data could be useful to rule out the possibility that the surround for the 21" has less linear excursion than for the 18".  However, assuming they have equal linear excursion and use the same motor, I argue that you will still see this difference.  It actually makes a lot more sense if you think about where the distortion originates and how it propagates through the rest of the system.

 

As an example, suppose 3rd harmonic distortion enters the system all the way back in a processor, early in the signal chain.  Maybe the processor adds 1% 3rd harmonic at 15 Hz.  That's the same as -40 dB.  But now you play this distorted sound through a sealed sub with an Fb at 60 Hz and Qtc of 0.707, and the difference in response between 15 Hz and 45 Hz is about -19 dB.  So now that third harmonic is only -21 dB from the fundamental, which is closer to 9%.  What happened here is that the sub played the distortion harmonics louder than the fundamental.

 

When you recognize that a roughly similar thing happens to the distortion introduced in a motor or in the compliance, it makes perfect sense that distortion relates to response like this.  The relationship is not perfect.  What you really care about is the variation of the transfer function between the thing that is distorting and the output.  The sensitivity/response is the transfer function between drive voltage and output.  To understand motor distortion, you want the transfer function between the motor force and output.  The motor force depends on current rather than drive voltage, so you also have to take impedance into account.  I think the curve you want to look at for motor distortion is actually (sensitivity * impedance).  I'm oversimplifying details here in that sensitivity is usually quoted as a logarithmic level and impedance is quoted as absolute, but I can fill in the math if anyone is interested.

 

Edit:  If Kyle is interested, I could work with him on plots of this (sensitivity * impedance) curve.  Another curve that I think would be very interesting to see is efficiency, which is a different curve that is calculated from sensitivity and impedance data.  I can also help with the calculations here if necessary.

Edited by SME

Share this post


Link to post
Share on other sites

All the data points to this being related to the efficiency of the system, current induced BL distortion and inductive distortion IMHO. The native response of the system should have some bearing on the harmonics as well but I'm not sure if it is the largest or even one of the largest factors. Also I don't think voltage sensitivity is the right tool to look at for that comparison between systems. You would want to look at the true 1w power response as this represents the behavior with equivalent energy applied at all frequencies. If that were the case how does it explain that the very efficient drivers with rising top end have less THD over the upper bandwidth for equivalent output as well? 

 

There is no doubt at all if you simply bolt up a much weaker motor with identical geometry on the same coil and soft parts, overall harmonic distortion will go up. 

Share this post


Link to post
Share on other sites

I think you may be right about current-induced distortion (in BL or Le) for THD at high frequencies but not so much for low frequencies.  High frequencies generally involve a much higher ratio of power to stroke, so distortion related to power or current is likely to be much more important at high frequencies than low frequencies.

 

If we focus on low frequencies, then inductance shouldn't matter much if at all because it's too low, relative to Re.  I brought up the 18" vs. 21" comparison specifically to argue against current-induced BL distortion as being a big factor in the low end.  To reiterate, the 21" failed CEA testing at 12.5 Hz with 400W and 7A vs. 600W and 8A for the 18".  That is definitely not a consequence of current-induced BL distortion.

 

I do agree that voltage sensitivity is not really the right tool, but 1W power response is not exactly right either.  For motor distortion, whether it is caused by excessive stroke or excessive current, the effect of this distortion will be on the motor force and the harmonics in the motor force will be amplified relative to the fundamental according to the transfer function of motor force vs. output.  That transfer function is not the same as either the voltage sensitivity or the 1W power efficiency but it has characteristics in common with both.  For example, it rolls off at 12 dB/octave once you go low enough.  I am certain that this accounts for the rapid rise in THD with decreasing frequency below resonance, which is observed in measurements of almost every sealed system that I'm aware of.  It occurs even if you keeps excursion constant through your testing.

