Jump to content

(8) Sealed Incriminator Audio Judge 21" build


lukeamdman

Recommended Posts

Too many Daves here but I'm going to assume me for this question.  Honestly...marketing.  I'm using a big driver in a relatively small enclosure although the BHS24 driver works better in the small enclosure design than the HS24.  I knew with that design that I'd have to boost the low-end to get stronger low end response from the smaller enclosure.  But frankly, much of the boost is there to make the middle of the field measurement look better.  Did I push it some?  Maybe, likely.  The alternative is that the graphs from the Mariana 24 not look as impressive as compared to other company's DSP'd subs even though in-room and high output would likely be much stronger with the Mariana 24.  That's a whole other can of worms that I didn't want to deal with as most consumers researching subs really don't understand how much the room reinforces and shapes low-end output, they compare graphs and they compare numbers.  Then again, the in-room response with the Mariana 24s has been flat down to 7Hz before it starts rolling off.  This has been in several rooms.  The question is, what would that response have been like without the low end boosted?  Good question and I'd have to measure multiple rooms before I could answer that one reliably.

 

I may have to take a separate tact by showing in-room response graphs, which are more representative of real-world, instead of middle of the field measurements.  Again, this is more of a marketing issue and I am sort of am mad at myself for getting caught up in it.  But I've seen really good products marketed poorly and really poor products marketed well.  Just like engineering, everything is a balance and I want to stay balanced here too.  But I definitely don't want to ever, ever be caught up in anything deceitful or misleading...I'd rather fail at business.

 

Forgive the follow up but I'd like some clarification.

 

You said above your product has been flat in room to 7hz in several rooms.  That's after a LT and other DSP as well, correct?

Link to comment
Share on other sites

  • Replies 458
  • Created
  • Last Reply

No limiter was engaged which is what lead to the failure in the first place.

There's nothing to engage because it's built into the amp, but it is slow to act (6 seconds according to the literature I posted for the SP1-4k).  It's probably there to prevent circuit breakers from flipping and the amp from overheating as quickly as it would if it kept on supplying 4 kW.

Link to comment
Share on other sites

Golly. Please show me where the word "distortion" appears in my post? It doesn't. You're using a +10dB signal boost to get the high Fs driver to a flat response at the seats. With the HS24, you would need NO signal boost to get the same result. That has nothing to do with distortion. It's just a fact and not a trivial one. I apologize for mentioning it, but you were hinting that you selected the IA driver over the SI driver because it's somehow more reliable. With 10X power for same response, it is not.

 

This is a load of crap.  We're talking about output at 10 Hz and lower, *way* below the resonance of any sealed subwoofer system I'm aware of.  At that point, Fs hardly matters at all.  It's all down to motor strength, power input, and system suspension stiffness.  Motor strength is practically limited by cost, magnetic flux density, and and other considerations like inductance.  In small boxes, the air spring stiffness dominates the overall system stiffness.  This is Hoffmann's Iron Law territory.  In that situation, the HS-24 is actually a poor performer because it has low motor strength relative to its displacement, whereas the HST-15s and HST-18s hold their own much better here.

 

Either way, the difference in performance between these subs is likely miniscule at 5-10 Hz.  But whatever.  You won't believe me because of your decades of experience, yada yada yada.  Too bad this fact makes your whole claim even more embarrassing.

Link to comment
Share on other sites

Just like engineering, everything is a balance and I want to stay balanced here too.  But I definitely don't want to ever, ever be caught up in anything deceitful or misleading...I'd rather fail at business.

 

I think the best of both worlds would be to use the more aggressive LT so you can get the prettier graphs and better performance but have a long-term limiter that activates after some time to throttle back the power and protect the driver.  Presumably, the SP amps already have this built-in for their own protection.  (See my earlier posts.)  You might enquire with SP about whether it's possible to adjust the long-term output limit of the amps to a level that the sub can sustain indefinitely.  Such limiting shouldn't actually activate during normal use.  It's only there to protect against abuse, accidental or otherwise.

Link to comment
Share on other sites

Shred brings up a very important point.

 

The 3 AVRs I have tested cannot handle a WCS with trims set at 0 and MV at -0dB.  For the Denon/Marantz Units I have measured and others have, you need to attenuate the SubOut to -10dB (or -10.5dB to be safe) to output unclipped waveforms of a near WCS (all channels in phase with same signal at -0.1dB) with bass management engaged at MV -0dB.

