|
Post by aespeakers on Nov 12, 2005 17:22:11 GMT -7
Actually I'm over 400 IB15's sold now and nearing in on 500 pretty quickly.
John
|
|
|
Post by shomrighausen on Nov 14, 2005 16:18:48 GMT -7
I welcome your offer for replacement drivers. Just let me know how you want to handle the return of the ones that I have now and I'll get them to you.
Also, what method do you use to test the drivers? I was also wondering if anyone would be willing to pay a bit extra for a spec sheet for each individual driver ordered ($5-10?) If possible, I'd like to test the drivers that I have using your test method before shipping them back... then you could compare when you get them back and see if we are on the same page.
All I can think of on the Mms is that there are parts of the driver that are interacting with each other in a manner other than what you expect. I have consistant results when measuring the drivers with different methods.
As far as the magnet structure not being fully charged, if they are charged to some percentage less than optimal (say 70%) but they are consistant at that percentage, then the results would be consistant.
I'm sure that the drivers that I have not are broken in, and I don't see the spider loosening up much more. I also test the drivers 'suspended' about 2 feet above the floor with no restriction for the pole vent (two feet between it and the floor).
I look forward to hearing from you. If you have specific instruction, or questions - you should still have my e-mail address from the PayPal purchase.
|
|
|
Post by rbuszka on Nov 15, 2005 10:51:21 GMT -7
While you still have the drivers, perhaps you should run some low frequency sine wave through them at pretty close to X-max (like 20 Hz or something) and then measure the drivers again before you send them back, to see if that can change anything.
|
|
|
Post by shomrighausen on Nov 15, 2005 11:00:02 GMT -7
I've run more than 15 hours of different sine waves and slow sine wave sweeps (10 minutes to sweep from 10Hz to 30Hz repeating for 8+ hours) through these drivers using 300-400 watts of power per driver. They are easily broken in at this point, my tests were taken after the break-in.
|
|
|
Post by aespeakers on Nov 15, 2005 16:29:15 GMT -7
I will get labels out to you to return the drivers as soon as I get the new ones ready.
I use praxis to test the drivers. I get pretty close results with delta mass or delta volume. I still wanted to know how you attached the nickels. If they are vibrating at all that will throw off the results.
With Mms there really isn't anything that can interact. It is a physical mass of the parts. No way around that. You can easily use Mms to double check your measurements. If Mms isn't very close to the physical weighed out mass of the parts it usually indicates an error in the test somehow.
Can you send me the impedance curves you measured? That may help me to figure out why things are off here.
In regards to the breakin, these will take a very long time go get to the final specs. A typical spider is lightly coated with whatever resin is used. These have a very stiff ring in the middle that is almost solid. They are the ones Nick at Lambda switched to. He liked them because the car guys that put in 2000W per driver and beat on them all day had trouble bottoming them. Even in a situation like that they would take months to fully break in. The inside corrugations are coated with a very heavy layer of resin. You can barely flex it by hand even. It almost has to loosen up from the outside in. The longer you drive it, the more that ring of resin loosens up more toward the middle. Once they do get to the final specs though they are very durable and last forever and are very linear.
In trems of magnet charging, even if all drivers were to 70% the parameters would be all over. If a magnet is not fully charged it tends to be "spongy" was Nick's word to describe it. Back EMF from the coil will make a mess out of the flux in the magnet. You'll never be able to repeat the same test results twice. The only way to get 70% charge correctly on a magnet is to fully charge and then demagnetize to 70%.
