|
Post by shomrighausen on Nov 2, 2005 9:04:46 GMT -7
Has anyone taken T/S measurements of the IB15 woofers? I received my batch of them (8 ohm version), and the specs seem to be off 'a bit' from published specs...
|
|
|
Post by ThomasW on Nov 2, 2005 15:01:01 GMT -7
Are you playing games? Just publish what you measured and lets see...
Most people are amazed to get a well designed and well built driver created by Nick McKinny for $100ea. As a result they aren't too fixated on minor deviations from the published in the T/S parameters.
If you're not happy there's certainly a line of people that will pleased take them off your hands.
|
|
|
Post by shomrighausen on Nov 2, 2005 15:25:04 GMT -7
I wasn't playing games, I just didn't want to raise any red flags if I had a 'bad batch'... average for the first four drivers that I measured - the published are in ( )
fs 32.24 (16) Qms 4.43 (4.9) Qes 1.2773 (.48) Qts 1.0201 (.44) VAS 190.57 (467) Re 5.50 (5.6) Le 2.906 (1.25) Mms 132.63 (218) BL 10.743 (16)
The ones in bold are WAY off spec... yes, I've tested these four drivers multiple times, and I get consistantly 'off' results. These measurements were taken using a PE Woofer Tester (original, not WT2). I have the means to measure them 'by hand', but the WT has been good to me on all other drivers that I've tested (over 100).
It almost seems that there may have been some 'wrong' parts used to build my IB15s... I still have four to check, but I'm not in a big hurry after testing the initial four.
|
|
|
Post by ThomasW on Nov 2, 2005 16:08:20 GMT -7
I know of no one else who's measured theirs, so no basis for comparison.
But WOW, you're being charitable, they're certainly more than "a bit" off, if the woofer tester is accurate.
If you have the time to hand test for comparision that would be interesting.
Other than that, it seems to be time for a phone call to John..
|
|
|
Post by shomrighausen on Nov 4, 2005 9:38:50 GMT -7
I have questions that I'm having a hard time wrapping my mind around. With these 'out of spec' drivers, is it possible to have a good sounding IB? The 'Qts' average for the drivers is ~1.0, the 'fs' doesn't bother me, they can play plenty below 'fs'. a lower VAS is to my advantage, though in my situation, I doubt that it matters much. 'Le' IS high, but it's a subwoofer, so I don't care too much how fast it rolls off on the top end. Having a lower Mms is a good thing, less mass to move = less mass to stop. Low 'BL' bothers me a little, the cone will be moved with less authority. I have a BFD to flatten the response. I have a Nady XA-2100 (51.6lbs of goodness) - plenty of power (over 400w per driver). I plan on a manifold with four drivers in my basement HT ~2500 ft^3. The back of the manifold will be open to approximately 450 ft^3. With a set of 'in spec' drivers this is ~ 7x VAS; with the my 'out of spec' drivers, this is ~17x VAS - I don't see this affecting the sound. My brain isn't letting me understand how the Qts of the drivers will affect the overall sound of the IB. I know that without a BFD the driver Q roughly equals the system Q. With a BFD to flatten the response, does it matter? Flat is flat? I realize that the rolloff of the system will be faster with high Q drivers/systems, but if it's flat down to 15-20Hz how much does that matter? I would LOVE to have a system flat down to 10Hz or so, but without equalization, it's not possible, regardless of driver/system Q; and since the BFD can only equalize down to 20Hz, I'll 'settle' for that. 'BL' concerns me - but with lots of power how much SHOULD it bother me. I have another four of these drivers - I'm going to use these in two other IBs, a pair in my family room - for music, and a pair in my garage/shop - for music. My family room and garage are near enough to each other that I'm going to get another BFD and use one channel for each IB. Each powered with 400 watts of older NAD power (1200w peak power!!!) THREE IBs in my house when I'm all done... Someone with less on their mind (or more IN their mind) help me out please. I've had a hard time getting a hold of John since receiving the drivers. I'm going to try to get the drivers measured 'by hand' this weekend if all goes well to verify the specs that I measured. I'm don't think I'll have time to build the IB and get it installed this weekend to listen, and the theory behind my situation has my brain working overtime. Thanks to all who can help.
|
|
|
Post by ThomasW on Nov 4, 2005 10:00:39 GMT -7
I wondered about this situation as well.