 

Everyone that listens with SubOut at 0 or + levels is asking for a clipped signal.  SubOut at '0' running at MV -0dB =REF means a bass managed signal at -10dB form a WCS will clip.  That means EoT, Lone Survivor, HTTYD all will clip when run at MV -0dB = REF.  So many folks here run with subs at + 'something-or-other' at or above MV -0dB = REF, and do not know that they are corrupting the signal sent to the subs.  

 

If you want a clean sub signal and want to run hot, turn your SubOut level down and use the amplifier attenuators to set the amount of dB 'hot', or restructure everything so that MV -10dB = REF with trims at 0 or below.

 

For the LF-heavy content we like on this site, do not let the weakest link in the chain destroy the signal.

 

For these helicopter scenes that have a strong fundamental and tons of harmonics, distortion/clipping is likely masked to a very large extent, so the system may 'sound good', until you playback the material without the AVR clipping the hell out of the SubOut.  I wager most people that listen above reference with the subs hot (all controlled by their AVR) may have never heard a clean presentation on the louder portions of films/trailers.

 

For really powerful low end, the limiting factor can be the AVR if you have a capable system, unless you set up the gain structure properly taking into account the AVR's limitations.

 

I would like to see the AVR manufacturers take bass management signal integrity and headroom into account, because they obviously don't at this point.

 

 

JSS

 

 

I will also note that my subwoofer trims at the g2g for my specific system is always set at -10 or thereabouts. I do believe it was bumped up some for the m24, but I don't recollect how much. 

Link to comment
Share on other sites

Forgive the follow up but I'd like some clarification.

 

You said above your product has been flat in room to 7hz in several rooms.  That's after a LT and other DSP as well, correct?

 

Correct, with LT and DSP.  Haven't disabled them in enough rooms to know how they'd measure.  Will work on getting those numbers though.

Link to comment
Share on other sites

I think the best of both worlds would be to use the more aggressive LT so you can get the prettier graphs and better performance but have a long-term limiter that activates after some time to throttle back the power and protect the driver.  Presumably, the SP amps already have this built-in for their own protection.  (See my earlier posts.)  You might enquire with SP about whether it's possible to adjust the long-term output limit of the amps to a level that the sub can sustain indefinitely.  Such limiting shouldn't actually activate during normal use.  It's only there to protect against abuse, accidental or otherwise.

 

You are correct the SpeakerPower has built-in protections as well as the option to place a limiter for protecting the driver.  This is fully controllable by monitoring current draw over a time frame.  And the limiter can be selectively and progressively enabled such that, as an example, under 15Hz can be attenuated some initially while leaving above 15 Hz unchanged, and if more aggressive limiting is required, under 15Hz can be attenuated more followed by some attenuation at 15-25Hz.  I am working with Brian Oppegaard at SpeakerPower to develop the limiters for each of the drivers HST18, HS24, BHS24.  We're going for a protective limiter that will limit to protect the driver but will otherwise stay out of the way and not to make itself known unless absolutely necessary, just like you state. :)

 

So yes, that is exactly what I'm working on and one of the reasons I haven't formally released the Mariana 24S Max with the BHS-24 driver as I want to have everything buttoned up before I do. 

Link to comment
Share on other sites

sorry to interject but.... what is it with (some of) the threads on this site?!  :mellow:

 

that gtg/smoking coils thing really should be the subwoofer equivalent of godwin's law

 

Yep.  No more soup for you.  No more GTGs for me. LOL!  Naw, they're too much fun except when you burn up your subs.  Aw, it was good fun and better I learned the lesson there than later.  A lot of knowledge and data was also available at that GTG so there was a lot to be gleamed from the incident. :)  And what else would we argue about?  People were getting bored of being pummeled by Beast's subwoofer system after the first hour. LOL.

Link to comment
Share on other sites

Yep.  No more soup for you.  No more GTGs for me. LOL!  Naw, they're too much fun except when you burn up your subs.  Aw, it was good fun and better I learned the lesson there than later.  A lot of knowledge and data was also available at that GTG so there was a lot to be gleamed from the incident. :)  And what else would we argue about?  People were getting bored of being pummeled by Beast's subwoofer system after the first hour. LOL.

lol, you have to refer to it as "the incident" (quotes are mandatory) from now on :)

Link to comment
Share on other sites

I'm setting the record straight here.  I don't care who you are and if you disagree with this you have every right to say something but you'll be 100% wrong. 

 

My comments about my assessment of the durability of the Judge vs. the HS24 had absolutely NOTHING to do with that GTG.  Period.  End of story. 

Link to comment
Share on other sites

"I do not use SI drivers in the Raptor."

