Scan-Speak sub kit? - Page 2

Originally Posted by Phast1

Since I'm the OP, thought I should chime in...

Lots of good info/debate here, thanks. I think I will need to research much more and be sure of what I really want/need.

I am also considering buying a decent used sub here locally (10" with 500w plate amp) and dissecting it. Maybe try to model it in winisd for a different enclosure, port etc...
This might be a good way to learn.

Originally Posted by exojam

Rick,

Below is a near field of one of my AV15H's in a cabinet that are 19.5x19.5x21.25 (no smoothing). The second graph below is the same measurement as the first, just moved over so the lower range can be seen.





These graphs are both of the subs combined measured at two different locations. No EQ added or smoothing on the graphs.

First location:


Second location, moved only about a foot back from the first.


James

Originally Posted by exojam

Rick,

Below is a near field of one of my AV15H's in a cabinet that are 19.5x19.5x21.25 (no smoothing). The second graph below is the same measurement as the first, just moved over so the lower range can be seen.





These graphs are both of the subs combined measured at two different locations. No EQ added or smoothing on the graphs.

First location:


Second location, moved only about a foot back from the first.


James

Originally Posted by spanish68

djarchow, all those different brand name subs that you have, do you run them all together? No integration issues? Also, you mention the F15 sub as one of the best sounding, better than the AV15 sub? The maelstrom is probably the spl leader, right?

Originally Posted by A9X-308

Can't give you Klippel, but what's provided is more than the Seas datasheet provides. Do you have Klippel or real nerfields on the 23W's?
Not sure what nearfield will confirm as I have never seen a completed driver + box that did not come very close to it's modelled response. Measurement technique will give larger errors.

The average 10" is about 350cm2 and 15" 830cm2. 2 x 350 is not greater than 830.

Not sure what you mean by 'real bass'. Can you be more specific?

Sure, a 10 in a suitable box can have a similar f3 and f10 to a 15 or an 18 for that matter. It's just that it won't go as loud for a given F because it can't move as much air given the same Xmax, and the Xmax on a smaller driver is typically lower, though not in every case.

Originally Posted by Rick Craig

How far was your microphone from the sub? Can you describe your measurement equipment and software?

Originally Posted by A9X-308

Can't give you Klippel, but what's provided is more than the Seas datasheet provides. Do you have Klippel or real nerfields on the 23W's?
Not sure what nearfield will confirm as I have never seen a completed driver + box that did not come very close to it's modelled response. Measurement technique will give larger errors.

The average 10" is about 350cm2 and 15" 830cm2. 2 x 350 is not greater than 830.

Not sure what you mean by 'real bass'. Can you be more specific?

Sure, a 10 in a suitable box can have a similar f3 and f10 to a 15 or an 18 for that matter. It's just that it won't go as loud for a given F because it can't move as much air given the same Xmax, and the Xmax on a smaller driver is typically lower, though not in every case.

Originally Posted by zero the hero

You might also want to check out the Seas L26ROY driver.

Excellent sound quality from all reviews on it, near SS quality at decent price level. Works well in smallish box tuned to 30hz (low enough for music)

Originally Posted by exojam

Rick,

Subs are sealed. Below is a picture of how far away the SPL meter was from the driver (maybe an inch or two). I also have the steps I took to level match them before placing them in the best spots; this was done before the near field measurement. I believe the net internal volume is around 3.2 or 3.5 cubic feet (sorry I do not have the exact volume).

Using a Onkyo SC-885, EP2500 amp, little laptop, REW and Radio Shack SPL meter

Main set up process:
1) Set my master volume control on my processor to -20
2) Set the trim in the processor to its lowest level -15
3) Put one sub in the middle of the room and turned up the gain knob on the EP2500 till it read right at 77-78db
4) Moved the second sub into the same place and did the same steps.
5) I then moved them to the general place where they would go until I got the best reading and then left them there
6) With both subs running I then adjusted the trim level in the processor till I got to 75db.

Picture of one of the subs:


Picture showing SPL meter:


James

Originally Posted by Phast1

Since I'm the OP, thought I should chime in...