The "Q" being high means underdamped operation. As you said I don't think Le is a problem since they're just subs. The higher Fs means a dab more EQ might be needed for the low end.
I don't think the other specs are problematic given you have 8 drivers.
I guess the only way to tell is to setup one manifold and try ti out. First I'd contact John and see what he says.
|
|
|
Post by formica on Nov 4, 2005 10:06:14 GMT -7
Other than that, it seems to be time for a phone call to John.. Please keep us up to dateā¦ I different batch of lighter cones (which would explain the lower Mms and therefore higher Fs)? I wish that I could help, given I have a batch of IB15s from the original production, but I'm not equipped to do T/S measurements. Rob
|
|
|
Post by rbuszka on Nov 5, 2005 6:29:00 GMT -7
I would say just contact John. He's been known to go far, far out of his way to make a wrong situation right. Just give him time - he's moving stuff out of his shop and back to his (decently big) home this week. So if he is hard to reach this week, that is probably why. Give him a little time and I figure he will get right on your issues. For the time being, probably box up all of those woofers for shipment back to AE Speakers for testing and possible rebuild.
|
|
|
Post by aespeakers on Nov 5, 2005 7:27:33 GMT -7
I see a few things that cannot be right. First off, Mms cannot be measured correctly. The physical mass of the cone should be about 100grams, plus the coil around 65, plus half surround, spider, glue, dustcap, plus the loading. Mmd should be a little less than Mms because it is the physical weight only.
Also, Vas is much stiffer than it should be. The spiders do have a very stiff ring of resin on them towards the middle, but that shouldn't affect things that much. My initial guess is that the method(delta mass) to measure Vas and Cms was a mass added method. This is often not very reliable. If mass moves at all it will greatly throw off the measurements. Both with the woofer tester and with praxis I used to get erratic results. A much more reliable method is the delta volume where you take measurements in different known volumes.
Although everything looks off by a lot, putting the Mms in the approximate range it should be in shifts parameters back to where it should be. Cms being about .44 instead of the .19 that it measured at shifts everything else dead on. Both of those parameters would be wrong together as I said if the mass added method wasn't right. The mass as measured cannot be correct because the parts alone weigh much more than that.
In regards to the Le measurement, this is about the most difficult thing to measure. Le is different at all frequencies. It depends where you take the measurements. Madisound specs Le at 1Khz and 10Khz. Neither of them is at all useful for a subwoofer really. If you take a meter and actually measure Le across the terminals you should measure pretty close to 1.25mH.
John
|
|
|
Post by ThomasW on Nov 5, 2005 9:13:36 GMT -7
Hi John, and thanks for dropping by to clear things up....
|
|
|
Post by shomrighausen on Nov 5, 2005 12:46:00 GMT -7
The Qes (and therefore Qts), and Le are measured without dealing with added mass. I've measured the Le with a meter (Ellenco LCR-1801), and using 'measure inductor' with the woofer tester - which measures at 1kHz. Both show Le of almost 3.0 on all four drivers that I'm measuring. I've now broken the drivers in with various test tones - 10Hz, 15Hz, 20Hz, 25HZ, repeated 5 minute sweep from 10Hz-20Hz, 15Hz-30Hz, 20Hz-40Hz. Total breakin time on the drivers is around 30 hours, about half of them using most of the Nady XA-2100 (~2000w total to all drivers), about half using most of an NAD 2600A (400w total to all drivers). I'd say that the drivers are well broken-in. I then retested the drivers, and removed the rubber boot that is around the driver (near the bolt hols), to eliminate any dampening that this would have caused. I also used 185g of added mass instead of 125g (it is NOT moving during the test). I have each driver clamped two feet above the concrete floor in my basement using three Bessey K-Body clamps. My clamping system weighs 3-4x the driver weight. These are the most recent averages for the four drivers that I'm testing - published specs in ( ): fs (16) 30.70 Qms (4.9) 7.84 Qes (.48) 1.3228 Qts (.44) 1.1319 VAS (467) 178.79 Re (5.6) 5.49 Le (1.25) 2.937 Mms (218) 154.98 BL (16) 11.135 Even if there is a problem with my Delta Mass method for finding "VAS", it doesn't change the fact that, electrically, these drivers are far from spec. I'm in the process of building my manifold and I'll put them in and see how they sound... I'm not willing to pay $75-80 shipping to get these back to AE Speakers for being so far out of spec. It's fairly disappointing after spending $800 for eight of these that they are so far out of spec. I suppose I should have ordered 4 and gone from there, at least I'd only be out $400.