 

Well the picture on the front page of your website, that says "Raptor" on it, is a Raptor cabinet with the HST-15s loaded in it.  That's a little confusing. 

 

"The HS24 is less expensive to buy and less expensive to power within the parameters of the stated goal of ULF reproduction in a small listening space. I do not believe that you will see a 1/2 octave lower F3 after a month or any length of time. I believe that the signal boost you've engaged will not change as time goes on. Stating the reduction in Fs from 36 to 26 Hz as though it has already happened or soon will is something I will always post an opposing opinion to without the data, sorry. Not sure how you would know that the IA is more durable than the HS24, but that's pretty important info and comment-worthy in my opinion."

 

The HS24 is a larger and more expensive driver than the IA Judge, and I already replied to you about the effects of the Fs lowering.  You obviously haven't read that yet.  http://data-bass.ipbhost.com/index.php?/topic/512-8-sealed-incriminator-audio-judge-21-build/?p=12292

 

"You need +10dB of boost with a higher Fs driver than you do with a lower Fs driver... period. I am not saying this to irk you. Sorry you always seem to take it that way."

 

I think you're wrong based on the HST-18 vs. Zv4 comparison.  That's not a blanket statement you can just apply to all drivers. 

 

EDIT:

 

When comparing those two drivers, the HST-18 has an Fs that's ~23% lower, but that didn't have hardly any effect at all on the raw response or ULF performance.  According to your statement above, the Zv4 needs +10db of boost over the HST-18.

 

Capture_zpsatpz4un7.jpg

 

 

 

"Displacement is not only about Sd. A 21" driver with 9mm Xmax has less displacement than a 15" driver with 45mm Xmax. I only meant to show savings, based on your originally stated objective of keeping it simple with minimal cost and time."

 

This type of post needs to stop.  Everyone here knows this, and in my opinion posting it to this specific crowd is condescending. 

 

"You said the HS24 was less reliable, that it only has 1 spider, that the difference in Fs is nearly identical, etc., that your goals were to limit time spent, money spent and to keep it simple. No one trashed your thread. I don't know what to say to most of your objections to what's posted. I really don't. But, that doesn't mean that I'll stop posting and so sorry for your objections and accusations."

 

1.  I said "I'd give a slight edge" to the IA Judge, and I stated several reasons, such as the slightly larger VC WW, thicker VC wire gauge, twice as many spiders, etc., for why I drew that conclusion.  I went out of my way to say "Again, I'm not trying to diss the HS24 in any way, and my assessment could be wrong" and "I'm not saying the HS24 isn't durable". 

 

2.  I never compared the Fs of the IA Judge to the HS24

 

3.  My goals about limiting time and money spent were specific to the enclosures. 

 

4.  To summarize your comparison of the HS24 vs. the IA Judge:  "The 2 drivers are apples and oranges, or, to use a better metaphor, HT and car audio."

 

You calling these "car audio" drivers was negative and derogatory.  Your summary above is basically saying these drivers aren't suitable for HT. 

 

 

I apologize for thinking that that's exactly what you were referring to when saying that 2 drivers that are completely unknown to you were down to reliability. Of course, I believe whatever you say on the subject, but without this disclaimer, what else would your stated assessment have been based on?

 

It was based on the exact reasons I posted.

Link to comment
Share on other sites

This is for you and Luke, as you both seem not to grasp the concept of native + L/T signal shaping + Room Gain = flat in-room.

 

1. Luke posted a close mic of the native response of his sub. (That is the first step for designing a L/T)

2. Luke posted a response at the seats. (In-room response) (not shown in this graph)

3. Luke posted a response at the seats with his version of a L/T. (signal boost of +10dB) (derived from the posted measurement, shown in red trace)

4. We have the measured response of the HS24 from several sources (shown in dashed blue trace)

 

2wG3m5r.png

 

If you have to boost the signal to the IA sub +10dB to equal the response of the HS24 with no signal boost, then the HS24 will be the same in-room response at the seats in Luke's room as the signal-boosted IA sub. Far from being a load of crap, it is the fact of the matter.

 

NOTE: The green trace, the L/T modified response of the IA sub, is incomplete because Luke did not say what other EQ filters he used to arrive at his posted in-room responses.

 

Something isn't right with that picture.  You're level matching to the peaks, but look at the levels at and above 100hz. 

 

Additionally, the IA Judge has more inductance than the HS24, therefore that peak in the response (caused by inductance) will be higher in volume.