 

Your example of an efficient driver with rising high frequency response having lower high frequency distortion deserves more discussion.  Inductance definitely interest the picture, both as a source of distortion and a modulator of that distortion.  How much does a rising high frequency response in one driver vs. another have to do with high efficiency vs. low Le/Re ratio?  If you looked hard enough, I'm pretty sure you could find a driver with high upper frequency efficiency that still distorts badly up there because of Le problems.  Here's an idea for testing the hypothesis that distortion depends on efficiency at high frequencies:  After measuring a sealed system, add some mass to the driver(s) and re-measure.  The added mass will reduce the upper end efficiency, but will leave Le and BL completely unchanged.

 

Lastly, I *do* doubt that a weaker motor with identical geometry, coil, and soft parts will have higher distortion everywhere.  FunkAudio suggests that the opposite is true at resonance, and I agree.  Again, I believe this can be explained by the relative difference in the motor force vs. output transfer function at the fundamental and its harmonics.

 

Edit: I'll concede that there are probably cases where current-induced distortion matters at low frequencies too, but I think this will only happen when the boxes get real small.  In almost all the sealed Data-bass sub measurements, they fail CEA at power levels well below what they are able to take at higher frequencies.

Edited by SME

Share this post


Link to post
Share on other sites

I don't see the logic behind saying that the 21 was taking more current than the 18 and producing less excursion yet producing more distortion down low in a similar box size, but the extra current has nothing to do with it. The same motor and coil design is on both. This points to the difference not being primarily excursion related. BL and LE are close within production tolerances. Both have their resonance at close to the same frequency.

 

Distortion is high below system resonance on all sealed systems tested. Excursion is high and so is current.Distortion is low at resonance despite high excursion levels. At resonance efficiency is high and current is low. Above resonance excursion is relatively low and efficiency is higher but distortion can still be high. Placement of the resonance within the passband has huge effects on distortion performance.

 

I have no doubt that if you add a bunch of mass to a driver the overall distortion will increase. Using a much weaker motor would increase it as well.

 

The distortion is a combination of all factors of which there are many. Low frequencies are worst case. Current is high, efficiency is dreadful low, excursion is high and the natural response is weighted towards higher frequencies.

Share this post


Link to post
Share on other sites

I don't see the logic behind saying that the 21 was taking more current than the 18 and producing less excursion yet producing more distortion down low in a similar box size, but the extra current has nothing to do with it. The same motor and coil design is on both. This points to the difference not being primarily excursion related. BL and LE are close within production tolerances. Both have their resonance at close to the same frequency.

 

No.  The 21" was taking *less* current and power than the 18" when they failed CEA.  So the difference is related to neither excursion nor current/power.

 

What it comes down to instead is the relationship between motor force and output for fundamental vs. harmonic.  The difference in output with respect to motor force between the fundamental and 3rd harmonic at 12.5 Hz is greater in the 21".  This is largely a property of the alignment with either lower Fs or lower Qtc reducing this difference.  Call it the difference in motor force sensitivity but note that motor force F = BLI, with I being current, so it's essentially the same thing as the current sensitivity.  Note that the motor force or current sensitivity is very high at resonance, especially in high Qtc designs, because there's a lot of output for very little current flowing there.

 

From a theoretical standpoint, it's also clear to me that this phenomenon applies equally to distortion caused by stroke and distortion caused by high current and/or power.  So shrinking the sealed box size harms low frequency performance in two ways at the same time, first by efficiency loss potentially leading to more motor distortion from high power and/or current and second by amplifying the harmonics relative to the fundamental even more than in the larger box.

 

With all this said, I don't believe motor force (or current) sensitivity is exactly the right metric either.  A few important things happen as you get closer to resonance.  First, the phase relationship between motor force and driver displacement shifts.  In the low frequency limit, motor force and driver displacement are in-phase, so unfortunately, the required motor force peaks at the extremes of the stroke.  OTOH, at resonance, the driver motion lags the motor force by 90 degrees.  This means motor force vanishes at the extremes, so the drop in force due to the coil leaving the gap won't be felt nearly as much by the system.  Second, back EMF has a substantial effect on the system at resonance also.  Like motor force, back EMF also vanishes at the extremes of stroke where cone velocity goes to zero, so this effect may not be especially strong but it could still alter the apparent Qtc of the system causing it to appear to be a bit higher.  Still, I think motor force (or current) sensitivity can give a lot of insight into distortion behavior.