Lots of good info/debate here, thanks. I think I will need to research much more and be sure of what I really want/need.

I am also considering buying a decent used sub here locally (10" with 500w plate amp) and dissecting it. Maybe try to model it in winisd for a different enclosure, port etc...
This might be a good way to learn.

Originally Posted by Rick Craig

Actually one of the AE woofers is coming to me but I don't know when it will arrive. I've worked with John in the past and think he means well but the delivery problems have discouraged me from incorporating the woofers into my designs.

I never trust any Salk claims for bass extension because there are no nearfield or ground plane measurements to back them up. Having used some of the same drivers I can tell when the numbers look "optimistic".

Originally Posted by penngray

There is zero reason to consider a premium scan speak driver for Subwoofer design (IMO).

Originally Posted by Phast1

This would be a primarily music sub, preferably sealed - no passive radiators.
If anyone could point me to some ready-made plans for this driver so I could check it out, that'd be great.

Originally Posted by Rick Craig

I've used the Scan-Speak 23W series drivers with very good results. For a small sealed box they have excellent bass extension and the low inductance motor gives you good upper end bandwidth. Yes, they are expensive but I know of no other sub that can do all of the above.

Originally Posted by Rick Craig

Yes I've seen those before. They verify the low motor inductance and that's good. But I would like to see nearfield measurements in real enclosures to see the actual extension in the lower octaves. Klippel tests would also show how well the motor and suspension perform.

Originally Posted by Rick Craig

Actually one of the AE woofers is coming to me but I don't know when it will arrive. I've worked with John in the past and think he means well but the delivery problems have discouraged me from incorporating the woofers into my designs.

I never trust any Salk claims for bass extension because there are no nearfield or ground plane measurements to back them up. Having used some of the same drivers I can tell when the numbers look "optimistic".

Originally Posted by zero the hero

You guys are jumping all over Rick (ok, not John so much) for little reason IMO. I've seen plenty of unprofessionalism around here, and this doesn't really qualify. Sure, we can all trust modelling (and we pretty much do), but let's face it, there are drivers out there where modelling is only accurate for very low power input (and I haven't seen too many 1 watt plate amps). A little skepticism is healthy, and probably necessary in many cases.

Originally Posted by Jeff Bagby

I didn't feel I "jumped all over Rick". He made a comment about my design work and I tried to clarify. The comment would be different coming from one of you, but Rick makes speakers and sells them too, so I felt I should comment. I think he would probably agree. I only tried to give the background and state the validity of what we are claiming. Rick and I know each other and have even had lunch a couple of times. He is a great guy, and I hope there are no hard feelings from anyone. I wish him and Selah much success. The more successful he is the better it is for all boutique companies.

Originally Posted by Jeff Bagby

Hello Rick. I have never posted here before, but you know me well by now. This thread was brought to my attention, and since I designed the bass alignments for the larger Salk speakers I thought I would comment and clarify.

A little history - when Jim and Dennis Murphy were first working on the HT3 Jim contacted me for woofer suggestions. They had originally looked at common commercial drivers and were leaning toward a Peerless driver. I suggested a custom woofer from TC Sounds. I already had a contact there, and connected Jim with them. Together we designed a 10" driver that met all of our requirements. We didn't need high sensitivity since it was being mated with the Seas Excel woofers, so we focused our parameter choices to trade off sensitivity for bass extension. When we were finished we had a 10" TC driver that was custom designed for Jim's application, and was not like hardly any other driver available on the market. This driver wasn't available to anyone else either, but anyone could have had Thilo make something like it if they bought in large runs like Jim did. This was the key. You can not buy drivers like these as one-off drivers.

This went well for a while but when TC Sounds ran into financial trouble and ultimately closed their doors, John at Acoustic Elegance agreed to duplicate the woofer for Salksound. Using identical motor parts in its construction, John made a woofer that matched our parameters and were confirmed with measurements. Incidentally, John's measurements and mine matched almost exactly. I can not say the same for most off the shelf drivers out there, even ones from SEAS and Scanspeak. These often measure quite a bit higher in Fs and Qes then advertised, but I digress.....

