Help choosing a 10 inch midbass driver.

[j-roadtatts 1,129]

I cooked the voice coil in one of my Aura 10 inch midbass drivers So now I need to replace them with something more heavy duty.

http://www.madisound.com/catalog/PDF/aurasound/NS10-513-4A.pdf

I know the t/s parameters should tell me which driver should be a good candidate but don't understand all of them to the fullest yet, SO this is where I am asking for help.

The Drivers will be mounted in my doors IB and will be getting 300 watts RMS. They will be bandpassed 50hz to 400hz, give or take whatever it takes to blend them.

Based on my installation I ASSUME that I need a driver with a total Q value of around .35?

I ASSUME that because I burnt the voice coil in the Aura I should get a driver that can handle a higher RMS for my long term abuse. I ASSUME that if I get a driver with a RMS of 500 watts and only feed it 300 rms long term that I will NOT burn it up as long as the signal is FLAT?

The third factor I ASSUME I should look at is sensitivity? I KNOW that this is a BAD thing to compare drivers by ALONE but still think its important.

The LAST THING I ASSUME plays a big factor is the Frequency Response of the driver?

I ASSUME that I am TRYING TO COMPENSATE for a POOR FR curve by turning up the volume to hear certain frequencys that are lacking? I ASSUME this played a big part in me destroying the other driver, is why I bring this up also. I DON'T know per say what FR curve I am after due to my lack of experience, BUT ASSUME it wasn't what I had with the Auras? I may need a better EQ or a better tuning ear, one or the other or BOTH?

My budget is 400 dollars for two 10" drivers. I am interested in the Daber Audio 10's for the following reasons, RMS rating, QTS of .35, and a sensitivity of 88. I ASSUME the MOST important thing for my install is going to be thermal power handling, a.k.a. RMS?

SO PLEASE HELP me get the correct driver for my application. My budget is 400 dollars for the drivers.

[j-roadtatts 1,129]

I guess I should add that only one driver got cooked. The other driver has blueing of the coils but NO BURNT spots and still tests good SO I can re-use it if I replace the one blown one.

I can only ASSUME that because I had them on seperate channels that the one that blew had the gain set alittle higher?

I Also want to add that I was listening to music that was heavily boosted in the studio when I blew the driver, I KNEW I was pushing the driver and probably NO driver will handle this abuse.

SO my next question is should I just replace the one Aura with a new one, and use the one thats still good, turn down the gain alittle and be more responsible with the volume? This is definenetly the most affordable option.

What do you guys think?

[ssh 924]

Will you be happy with the output if you turn the gain down a bit? Or are you going to want more?

[Impious 1,476]

I wouldn't go for the Daber given your intended application. Just going by inductance, the driver will have a self-induced roll-off beginning around 280hz, which is more than a half-octave below your desired bandwidth.

I'm really not sure where this Qts of .35 thing came from?

Qts alone isn't going to help you entirely. There is a significant difference in response between a driver with a Qts of .35 and an Fs of 70hz, and a Qts of .35 and an Fs of 35hz. The 2nd driver would have an F3 approximately one octave lower than the 1st in the same alignment. So you need to consider both the Qts and Fs to really determine the response. Really, what you're actually looking at is Qtc and Fc....in an ideal "IB" environment, Qts = Qtc and Fs = Fc. However, your door may not end up being a true IB environment. So the Qts and Fs may be different than the Qtc and Fc. You need to consider how your door is going to affect the driver, and the effect will be slightly different for every driver depending on it's Vas.

The 2nd thing to consider is differences in driver behavior with different Qtc's. A lower Q is going to be more overdamped with a shallower roll-off and generally the least output above F3. Subjectively this is typically said to be a little "dry" or "sterile". A higher Q is going to be a little more underdamped with a steeper rolloff but slightly more output above F3. Subjectively this would be a little "thicker". A Qtc of .5 has optimal transient response at the cost of peak output and a higher F3. A Q of .707 has optimally flat frequency response, the lowest F3 and transient response that is still considered good. A lower Qtc typically has slightly more output at the lower frequencies (somewhere below F3) due to it's shallower rolloff even though the F3 may be slightly higher. F3 is the frequency at which the response has rolled off to a level of -3db.

So, the main point of those brief paragraphs was to explain that your target Qts may or may not be ideal depending upon where you want your final Qtc to end up, and also that you can't look at Q alone you also need to consider Fs/Fc. If you were pushing the power limits to get the output you wanted in the midbass, a higher Q alignment may give you a little more output in the 60-100hz (depending on the Fc).