 

Also, even after taking into account 38% more Sd, there's no way the HS24 will have 10db extra on a 21" driver with the same amount of power applied to each driver.  For that chart to be accurate, the HS24 would have to be receiving more power AND have more cone excursion. 

 

I'm not disputing the HS24 may require less boost down low, but 10db is greatly exaggerated. 

Link to comment
Share on other sites

Q: Is the close mic you posted accurate? If so, there is nothing wrong with the graph within a dB or 2 regarding the subject of the comment. You're just not getting what I'm saying and worse than that, you're making it a personal thing. Really wish you wouldn't let your feelings get in the way of understanding.

 

Like: "That's a derogatory statement", referring to my HT driver vs car driver reference. One driver is designed for HT and the other is designed for car audio. If anything is clear to 'everyone', it should be that. That's not a derogatory statement (and certainly was not meant to be), it's a fact.

 

We did a shootout using the HST-15, BHT-15 and the LMS-R-15. The HST-15 required less L/T boost to give the same response at the seats than the BHT-15 and the LMS-R-15. The reason for this is simply that the BHT-15 and the LMS-R-15 both have an Fs of 24 Hz but the HST-15 has a lower Fs. Thus, the F3 in box, and subsequently, the native FR and the FR at the seats reflect that.

 

It has nothing whatever to do with the quality or performance of each of the drivers. Those parameters can only be seen in the comparison testing, thus the shootout.

 

Close mic:

xrObxk3.png

 

At the seats:

GGLDR2Y.png

 

The same L/T boost is applied to all 3 drivers. In order for the BHT-15 and the LMS-R-15 drivers to have the same response at the seats as the HST-15, they will require more boost in the L/T. The difference is "only" 4-5dB. That translates to 2000W vs 5-7000W, assuming the drivers could take that much more, etc. Not trivial.

 

The native response you posted vs the known measured HS24 responses we have show a 10dB difference. That's FR only. Don't read other things into the situation like inductance hump, smaller box induced higher Qtc hump, displacement, power handling, etc. If your response truly does change as radically as you have been led to believe it will, there will still be a significant difference in power requirement <20 Hz.

 

Again, not sure why you take this information as a personal attack, but seriously, you need to get a grip. It's a fact of ULF life, Fs is directly related. If the drivers have enough power handling to get you where you want to be, then great, boost away, buy big amps and everyone's happy. But, if the stated goals include price, the $ per liter of displacement and $ per watt of power heavily lean toward the SI driver. Less boxes, less drivers and less power amp. It's not even close.

 

You directly insult, belittle, and are extremely condescending to Scott but I'm the one that needs to get a grip? 

 

Are you just going to keep ignoring the HST/Zv4 data?  You're exaggerating the effect of Fs and your own graph proves it.

Link to comment
Share on other sites

So for a dummy like me that has trouble reading all the technical stuff, are you saying I should have my sub level at -10 give or take, and then adjust the gain up accordingly to offset the signal being lowered from the AVR?

 

I've got a Denon X4100a and I think the sub is like -4.5 right now and I'm powering my DIY Sub (4 cubic foot ported Cube15 from DIYSoundgroup w a HT-15) with one of the Dayton SA1000 rackmount amps.

 

Just trying to make sure I'm not screwing something up in my chain as you mentioned.

 

-10dB on SubOut only if you playback at -0dB Master Volume.  If you mainly listen at -10dB Master Volume, you can set the Sub Trim at 0 and be OK.

 

JSS

Link to comment
Share on other sites

Wow I missed a lot.

 

Bosso your graphs are from my measurements of the 24 correct? If so that is a 17cu ft box or so. Not representative of the Mariana enclosure which is probably half that. The 21 boxes Luke has are even smaller per driver still. That would be apples to oranges. The 24 would likely be about 6db down at 20Hz in a box half the siz and have a much higher resonance peak.

Link to comment
Share on other sites

Like: "That's a derogatory statement", referring to my HT driver vs car driver reference. One driver is designed for HT and the other is designed for HT BY A car audio COMPANY. If anything is clear to 'everyone', it should be that. That's not a derogatory statement (and certainly was not meant to be), it's a fact.

Corrected.

 

And, I'm betting I'm not the only one who doesn't believe that last statement.

Link to comment
Share on other sites

This is for you and Luke, as you both seem not to grasp the concept of native + L/T signal shaping + Room Gain = flat in-room.