Between Dennis Murphy and myself response measurements have been taken on all of the Salk line of speakers, up to and including close-mic'ed nearfield measurements. These confirm that Thiele and Small were correct in their ability to predict the low end response. Seriously now, that's why we all use box modeling software - because it is well defined and accepted mathematics, and has proven itself time and time again. And having written a box model program of my own based on Theile, Small, and J.E. Benson's work I understand the math quite well. We do not post these as "proof", because we don't feel every measurement of a speaker needs to be posted and used as marketing gimmicks. Further, there is nothing to keep any company from posting anything they want, and none of that really proves anything. Jim's speakers are sold on the basis of how they sound and from the satisfaction of owners sharing their glee with others. But, suffice it to say, Jim is quite conservative in his claims. He is adamant that nothing be exaggerated in any claims on his speakers and Dennis, Paul Kittinger, and I back up all of our claims to him with proof beforehand. For example, with my Pharos design, even though we are using a 12" JBL woofer we state that is begins to roll-off at 44Hz and is -3dB at 40Hz. We are not exaggerating anything.

Moving on - the transmission line speakers were all designed for low end extension by Paul Kittinger using Martin King's Math Cad sheets. These are widely accepted as the most accurate transmission line modeling program extant, and our measurements of them confirm Martin's predicted extension as well.

Now, finally, to the new as-yet-to-be-name speaker, currently called the HT4. We are claiming a bass extension of 18Hz. I apologize if this ruffles feathers, but I stand by this claim, and put my reputation on it (whatever that may be worth). Jim called me and asked if it were possible to achieve this cut-off in a reasonably sized three-way speaker. I told him I would see what I could do. At this time a SEAS L26ROY driver sent for us to test. However, it just did not deliver the bass extension, or possessed the excursion we were looking for. Based on that, I decided that we needed to go in another direction. Given the fact that John J. at Acoustic Elegance could, once again, make us the woofer we needed based on our parameter selection, we arrived at a custom 12" driver for the job.

This speaker is not tiny; the bass bin is approximately 3 cubic feet+. It uses two 12" passive radiators each mass loaded with over one kilogram of weight and the Fb of the enclosure is tuned to 20.7Hz, as measured. Passive radiators were chosen because it allowed for us to use the equivalent of a vented alignment without having to deal with excessive port noise and restrictions. The 30mm Xmax capability of the two radiators can easily handle this job when combined with the high excursion woofer we are using. The predicted F3 for the speaker based on J.E. Benson's passive radiator model is 18Hz. Based on close-mic'ed measurements of the cone and the PR's in a complex summation we can confirm this cut-off. You are welcome to visit Dennis Murphy and listen for yourself if you like, he has one in his home. How's that for a friendly invite?

And finally, finally, John at Acoustic Elegance, in my opinion produces some of the finest drivers available in the world. His Lambda TD12H is the best 12" true woofer I have ever measured or worked with. The low distortion motors with the copper shorting sleeves are the real deal, and these measure with a very low Le for such large, high excursion drivers. His ability to customize drivers for us, if bought in quantity runs, adds to the advantage of using Acoustic Elegance. John has been excellent to work with, and his products are truly top notch.

Now, I do not believe anyone associated with Acoustic Elegance or Salksound has said anything negative or disparaging about Selah Audio or any of your designs, nor do we intend to. I am sure your designs are excellent as well. With that in mind, when you make a statement like you did about not believing Salk's claims you called into question the integrity of Jim Salk, Dennis Murphy, Paul Kittinger, John Janowitz, Martin King, and Jeff Bagby. So I decided to speak up and ask you to reconsider making such off-the-cuff comments. As a professional speaking about another professional's products you really should know better about being professional. Agreed?

Many blessings to you in 2010 with Selah,
Jeff Bagby

Originally Posted by John_E_Janowitz

For a subwoofer design, it's all about moving air, and doing so cleanly. The biggest part in creating bass is displacement. The 23W scan speak alum cone drivers are quite expensive. You get .6L peak-peak displacement for $385 retail. The AV15 provides you 3.6L peak-peak for $249. You'd need 6 of the 23W to equal the low end output of a single AV15. The AV15 is also over 10dB more efficient than one of the 23W scan speaks. So to get equal levels you'd need 10X the power to the 23W to get the same level of the AV15. Obviously this is a 10" vs 15" and not an apples to apples comparison, but it verifies the point that the scan speak isn't a typically good choice for a subwoofer.