Another thing to look at is inductance, or Le. A driver's inductance and resistance create a classic 1st order (6db/oct) lowpass crossover. You can determine the corner frequency with the formula;

Frequency = Re/(2*Pi*Le)

*Le in henries

Ideally you would want this to be one octave above your intended crossover frequency. Additionally the response should be relatively flat within this bandwidth. If you can find some distortion and other measurements to make sure it doesn't experience any cone breakup or distortion issues within your intended bandwidth, that would be a bonus.

I would also take displacement into account. The more the merrier. For example, with the Sd of that Aura, it will take 9mm of excursion to reach 100db. If you plan on listening at or above levels of 100db, you will need a driver with atleast 9mm one-way Xmax (with 346cm^2 of Sd). The louder you want to go, the more excursion you will need.

Sensitivity.....meh. The only reason I ever really care about sensitivity is in regards to Hoffman's Iron Law.

That's what I can think of off hand.

[lithium 358]

I wouldn't go for the Daber given your intended application. Just going by inductance, the driver will have a self-induced roll-off beginning around 280hz, which is more than a half-octave below your desired bandwidth.

I'm really not sure where this Qts of .35 thing came from?

Qts alone isn't going to help you entirely. There is a significant difference in response between a driver with a Qts of .35 and an Fs of 70hz, and a Qts of .35 and an Fs of 35hz. The 2nd driver would have an F3 approximately one octave lower than the 1st in the same alignment. So you need to consider both the Qts and Fs to really determine the response. Really, what you're actually looking at is Qtc and Fc....in an ideal "IB" environment, Qts = Qtc and Fs = Fc. However, your door may not end up being a true IB environment. So the Qts and Fs may be different than the Qtc and Fc. You need to consider how your door is going to affect the driver, and the effect will be slightly different for every driver depending on it's Vas.

The 2nd thing to consider is differences in driver behavior with different Qtc's. A lower Q is going to be more overdamped with a shallower roll-off and generally the least output above F3. Subjectively this is typically said to be a little "dry" or "sterile". A higher Q is going to be a little more underdamped with a steeper rolloff but slightly more output above F3. Subjectively this would be a little "thicker". A Qtc of .5 has optimal transient response at the cost of peak output and a higher F3. A Q of .707 has optimally flat frequency response, the lowest F3 and transient response that is still considered good. A lower Qtc typically has slightly more output at the lower frequencies (somewhere below F3) due to it's shallower rolloff even though the F3 may be slightly higher. F3 is the frequency at which the response has rolled off to a level of -3db.

So, the main point of those brief paragraphs was to explain that your target Qts may or may not be ideal depending upon where you want your final Qtc to end up, and also that you can't look at Q alone you also need to consider Fs/Fc. If you were pushing the power limits to get the output you wanted in the midbass, a higher Q alignment may give you a little more output in the 60-100hz (depending on the Fc).

Another thing to look at is inductance, or Le. A driver's inductance and resistance create a classic 1st order (6db/oct) lowpass crossover. You can determine the corner frequency with the formula;

Frequency = Re/(2*Pi*Le)

*Le in henries

Ideally you would want this to be one octave above your intended crossover frequency. Additionally the response should be relatively flat within this bandwidth. If you can find some distortion and other measurements to make sure it doesn't experience any cone breakup or distortion issues within your intended bandwidth, that would be a bonus.

I would also take displacement into account. The more the merrier. For example, with the Sd of that Aura, it will take 9mm of excursion to reach 100db. If you plan on listening at or above levels of 100db, you will need a driver with atleast 9mm one-way Xmax (with 346cm^2 of Sd). The louder you want to go, the more excursion you will need.

Sensitivity.....meh. The only reason I ever really care about sensitivity is in regards to Hoffman's Iron Law.

That's what I can think of off hand.

excellent post.. where did you learn all of this !?

[j-roadtatts 1,129]

Thanks for the help Impious. You are World Class.

I will do some searching and see what drivers I think will suit my needs and report back.

I plan on buying the loudspeaker cookbook soon and doing some serious reading also. Thanks again.

[Kanaworm$ 292]

This might be something to look at .... Solen Electronique Inc.=

The specs

http://www.solen.ca/pdf/peerless/850146.pdf

[Duran 387]

Thanks for the help Impious. You are World Class.

I will do some searching and see what drivers I think will suit my needs and report back.

I plan on buying the loudspeaker cookbook soon and doing some serious reading also. Thanks again.

It's a very good read Alibris: Used Books, Used Textbooks, Rare & Out-of-Print Books

[j-roadtatts 1,129]

What is the correct equation for calculating qtc? Better yet does anyone have a link to a caculator that I should be using, all the ones I googled ask for different t/s numbers and come up with different awnsers.