|
|
|
Post by aespeakers on Nov 5, 2005 13:06:48 GMT -7
Well, something is still obviously wrong that we need to track down. You can't just measure the electrical parameters without the mechanical ones. Qes is dependent on Qms which is dependent on Rms which is dependent on the Mms and Cms. If your compliance and mass are off it throws off everything. In praxis if you measure at first without going through the mass added measurements it will give you a partial parameter set that is way off. As you do the second measurements with the added mass it goes back and fixes everything.
Your Mms went from 132g to 154g now. That alone is enough to show that something isn't correct in the measurements. Breaking in the driver is not going to add 22grams to the Mms. In any case, the cone and coil alone weigh about 165grams. Add in 8-10grams for the dustcap, half surround and spider, plus glue for the dustcap, glue from spider to coil and coil to cone and you should get an Mmd of right around 200grams. That is a physical weight. It can't be changed. Then the Mms is slightly higher due to the loading on the driver.
Your Qms changed from 4.9 on your initial measurements to 7.84. Again, that is not going to change that drastically based on just breaking in the drivers.
Bl are inversely proportional. If they were independent of the mechanical parameters, As Qes goes up from 1.27 to the 1.32 you measured, your Bl should go down, but it went up about 4%.
How much mass are you adding when you do these measurements? What are you using for mass? How are you attaching it to the cone? where on the cone is it attached?
John
|
|
|
Post by ThomasW on Nov 5, 2005 13:23:58 GMT -7
shomrighausen,
I realize you've used this testing device for many many drivers, but the original Woofer Tester was taken off the market due to problems with it's accuracy.
So before you start returning drivers, it would be interesting to see a comparison between hand measured units, and the Woofer Tester results.
Doing that has to be less time consuming than packing up and returning the drivers.
|
|
|
Post by gonefishin on Nov 5, 2005 20:55:38 GMT -7
Hi guys... I just wanted to thank the both of you...well, ok. The three of you... shomrighausen, John (AES) and ThomasW for dealing with this in an adult manner. Too many times you see people on internet discussion forums get defensive or start with insults. There does seem to be a discrepancy in the measurements that shomrighausen made on the AES woofers. It's nice to see that all parties involved are simply trying to find were or how the measurements were taken. Then, once that is established, I would imagine you then...go from there. But thanks for giving this topic a chance to get to some conclusion. Mean while...yes, I will be watching closely to not only the determination of the spec results...but also the opinions on the sound. I've been contemplating getting four IB15's for my system. (thread here --> ibsubwoofers.proboards51.com/index.cgi?board=projects&action=display&thread=1128991543) Due to funds dwindling because of unexpected tube faliures and the fact that I spent more money than I thought I would on some room treatments...I'll have to wait a bit before ordering (while funds replenish ) John, would it be any trouble for you to perform some measurements on a couple of production woofers. I know that there may be some variation...but it would be interesting to see what you come up with? (hope I'm not being out of line with you...I'd just like to see this come to a conclusion) again...thanks a bunch guys! dan
|
|
|
Post by shomrighausen on Nov 6, 2005 14:33:34 GMT -7
I decided to spend my time and effort getting my HT IB up and running instead of measuring the drivers by hand. I ordered a Woofer Tester 2 from PE, it should be here mid-week. I will run a full set of tests on the remaining 4 drivers (yet to be unboxed) on the original WT, the WT2, Speaker Workshop, and by hand (maybe Praxis if I get the time). I'll post the results from all of the tests.