 

I found a paper by Klippel that outlines some of the things I speak of here:  http://www.klippel.de/uploads/media/Assessment_of_Voice_coil_peak_displacement_XMAX_01.pdf:

 

The harmonic distortion is maximal at excitation frequencies below resonance. This is not only caused by the low excitation force F=Bl(x)i due to the coincidence of current maximum and Bl(x) minimum but more by the lowpass characteristic in the radiation of the frequency components below fs.  At the resonance frequency there is a pronounced minimum and the total harmonic distortion measurement has a blind spot for detecting Bl(x)-nonlinearity.
 
I find this very interesting because it means that there is a lot more to getting low distortion in a system than just choosing a low distortion driver.  The acoustic response of the system is very important as well, even if indirectly.  A big advantage had by many ported and horn system is the relatively flat acoustic response in their bandwidth compared to sealed.  In addition to the big gain in output at frequencies at the bottom of the pass band, any distortion generated in the motor will tend to be amplified a lot less than in a sealed system driven below its resonance to similar levels of excursion.

Share this post


Link to post
Share on other sites

No.  The 21" was taking *less* current and power than the 18" when they failed CEA.  So the difference is related to neither excursion nor current/power.

 

 

I found a paper by Klippel that outlines some of the things I speak of here:  http://www.klippel.de/uploads/media/Assessment_of_Voice_coil_peak_displacement_XMAX_01.pdf:

 

The harmonic distortion is maximal at excitation frequencies below resonance. This is not only caused by the low excitation force F=Bl(x)i due to the coincidence of current maximum and Bl(x) minimum but more by the lowpass characteristic in the radiation of the frequency components below fs.  At the resonance frequency there is a pronounced minimum and the total harmonic distortion measurement has a blind spot for detecting Bl(x)-nonlinearity.

 

 

Actually the 21 was taking a bit more current but only a small amount. The 18 a bit more voltage. Overall though small and likely insignificant differences. Agreed it appears to be neither excursion or current/power input. That leaves the efficiency response as the most likely culprit.

 

Link to the paper doesn't seem to work? I've probably read it a few times. Most of their literature is available for download and is recommended reading.

 

Agreed on most of the rest. In a nutshell at resonance the system is at maximum efficiency and it just so happens that many sealed systems with modern drivers end up with a resonance right around 30Hz in a moderate sized enclosure. That puts the 3rd harmonic of 10-12.5Hz at the resonance which is greatly boosted relative to the fundamental output so far below resonance. I've done a bit of looking and there is a large amount of variation in the roll off below 30Hz of the systems I've measured largely related to the driver efficiency and resulting Qtc but also likely impacted by inductance effects by some amount as well. Some of the outlier 18's like the LMS5400 are only 17 to 18dB down at 10Hz from 30hz indicating a more damped alignment. On the other end drivers like the PSI 13AV2 recone and the Obsidian 18 are more like 20 to 21dB down. We see the same thing with the GUJ21 vs GUJ18 in the big sealed cabs. The 21 has 3dB greater difference between 10-30Hz than the 18 as expected by the larger cone in about the same airspace. Interestingly the drivers at the very extremes are the T3 19 and the XXX. The XXX split coil shows a nearly 23dB slope from 30Hz down to 10Hz and the RF T3 only 14.5dB. I knew they would be on either end of the spectrum but not the book ends. I suspected that this mattered as far as distortion goes but not by the amount that appears to be indicated after looking at it a bit harder.

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
Sign in to follow this  

×
×
  • Create New...