I looked at the Peerless and think it would make a good candidate for testing based on the high VAS and low FS? The specs don't show XMAX but I came up with 18mm, using the following math. VC lenght minus height of gap. Is this correct? Also they will handle 200watts RMS.

[Impious 1,476]

Qtc = Qts * [ ( Vas / Vb ) + 1 ]^.5

[j-roadtatts 1,129]

So in other words I need to figure out the Vb of the door. How accurate do I need to get with the Vb? I can get pretty close with pulling numbers and OR I can stuff the door full of packing peanuts and measure their volume to get even closer. Anyone have any other great ideas?

[Impious 1,476]

I would just get rough measurements. Your door is not completely sealed, it's more of a leaky sealed box, so your physical Vb is going to be different than the effective Vb. Without knowing the loss factor (i.e. the amount of the leaks) you aren't going to get it 100% accurate anyways.

IIRC you basically have a pod built onto the door? You might get a rough measurement of that as well.....one thing you may consider doing, depending on what mids you are looking at, is to seal the pod. If you knew the volume you could calculate how that would affect the Qtc and Fc. You might be better off using a lower Qts lower Fs driver in a sealed enclosure to bump the Qtc up to the .6-.7 range since a high Q driver is going to be harder to find, and in your situation I would probably go for a higher Q alignment to try to get some more response in the 50-100hz range.

[cobra93 47]

I followed your thread and wish I could make enough room for some tens in my doors. Then i read you lost one, sorry to hear that.

I'm following this thread also because I'm looking for the best mid bass I can get as well.

It looks like you're getting some good help, Good luck with what ever you choose next.

[j-roadtatts 1,129]

I measured the door to be around 2-2.5 cu ft. , and I guess the pod to be around .5 cu ft..

second I am willing to cross the midbass as low as 200hz, so this puts the Daber or similar back on the table. I calculated the self induced roll off of the Peerless at around 300 and the Aura doesn't give a Le value.

Third I calculated the Qtc of the three drivers that are mentioned in this log to make sure I am understanding this right.

On the Aura I came up with a Qtc of .45 in 2.5 cu ft. Vb and a Qtc of .7 in a .6 cu ft. Vb.

On the Daber I came up with a Qtc of .4 in 2.5 cu ft. Vb and a Qtc of .7 in a .5 cu ft. Vb.

The Peerless I came up with a Qtc of .61 in a 2.5 cu ft. Vb and a Qtc of .7 in a 1.75 cu ft. Vb.

So IF i understand this than I think the Peerless is going to do best with the enclosure alignment I have now? BUT IF i make a sealed pod in the .5 cu ft to .6 cu ft range than the Aura or Daber are going to shine? Am I correct?

[eggyhustles 33]

Sundownz mentioned someone in Russia using sundown sa-8's for midbass duty

why not try out the 10's?

[j-roadtatts 1,129]

I ASSUME my math and assumptions on Q values are correct, BASED on no one answering????? I ASSUME this is the math box programs use to gragh out FR. I will admit I never thought to run anything other than my sub through winisd. I will redownload winisd today since I erased it due to a bug before.

On another note, I have made MAJOR break throughs this week in my quest for true SQ. I realize where I went wrong the most is too shallow of a slope on the low end of the Aura's. I will admit that my novice ass liked the extra colored sound that I was getting with the shallow slopes, BUT I had to remind myself this isn't an SPL car. My stupid novice ass willl take ALOT more time dialing everything in this time, I lost track of my goal I guess. Hard to kill the monster.

SO where does this leave me in my quest for a new set of 10" midbasses, you may ask. Well now that I spent the last couple weeks tuning my ear and brain for SQ, and getting back on track of my SQ goals, I think I should give the Aura's another chance. I THINK I should build a sealed test pod at .6 cu ft and do some A B comparisons between the test box and my door with better crossover settings AND the sub turned off, SO I can develop a more subjective opinion of what SQ sound I am really after out of the mids.

Let me know if you think I am on the right track. Thanks for helping out the:noob: .

[cobra93 47]

I ASSUME my math and assumptions on Q values are correct, BASED on no one answering????? I ASSUME this is the math box programs use to gragh out FR. I will admit I never thought to run anything other than my sub through winisd. I will redownload winisd today since I erased it due to a bug before.

On another note, I have made MAJOR break throughs this week in my quest for true SQ. I realize where I went wrong the most is too shallow of a slope on the low end of the Aura's. I will admit that my novice ass liked the extra colored sound that I was getting with the shallow slopes, BUT I had to remind myself this isn't an SPL car. My stupid novice ass willl take ALOT more time dialing everything in this time, I lost track of my goal I guess. Hard to kill the monster.