I guess there are a few things to look at here. For music you'll typically want flat response to 30hz or so. The low B on the bass guitar is 30.87hz and low A on the piano is around 28hz. There is also a lot of electronic content on keyboards, etc that can be in this region or lower. Obviously you can go sealed and a sealed AV15 will give you much more output than many of the scan speaks.

However, what people often think is that sealed boxes will be more accurate and sound better than sealed or vented boxes. This is not necessarily the case. Sealed boxes take multiple times more power and woofers need to move to much higher excursions for the same output levels as vented or PR cabinets. You do have the "damping" that comes from the sealed box, but on the other hand you have much higher distortion for any given SPL due to thermal issues and non-linearities at higher excursions. In most cases for subwoofer duty, a vented/PR system will provide more cleaner output than a sealed.

You may wish to design the box sealed but large enough to accommodate PR's on the sides if you decide you need more output later. You can get 6-9dB more output at and around tuning with the same input power.



I happen to know of quite a few. The low inductance and shorting rings used to lower distortion are all good things. We do the same, although in a more effective way. Scan speak and Roger Russell had the original patents on 2 variations of shorting rings. One was a cap at the top of the pole. The other a ring at the bottom of the pole. Lambda put a sleeve on the entire pole which is more effective than either because the coil is adjacent to this large sleeve of copper at all points in excursion. As a result Le has nearly no variance with excursion. That is not the case for scan speak. You can see augerpro's testing on a few woofers that verifies this fact for the Lambda motor. The other difference between the scan speak and our AV woofers is that we also have high excursion and can displace large amounts of air as well.

We do have an AV10, which is not the same as the Jim Salk woofer. Inductance is about half that of the Scan Speak, efficiency is about 5dB higher, Xmax is about 2x as long (23mm) and power handling about 4x as high as the scan speak. Oh, and it is fully hand made in the US with mostly US components and it's about half the cost too. I guess you are right. There really is no comparison.



I'm not sure what nearfield measurements will really tell you that modeling won't. As pointed out, the T/S parameters give quite a good indication of what a woofer will do in the enclosure. We go through great lengths to make sure every woofer meets our specifications quite closely. Tolerances are about 1/2 that of what the industry standard is. I recently sent samples of the AV woofers to someone over in Finland that many of you may remember quite well. The klippel results he tested verified that the drivers were quite linear but I unfortunately can't publish this data.

Ground plane measurements at increasing volume levels would verify the linearity of parameters. As Bl drops, the Q of your system increase and as a result the rolloff will get steeper.



We have had a large share of supplier issues over the years, but things have mostly stabilized now. We do build every driver as it is ordered. If all parts are in stock we can deliver within 2 weeks mostly routinely. The difficulty is in keeping parts in stock. We aren't in a position to carry huge parts inventory but we try to keep realistic reorder points. However we can have a reorder point of 50 cones for a certain woofer and an OEM customer calls to take 50 and they're all gone. Then word gets out they are using our driver and 20 people all want to clone that system. At that point our lead time is based upon when our vendor gets us the parts. We try to be realistic and base our lead times not only on the times we are quoted but the experiences we've had with that vendor in the past. That's all we can really do.

The solution is to continue to find ways to do more and more things in house. We recently found out our santoprene surround vendor will no longer be making surrounds. The solution is to build a thermoforming machine to make them here. We couldn't get satisfactory aluminum cones from china. They are spun in the US and we sand and finish every cone in house. We couldn't get spiders made with lead wires attached, so we sew every one in house. Steel parts didn't come in with correct tolerances so we hand machine and verify each t-yoke and top plate within .001" tolerances. In the future we'll be moving even more production in house until only the raw materials are purchased elsewhere. Then we only need to keep track of raw materials and individual components can be made much quicker as needed.

John