John
|
|
|
Post by shomrighausen on Nov 15, 2005 17:17:31 GMT -7
Here are some quotes the Parts Express forum. The poster is the original maker of the WT1, and a partner in the WT2 (goes by csaudio on the PE forum). ---------------------- "The WT2 measures at a few milliamps. If the mfgr measures Fs at some watts that will explain the difference. I had a manufacturer with this problem many years ago. He was used to testing Fs at several volts and the WT1 tested at milli-volts. He was unhappy but there was no help for him as he was testing outside of accepted small-signal Thiele/Small theory. " "In your case it may just be a matter of the driver not being broken-in. I suggest applying a 15Hz signal that makes the driver excurse +/- Xmax for at least a couple of hours. My practice is to excurse a driver overnight. Then you have to let the driver cool for a few hours. Then test for Fs, Qt. " "If at the end of all this the WT2 is very different than your driver's specs you can bank on the WT2 results. The WT2 is +/- maybe 2% max from the really high-dollar Leap/LMS type gear for about 1/10 their price." "I have looked at you results. Fs = sqrt(1/(Mm*Cm)) Cm is proportional to 1/Vas. So, your Fs is off because Vas is off by a factor of 2+. Vas is a function of the spider. So, the problem is that the spider in your production model is twice+ as stiff as specified and that makes Fs twice as high. " ---------------------- This is consistent with what you are telling me about the spider. The problem that I have is that these drivers were driven for 15+ hours with 300-400 watts per driver. I'd say that this is longer than just about everyone out there - my drivers are still far from published spec. I'd also guess that they will not loosen up much further under 'normal' conditions within a manifold IB. Have you checked any of your recent production IB15s to see how close to published spec they are? I'd understand if these were 20-30% off from spec, but the Qts is about 230% (!!!) off from spec - and that's after breaking them in for 15+ hours. You said that these spiders have a stiff ring of resin in the center that will take months for people that are putting 2000 watts into them, how do you expect that these drivers will loosen up in an IB setup with people driving them with maybe 500 watts per driver? Will they every meet spec for these people? I do look forward to receiving the replacement drivers, measured at least CLOSE to published spec. If you want to send me labels now, I will return four of the drivers now (the ones that are not in a manifold), and the other four when four replacement drivers come in. That way I'll at lease have something to listen too.
|
|
|
Post by gonefishin on Nov 15, 2005 21:50:26 GMT -7
Have you checked any of your recent production IB15s to see how close to published spec they are? I'd understand if these were 20-30% off from spec, but the Qts is about 230% (!!!) off from spec - and that's after breaking them in for 15+ hours. You said that these spiders have a stiff ring of resin in the center that will take months for people that are putting 2000 watts into them, how do you expect that these drivers will loosen up in an IB setup with people driving them with maybe 500 watts per driver? Will they every meet spec for these people? I do look forward to receiving the replacement drivers, measured at least CLOSE to published spec. If you want to send me labels now, I will return four of the drivers now (the ones that are not in a manifold), and the other four when four replacement drivers come in. That way I'll at lease have something to listen too. Hmmm...good questions shomrighausen. I know I'd be willing to pay...a little more (but hopefully not too much) to get four drivers that were broken in and each measured close to spec. But if these are indeed this tough to break into spec...I doubt the use I would give them would ever get them within working parameters. So any helpl that would be available would be appreciated. again...thanks for being adults guys. John (AES), it's also nice to see you stand behind and along with your products and customer. take care, dan
|
|
|
Post by hdjunkie on Nov 17, 2005 11:01:30 GMT -7
So, these will literally take years to break in to spec with typical 4-6 hours a week in an HT environment. Tests and deductive reasoning of minds in the know have narrowed it to a stiff spider and John has confirmed it.
Obviously the FS of 16hz is a huge selling point.
Breaking them in with test tones for hours on end for months would surely force my wife to break my head in.
Silver lining...Everyone who already has them installed and really likes them, will find that they will get even better as time goes on. Provided they have always been this way and this is not a recent production error.
With that final thought in mind, aren't they still a good deal just needing a little more amp headroom and EQ until they do fall into spec? (although a normal break in peiod would certainly be more desireable)
Honestly, and with no disrespect to John, would I be better off with the Dayton IB 15's? Considering the benifit of the lower FS NOW as opposed to what the AE IB 15s may/will sound like in a couple + years.
I know me... by the time two years is gone I'm going to already have upgraded drivers and sold what I have to a freind.
|
|
|
Post by ThomasW on Nov 17, 2005 11:29:30 GMT -7
Let's not forget that room gain plays a role in the output of the lowest frequencies.