SO where does this leave me in my quest for a new set of 10" midbasses, you may ask. Well now that I spent the last couple weeks tuning my ear and brain for SQ, and getting back on track of my SQ goals, I think I should give the Aura's another chance. I THINK I should build a sealed test pod at .6 cu ft and do some A B comparisons between the test box and my door with better crossover settings AND the sub turned off, SO I can develop a more subjective opinion of what SQ sound I am really after out of the mids.

Let me know if you think I am on the right track. Thanks for helping out the:noob: .

Hey j, sounds like a good idea.

You're using the PXA-H100, what's the crossover slope on that ( 12db per octive)?

Are you using the crossovers on the alpines?

If not, try using both, combine them for a 24 db per octive.

I was just looking back in my thread and found where you stated

They blended really well crossed from 50 to 400 with a 12 db slope, but I should of had a steeper slope down low for wanging on them.

I'd try a test box with the one that's still working.

I've always preferred some type of enclosure for my speakers for that fact if nothing else.

However if you smoked the coil, the box probably won't help.

You stated you were showing them off to your buddy (a drummer), how much harder were you pushing them than normal?

With the speaker open to the door cavity can you see anything in your mirrors?

One last thought, with the setup in my cobra, I always turned the 8w3's down when I was cranking the rear sub (eclipse) because I couldn't tell if I was trying to kill them from one song to the next.

Good luck , I hope your next reply is that you got it sorted out.

[Impious 1,476]

I ASSUME my math and assumptions on Q values are correct, BASED on no one answering????? I ASSUME this is the math box programs use to gragh out FR. I will admit I never thought to run anything other than my sub through winisd. I will redownload winisd today since I erased it due to a bug before.

I haven't ran the numbers myself, but using an accurate modeling is a good idea. It helps to visualize the differences.

Building some test boxes to see what you think of the differences would also help you along.

[j-roadtatts 1,129]

I ASSUME my math and assumptions on Q values are correct, BASED on no one answering????? I ASSUME this is the math box programs use to gragh out FR. I will admit I never thought to run anything other than my sub through winisd. I will redownload winisd today since I erased it due to a bug before.

I haven't ran the numbers myself, but using an accurate modeling is a good idea. It helps to visualize the differences.

Building some test boxes to see what you think of the differences would also help you along.

Thanks for the help guys. I didn't really expect anyone to do the math, I just was making sure my understanding of the physics and concept of theory where solid. I am pretty sure I am good.

I will report back with the test results.

[sandt38 448]

I ASSUME my math and assumptions on Q values are correct, BASED on no one answering????? I ASSUME this is the math box programs use to gragh out FR. I will admit I never thought to run anything other than my sub through winisd. I will redownload winisd today since I erased it due to a bug before.

On another note, I have made MAJOR break throughs this week in my quest for true SQ. I realize where I went wrong the most is too shallow of a slope on the low end of the Aura's. I will admit that my novice ass liked the extra colored sound that I was getting with the shallow slopes, BUT I had to remind myself this isn't an SPL car. My stupid novice ass willl take ALOT more time dialing everything in this time, I lost track of my goal I guess. Hard to kill the monster.

SO where does this leave me in my quest for a new set of 10" midbasses, you may ask. Well now that I spent the last couple weeks tuning my ear and brain for SQ, and getting back on track of my SQ goals, I think I should give the Aura's another chance. I THINK I should build a sealed test pod at .6 cu ft and do some A B comparisons between the test box and my door with better crossover settings AND the sub turned off, SO I can develop a more subjective opinion of what SQ sound I am really after out of the mids.

Let me know if you think I am on the right track. Thanks for helping out the:noob: .

There are a couple things I would like to point out here...

First, any modeling program is going to force a midbass driver into a ported application, based on the higher Fs which boosts EBP. Please bear in mind that EBP (efficiency bandwidth product) is simply Fs/Qes. The majority of the modeling programs are looking to achieve a very flat FR and the lowest F3 possible. So take that with a grain of salt. Why? Because you don't really need to worry about the low F3. First, this is a midbass driver, so an F3 of 30Hz is going to be unnecessary. Second, cabin gains in the car will help fatten up anything around 100Hz or less (I know I will get an argument from someone saying 80Hz, and vehicle specific, or whatever... I am just making a quick generalization here), so you don't necessarily want to see a flat anechoic response curve that low, or you will have to skinny up the bottom end via EQ if you are looking for a flat response.