I too would be interested in having John test a few of the current production woofers, and share the results and method for testing.
As a side note, whatever the actual specs of the four drivers that I've been working with - the IB is in and sounds GREAT!!! It only took a little EQ from the BFD, and it's within a few dB below 16Hz on up. If the specs that I measured are accurate, it seems that they have little effect on the performance in an IB setup. If they are wrong, I'll donate my Woofer Tester to a friend (it seems to be accurate when measuring smaller drivers).
Happy listening... I'll keep you posted...
|
|
|
Post by shomrighausen on Nov 9, 2005 12:44:37 GMT -7
The WT2 showed up today, I had a chance to measure one of the remaining woofers (break in 10Hz-30Hz 10minute sweep for 15 hours). The new tester is much nicer than the old. They give similar results from what I've seen. I'll do a full suite of test with multiple test methods on the remaining four drivers as soon as I can. Here are the results using the WT2... I'd say that there is something wrong with these woofers...
fs (16) 31.3845 Qms (4.9) 4.6271 Qes (.48) 1.2342 Qts (.44) 0.9743 VAS (467) 210.6671 Re (5.6) 5.4927 Le (1.25) 2.1418 Mms (218) 128.4214 BL (16) 10.6311
I still hope that John builds a few IB15s using the current parts (same as the ones that he used when he built my eight drivers), and has an opportunity to test them. I'll post again after I do some more testing.
|
|
|
Post by aespeakers on Nov 9, 2005 22:36:39 GMT -7
I weighed up some parts today just to make sure things were correct on my end. Here are the weights:
Cone plus half surround 133 grams VC, 45 grams. Dustcap about 9 grams Spider is about 10grams so half is about 5grams Figure 10-15 grams of glue between the good bead around spider and cone and then the dustcap.
So physically we have 202grams. That is a constant that is the physical weight that comes up with Mmd. Mms is then that weight plus the resistive loading which usually adds another 2-4% if you go by what most companies calculate. So physically it is right in the 210gram range which is pretty close to the 218gram spec.
Steve, how much mass are you adding when you do the mass added method? What are you using for mass, and how are you attaching it to the cone?
John
|
|
|
Post by shomrighausen on Nov 10, 2005 7:05:08 GMT -7
I am using nickels for weight. With the WT2 it does a good job of telling you what the ideal weight to add will be for the driver. If I'm more than 5 grams off one way or the other, I add or remove nickels to get within 5 grams. It has varied between 15 and 17 nickels (75 to 85 grams). The cone is 'face up', and I'm adding the nickels by laying them around the dustcap, flat on the cone. With the original WT I was using more mass (125-185 grams) based on the published specs.
I am sure that my test method is valid. I am also sure that the drivers are out of spec. There are a fair amount of parameters that are figured electrically, VAS, Mms and some others are measured using Delta Mass. IF I am having a problem with my method of Delta Mass (which I don't feel it the case) - it still doesn't explain the Qts, etc.
The fact that both WT and WT2 have similar results - makes me believe that these drivers are far from spec.
Another reason that I think that the measured specs are valid - I had to use two broad filters on my BFD to cut a peak for my IB in the range that makes sense for having a high Qts set of drivers. Coupled with the rapid drop off of the low end, which is in line with simulations using the measured specs.
I still plan on measuring the drivers using alternate means, but my gut tells me that I'll measure similar results regardless of test method. I'm also going to build a sealed box for measuring the VAS, etc.