And since IB subs use more of their excursion than 'normal' subs, their break in time will certainly be significantly shorter than "years" (think a couple of months max)
The price and build quality of the AE-IB15"s is very impressive compared to mass manfactured drivers with stamped frames.
|
|
|
Post by hdjunkie on Nov 17, 2005 11:51:45 GMT -7
You are the man! Thanks for putting an educated spin on my thoughts. So, provided the stiff spider is the cause of the anomoly (it appears to be) The AE IB 15 is the deal of the day. I feel much better
|
|
|
Post by shomrighausen on Nov 17, 2005 12:26:03 GMT -7
As far out of spec as these drivers are, you'd need some EQ at 25Hz and below to make them the have the 'same' roll-off as the published specs. With the problem below 20Hz it would require a different (and more expensive) parametric EQ than the BFD (which only goes down to 20Hz center frequency). Depending on the size of your room, and how low you want to be 'flat', it could be a substantial amount of EQ required.
An alternative would be to lower the level using broad cuts (two octaves or 120/60) four octaves apart (20Hz, 80Hz, 320Hz, etc.) By lowering all of these center frequencies it would have the same effect as boosting below 20Hz. The 20Hz center would cut less as you went lower… Which is what I’ve done for now – one down side is having to turn up your gain on your amp and preamp level going to the sub, another downside is that it uses a handful of your 12 filters (though you could always run one channel into the other for a total of 24 filters if you needed to.) It does a nice job of ‘boosting’ to flatten below 20Hz. It only gives you ‘flat’ response if you have a consistent roll-off below 20Hz that can be compensated for with some amount of broad cut at 20Hz.
I have a driver that I’ve been sending a 10Hz sine wave near Xmax for the past couple of days – on top of the 15 hours of slow (10 minute) sweep from 10Hz to 30Hz. Each 24 hour period that this driver is receiving this 10Hz sine will flex the spider 864,000 times (10 per second x 60 seconds x 60 minutes x 24 hours). My plan is to let it run for four days bringing the total to 3,456,000 cycles. Plus the 15 hours of break-in before (about 1 million more at 20 Hz average for 15 hours). I’ll let the driver cool overnight and retest the driver. If this isn't enough to break-in the driver, I doubt that it will ever 'break in'.
My guess is that the specs may be closer to published specs - but will be FAR from published specs. They didn’t move much after 15 hours of break-in. If they are far from published specs after the extra four days, I’d say that John probably used a few of the wrong parts when he built these drivers. I really WANT these drivers to magically ‘break-in’ and measure close to published spec. Running break-in on these drivers for days/weeks isn’t acceptable for me. I’ll post my results when the four days is up.
If the new drivers are close to published specs when I receive them, I’ll be happy.
|
|
|
Post by shomrighausen on Nov 20, 2005 8:42:41 GMT -7
After four days of break-in added to the original 15 hours of break-in (one driver tested)...
Fs 29.89 Qts 1.096
It doesn't look like these drivers will make it to published spec. Not much of a move from what I had tested after 15 hours of break-in. I'd guess that this driver is as 'loose' as it will ever be (under normal conditions).
|
|
|
Post by hdjunkie on Nov 20, 2005 12:06:19 GMT -7
Wow! Somebody got some splaining to do. If the spider is the culprit thats completely unacceptable for me.
111 hours is the equivalent of 18.5 weeks of normal IB use for me and of course thats not all program material at max xmax so I would conservatively add 50% making it ~30 weeks. and not even close to spec.
|
|
|
Post by ThomasW on Nov 20, 2005 12:50:28 GMT -7
We need to figure out the discrepancy between John's hand weighed measurements of the Mms, and the Mms in Steve's testing.
|
|
|
Post by shomrighausen on Nov 20, 2005 14:53:16 GMT -7
That has bothered me as well. For some reason, after the LONG break-in, this woofer has a Mms of 201.8944g. The only thing that I can think of is that there are parts of the driver that are SO stiff that without substantial amounts of break-in, they are not moving (effectively stationary). After LOTS of break-in, they are counted as part of the Mms.
all specs for this driver:
fs (16) 29.8922 Qms (4.9) 8.1086 Qes (.48) 1.2676 Qts (.44) 1.0963 VAS (467) 144.6134 Re (5.6) 5.3145 Le (1.25) 2.1577 Mms (218) 201.8944 BL (16) 12.6085
I don't understand why the Mms has such a dramatic change from previous tests, but it is consistent regardless of test method on this driver.