Second, I suggest you not focus on a single crossover slope. Different vehicles with different driver combination will respond differently to various slopes. While we like to use steep crossover slopes on tweeters, and midranges, we don't necessarily want to use them on midbass drivers. Why? Well, the lower we extend in the frequency range the more air the driver has to move in order to maintain linear output. Tweeters have very small diaphragms, and too much extension can cause distortions, and can be mechanically dangerous to the driver (if you would like to see a ring radiator I have one torn apart, and I may even have a silk dome torn down, just let me know and I will get a few shots for you.). So rather then demand them to go beyond their physical limitations we elect to cross them sharply. The same can be said with midrange drivers, and I don't mean extended range midrange/midbass drivers, I mean 3 or 4 inch midrange. We don't want to start seeing them overextend themselves and fall apart (audibly I mean), but where we want to cross a tweeter at say a 24 octave slope, we would be more comfortable crossing the midrange at 18 or even 12db slopes. Keep in mind, the frequency range per octave is very broad when we get into higher frequencies, where they get much smaller at lower frequency ranges... for example a 1 octave slope from a 60Hz crossover point is only 30Hz, where a 1 octave slope from a 5000Hz crossover point is 2500Hz, which is 1 reason why we are more comfortable using a more gradual slope when we cross lower and lower. With a midbass driver, it should be pretty capable of getting down low without much breakup, and our slopes are significantly more steep, just based on how narrow an octave becomes at such low frequencies.

So don't go worrying so much about your bottom end slope with regards to the crossover point in the car. It is really something you should just play with until you find what works best.

I think the reason you pushed the Auras is because they are pretty sterile. They just don't seem to deliver quite as loud as a distorted driver. I like to remind people that the very limits of our hearing are the places where music becomes emotion, not sound. This is why a linear, undistorted subwoofer (as well as midwoofers) sounds so sterile, because it is more of a feeling then an audible perception. So you shouldn't try to hear the music, in as much as you should try to gather the emotion of it... That is what sound quality is all about.

[j-roadtatts 1,129]

Thank you for your thoughts Sandt38. For THIS set-up, I am trying to acheive the best sound stage possibly I can, even if it takes changing my perpective of "good sounding" to sterile.

I know as soon as I change midbass drivers or enclosure setup I will have to reblend everything. My biggest decovery with the steeper slope on the BTL is how much it cleaned up the sound, whether or not it will blend that way when I am done is another story.

I like the warmth that the sub gives with the shallow slope, IS why I am pretty sure I just need to change My Q alignment on the midbass and hopefully I can get the warmth without the sub. The TRUE sub notes sound AMAZING even with the steep slope, SO I am just missing the midbass notes up front to give me the sound stage I am after.

[sandt38 448]

I was looking at you sig, and now I have to ask, are you using the extended range drivers for midrange and top end? I thought I saw a build in your car that was using TypeR comps before... maybe it was someone else, and if so I apologize I am still trying to get to know people here. But I remember the trunk build and the attention to detail, and I could have sworn you had a comp set in there.

I looked through your gallery and there were so many pics, I got lost in them

I'll need to revisit your page and see the car again.

Take a look at this chart. If you look on the bottom of the chart you will see descriptions of the sound qualities of the various frequency ranges listed there. By using these descriptors, perhaps you can find the frequency ranges you are looking to accentuate and perhaps play with your EQ a bit and see if that is the range and the result you are looking for.

[j-roadtatts 1,129]

Thanks for the chart Sandt38. I have been using one just like this for the last few weeks for learning what I am hearing.

I know my build log got kinda long.lol. the type-r's where my last setup. Heres a few pics of the current setup.

I had the xovers set at:

FI Sub-50hz 12db slope

Aura mids-50hz 12db slope ----400hz 12db slope.

Fostexs highs-500hz 12db slope

I had a little underlap between the mids and highs.

[sandt38 448]

Actually, based on the 12DB slope and acoustic coupling, I would guess the transition is pretty damned smooth.

I knew you had the TypeRs. I am involved in a few midbass discussions, but I remembered your install.

And I have an awful memory

[Impious 1,476]

First, any modeling program is going to force a midbass driver into a ported application, based on the higher Fs which boosts EBP. Please bear in mind that EBP (efficiency bandwidth product) is simply Fs/Qes. The majority of the modeling programs are looking to achieve a very flat FR and the lowest F3 possible.

I probably should have mentioned when I suggested a modeling program, to model them in your airspace and not look at the "recommended" enclosure alignments for this reason.

The physics are the same, so it will "model" the low end response in any alignment accurately but the programs suggested alignments will be biased to ported as sandt point out.