John - what method do you use to test your drivers? I would like to use the same method that you use to try to duplicate the published specs of these drivers. Also, will you build a driver using the same batch of parts that were used to build my drivers and test the driver? I'm not a driver manucacturer, and I'd like to find a resolution to the difference in specs (published vs. measured).
|
|
|
Post by shomrighausen on Nov 12, 2005 8:40:35 GMT -7
Well... it's official... these drivers are out of spec... all of the measurements are close between the original Woofer Tester, the Woofer Tester 2(v1.01 and v1.03 software), Speaker Workshop using the Wallin Jig II, and Praxis v2.32a. I've measured the VAS using Delta Mass (nickels), and Delta Compliance (5 ft^3 box).
I've also posted my measurements on the Parts Express tech forum, and they agree that these drivers are out of spec. Darren K. from Parts Express says this - "Yes, this is indeed a strange problem. It is important to note that the Fs and the Q specs will not be affecting by changing the Vas test method, so that will probably not make any difference.
As far as what could be wrong with the drivers? 1. wrong spider. If the spider has a compliance that is way lower than what it is supposed to be, it will affect the Vas and the Fs of the driver inversely proportionally. (which looks quite plausible considering the Fs is way too high and the Vas is way too low). 2. magnets aren't charged all the way. If the magnets in the motor aren't getting fully magnetized, he won't have enough flux, and the BL will be low and the Qes and Qts will be high."
As far as I'm concerned this is a closed case. John from AE Speakers hasn't been able to provide me his test method, and I've used several valid test methods and receive similar results (none of which are even CLOSE to published specs).
The simulations that I run show why I'm not satisfied with the difference between the drivers that I have and drivers with published specs:
at 25Hz they are both down approximately 3.8dB at 20Hz my drivers are down 2.5dB more than spec at 16Hz my drivers are down 4.9dB more than spec at 12.5Hz my drivers are down 7.0dB more than spec at 10Hz my drivers are down 8.5dB more than spec at 5Hz my drivers are down 11.5dB more than spec.
One of the advantages that an IB has over a conventional boxed sub is low end extension - which I don't have with these drivers. I think Darren K. from Parts Express hit it on the head. I wait to hear from John on what to do next.
|
|
|
Post by aespeakers on Nov 12, 2005 17:10:22 GMT -7
Hi Steve,
If you aren't happy with the drivers I can build you 8 more drivers fully from hand. These motors are ones that were built in China, but have had no problems up to now. I have a run of steel parts that needs to get to the plate shop next week anyway. If you want, use these for a week or two and I can machine new steel for you and build 8 new drivers to get to you. I'll measure specs before I ship them out and include a printout with each driver. That way I will go from the ground up and make sure everything is where it is supposed to be.
The thing that confuses me is that your Mms is still not coming out anywhere near correct. I had some cone/spider/surround/ coil assemblies that were cut out from rebuilds on some other drivers. This was a physical piece weighed on a small postage scale of cone, plus half spider, plus half surround, plus VC. I weighed 3. One was 193 grams, one was 195 grams and the other had almost no surround left at all which was 189grams. There was no dustcap and dustcap glue which will add another 10-15 grams. Add that in and you are very close to spec.
I also weighed up cone/surround assemblies from 2 different runs of cones and several different boxes. All were were within 3 grams of each other. This is from 117-120grams with the full surround.
The spiders used have a very heavy resin coating on the inside corrugations. This takes a long time to break in and get to the specified compliance. Nick at Lambda Acoustics had many of the same issues with people feeling drivers were of spec, but over time they soften up dramatically. I have done over 250 IB15's now using them and this is the first complaint I had on the actual drivers.
In terms of magnets not being fully charged, that would give various other things that are way off. Your Q's would be all over the place and you wouldn't be able to get the same results or even close twice in a row.
As I have said originally if you soften the spider which gets compliance where it should be and then figure the proper amount of Mms, the drivers will model nearly right on spec. The only other thing I can think of is that I always put my woofer up on a couple pieces of mdf to make sure the pole vent isn't blocked. I've had various different results that have been off by having the pole vent blocked.
In any case, I'll get going on machining new steel parts on monday and should be able to get them to the plate shop Wed. I will have them back within a week and can get the assembly going. I hope this is fair to you.
John
|
|