Now that the Mms is close to published, and the driver is well broken in, and the other specs are still far from published spec - I'm just waiting for some return labels from John. I'm hoping to get four now so I can get these four sent back to him before the 'new' ones are finished. Then I won't have to ship all eight back at once and be without drivers for a couple of weeks shipping time. Four at a time would work out better for me.
|
|
|
Post by gonefishin on Nov 22, 2005 7:32:57 GMT -7
I sure hope the current batch is on spec. I'm on the list for four
|
|
|
Post by aespeakers on Nov 22, 2005 18:34:25 GMT -7
Steve,
I replied on my forum but figured I would comment here again also. I asked you a few questions that were not answered. Most important, can you send me the impedance curves you measured? That may help me to figure out why things are off here. If you have blips or double peaks in the curves or any irregularities that will greatly change the results. I also wanted to know how you attached your nickels to the cone. If they vibrate at all they will caue results to be way off. Also, what are you specifying for the diameter of the cone when you are doing the measurements? Either specifying the wrong diameter or irregularities in the impedance curve will cause everything to be way off.
In regards to the shipment, I let you know that I would get the drivers done, then send you the labels. As soon as I see from the tracking info the the 8 are on the way back I will get your 8 new drivers out. It will take at most 2 days to get to you in Iowa from Wisconsin.
John
Obviously something is very different if you are measuring results for Mms in the 132 gram range, then 154grams, now over 200. Your Bl changed from 10.74 to 12.6 also. This shows the test method cannot be repeated accurately so there must be an error.
|
|
|
Post by shomrighausen on Nov 23, 2005 8:37:35 GMT -7
The impedance curves look normal... no blips or double humps... one hump at fs, and the normal rise as frequency increases. Where would you like me to send the curves? I can try to post them here, or send them to you via e-mail?
I used 13" for the diameter of the cone - WT2 shows a calculated Sd of 856 (855 is published spec).
I have tested these drivers with the nickels spread around the dust cap (on the cone itself, and on the dust cap). I've also made stacks of nickels, fastening them with rubber bands and placing them on the cone. When I do this, I try to split the stacks up into groups of three and place them with equal spacing around the cone. I measure the weight of the stacks with a digital scale. Regardless of which method I use for the nickels, I get the same results. I don't think they are vibrating using the low signal level that is used by all of the software that I've used to test.
The Mms puzzles me as well - my thought is that there are parts of the suspension that are so stiff that they aren't moving until the driver is WELL broken in. This would make sense, as the Mms has increased as the driver is broken in for longer periods of time. The Mms measurement is repeatable on the driver between break-in sessions - once they are broken in for a longer period, they have higher Mms. It never goes down, from what I've measured.
I'm just reporting what I've measured. I feel that I've done a good job of using valid measuring practices. I've also measured these drivers using different software (as outlined in previous posts). After the comments on the PE forum from Darren (from Parts Express), and from the maker of the WT & WT2; I feel that the accuracy of measurements from the WT2 (and the WT) are right up there with more expensive testing platforms.
I'm excited to hear from you after building the replacement drivers. I realize that it is taking two weeks to build them this time (though you built the first eight in two days). When I receive the drivers I'll test them and compare the results that I measure with the measurements that you are including with each driver.
|
|
|
Post by gonefishin on Nov 23, 2005 12:38:25 GMT -7
The impedance curves look normal... no blips or double humps... one hump at fs, and the normal rise as frequency increases. Shroom...there's a hump at fs...but which fs? 16Hz or 29Hz? thanks, dan
|
|
|
Post by ThomasW on Nov 23, 2005 12:47:19 GMT -7
|
|