 

1. Luke posted a close mic of the native response of his sub. (That is the first step for designing a L/T)

2. Luke posted a response at the seats. (In-room response) (not shown in this graph)

3. Luke posted a response at the seats with his version of a L/T. (signal boost of +10dB) (derived from the posted measurement, shown in red trace)

4. We have the measured response of the HS24 from several sources (shown in dashed blue trace)

 

2wG3m5r.png

 

If you have to boost the signal to the IA sub +10dB to equal the response of the HS24 with no signal boost, then the HS24 will be the same in-room response at the seats in Luke's room as the signal-boosted IA sub. Far from being a load of crap, it is the fact of the matter.

 

NOTE: The green trace, the L/T modified response of the IA sub, is incomplete because Luke did not say what other EQ filters he used to arrive at his posted in-room responses.

 

The fundamental problem I see with your comparison here is that you don't have enough information to decide how far apart each curve is on the vertical axis.  There's absolutely no physical reason that the peak output from each sub should be equal.  You have to take into account what the amp was actually doing in each measurement, and the only thing you know for certain is that in each case, the amp was operating at a constant voltage output throughout the sweep.  The voltages used may be totally different, so you can't compare these two curves on dB by dB terms.

 

The issue of relating voltage to power via impedance is a different issue, but suffice to say, I agree with Josh Ricci's approach of doing all his response measurements such that V^2/Z_nominal is always equal to 1 watt, where V is rms volts and Z_nominal is nominal impedance.  This doesn't mean the sweep is done with 1 watt of power, but it's probably the best way to be fair as with drivers, typically, sensitivity goes up by 3 dB when impedance gets cut in half but efficiency stays the same.  (See my earlier posts for definitions.)

 

Now think about what happens if you shift the curve for the Judge up until it matches the HS-24 at 10 Hz.  Note that with the information we have, this is no less correct than what you've done in your picture.  With its curve shifted up, the Judge now appears to have a lot more mid-bass (40-100 Hz) output than the HS-24, which is entirely possible despite the larger radiating area of the HS-24.

 

If the Judge has enough extra motor strength and has a lighter cone, it could easily exceed the HS-24 in efficiency above resonance.  Why is this important?  Because, if you allow for that possibility in your comparison plots, you'll see that the gap in ULF efficiency may not be all that different after all.  The same goes for the plots comparing the BHT-15, HST-15, and LMS-R-15 woofers.  There's absolutely no reason they should all have the same efficiency from 50-150 Hz.

 

Unfortunately, without T/S parameters it's hard to say what's going on, and there is another potential complication.  IIRC, Luke's measurements were taken with all the subs running.  I think it would be wise to measure with only one sub on at a time, or at least one cabinet at a time and compare with all of them running.  There might be some weird interference going on that's making the response look a bit more humped than it should.

 

Another point relates to Mms.  Increasing Mms decreases the resonant frequency of the driver, both in free air and in a sealed box, but this comes at the cost of reduced efficiency above the box resonance.  Increasing Mms also increases Qts and Qtc, so even though the roll-off is shifted to a lower frequency, the roll-off is steeper.  In fact, increasing Mms has little effect at all on efficiency at frequencies well below resonance.

 

Another way to drop Fs is to increase Cms.  But note that it's actually Fb (the sealed box resonance) that's of more interest of more interest, and it depends on the overall system compliance, which is always less than both Cms and the compliance of the air spring.  That's a real shame because increasing Cms actually increases efficiency, but unlike Mms the effect is predominantly for frequencies well below resonance rather than well above resonance.  Increasing system compliance (where possible) also decreases Qts and Qtc, and this tends to make the roll-off start earlier and but drop more gradually.  Unfortunately as noted, the system compliance will never be greater than the compliance of the box air spring.  This is essentially Hoffman's Iron Law, and indeed, ULF efficiency depends strongly on box size for small box builds.

 

So you see that the only way practical way to lower box resonance in a *small* box is to push Mms higher, and that totally kills higher frequency efficiency.  There are legitimate reasons to do this in a woofer design.  If the buyer will not be using signal shaping, then the lower resonant frequency may provide a flatter response in actual use.  This is all the more important when the product is intended to be used in a small box.

Link to comment
Share on other sites

-10dB on SubOut only if you playback at -0dB Master Volume.  If you mainly listen at -10dB Master Volume, you can set the Sub Trim at 0 and be OK.

 

JSS

 

Thank you sir.  Since the volume is normally under 65 on my Denon which has a max of 80 (IIRC), and the Denon has the level at around -5 (need to double check just for my own knowledge), I think for now at least I'm good.  But I will keep this in mind.

 

That was a very helpful piece of knowledge.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.


×
×
  • Create New...