How do _you_ choose drivers? (1 Viewer)

[Brian J Dupuis]

Hello,
First-time poster with hopefully a reasonably non-idiotic question. How do you choose drivers, from the seemingly endless array of available ones, for your projects? Do you go by respected manufacturer and choose from their offerings? Do you rely on peer reviews of particular drivers (as I see a lot of excitement surrounding the Shivas and Tempests around here for subwoofers)? Do you utilize your experience with T/S parameters and peruse countless lists of drivers to find the ones that match your selected criteria?
I am soon to start my first DIY project (a subwoofer... seems easiest to get my feet wet and is my current need in my system) and am really just curious to know what experienced DIYers use as a 'filtering mechanism' amongst what, to a rather inexperienced newbie, looks like a bewildering quantity of (somewhat) similar offerings.
Thanks in advance... I'll try to be less annoying than some rather demanding newbies I've seen around .
P.S. 'subwoofer' is not recognized by the spell checker for the Home Theater Forum. Struck me as ironic

.

 

[Jack Gilvey]

With experience, you do get a general feel for what a driver can do by looking at the T/S parameters. There is a vast array of drivers on the market, but this forum is very good for bringing those that are worth attention to the top.
I don't have a lot of experience with multi-way speakers and crossovers, as I prefer the full-range route, but my personal "filtering mechanism" for sub drivers (which seems to be your interest) would be a low Fs (~20Hz), low Qts (~0.4) and reasonable Vas . There's an increasing number of really good drivers entering the market, the likes of which would have been extraordinarily expensive a few years ago. Fairly low-priced offering from Adire, Stryke, PE etc. are good examples of what the modern enthusiast has to work with.
quote: Thanks in advance... I'll try to be less annoying than some rather demanding newbies I've seen around [/quote]
You're already OK in my book, then, and rather perceptive.

Some seem much more ready to "teach" than to learn.
What type of project do you have in mind, as far as budget, size, performance goals?
[Edited last by Jack Gilvey on July 12, 2001 at 09:07 AM]

 

[ThomasW]

Brian
Welcome to HTF
Chosing drivers is a wierd set of compromises. Any driver must have the desired performance characteristics (yes this is primarily the T/S parameters). Then it must be in the desired price range. If two drivers have the same or very similar performance characteristics; and are similar in price, then I chose by mfgr.
One reason you see so many references to the Adire products is that they are an extremely good value for the cost, the performance is very good. Also the boys at Adire provide some of the best technical support/customer service in the business. They will literally help the customer custom design a personalized enclosure. In addition their drivers are a "known" item, they perform as per the published specs. In this business it's not unusual to receive a driver that tests out different than the published T/S parameters. This isn't the case with Adire products
Hope this helps
ThomasW

 

[Brian Bunge]

Jack,
That's how I decided to go with ACI's SV12. It has an Fs of 17.5Hz, Qts of .395 and Vas of 250 liters. These specs are somewhat similar to the Shiva (except the Shiva's Vas is 151 L), which is highly respected around here. I went with ACI long before I had heard of Adire but they are both very respectable companies.
I've stuck with ACI for most of my DIY stuff since building my first SV12 sub. In this respect I guess I've gone more on company loyalty than anything else. ACI hasn't let me down yet!
Nevertheless, I've been very tempted to take the Tempest plunge! But, I think my next sub project will be something along the lines of an 18" driver so I may take a look at what Adire will be offering in the future as opposed to ACI's SV18.
Brian
------------------

 

[Dustin B]

Another advantage to the Adire drivers is Adire provides a version of LspCAD that only works with their drivers. LspCAD is a very expensive by accurate program, and access to it allows you to do a much better job modelling your design. Plug in the Tempest or Shiva specs into another program and come up with a design, then see how the numbers differ when you do it in LspCAD. When you don't want to spend the money on software to get your design modelled as close to the real thing as possible this adds even more value to the Adire offerings.

 

[Brian J Dupuis]

Thanks for the responses guys.

What type of project do you have in mind, as far as budget, size, performance goals?

I'm lucky and unlucky for this project. The primary limiting factor for my sub project will be size (WAF in full effect). If I can stay under 2 cubic feet internal volume (preferably less), I think I might gain the required authorizations . Two designs I've been considering are:
1) An isobaric, sealed enclosure based on NHT 1259 drivers.
2) A PR design based on the Peerless XLS series of driver/PRs.
Havent's decided on which plate amp yet, though I'm mostly convinced that it will be a plate amp of some kind.
What do I want in the system? What everyone wants... deep extension, small size, excellent transients, high efficiency, and all for $19.99 delivered

. What I am realisticly shooting for is relatively deep extension (as close to 20 as I can get, but willing to sacrifice some here), excellent transients (less willing to sacrifice the transient ability... I like tight, controlled bass... not 'wobbly'), reasonable loudness (the wife already gets a little anxious with my modest system during movies, so foundation crackers would be overkill). The budget is relatively unconstrained, but not overly. Let's put it at $500 for drivers and amp.
I'm still doing my homework and trying to weigh options, but those are the two I have right now.
Again, I wasn't necessarily looking to have the gurus bless my choices... sure, you guys probably would do just that, but I'm almost certain you get tired of hearing the typical "I want to build a sub, should I choose the FlabbyBottom 12" in a 159.3 cubic foot enclosure or the 4" NoisyCricket in a 6th order bandpass spherical flomdidget?" . I was trying to get a more general feel for the thought processes used to sift through the mountain of choices... to me that's a more appropriate and efficient use for you guys' time.
But thanks for asking!

This looks to be an extremely valuable forum!

 

[ThomasW]

1) An isobaric, sealed enclosure based on NHT 1259 drivers.
2) A PR design based on the Peerless XLS series of driver/PRs

I'd suggest studying the T/S parameter of the Peerless XLS series very carefully.......
How about a isobaric dual Shiva in a 62L box?
Link Removed

 

[Brian J Dupuis]

How about a isobaric dual Shiva in a 62L box?

Very nice! Yours? I don't know if the exposed rear-end of the speaker would survive the WAF test, nor if I could squeeze a doubling of volume out of the bargain either . But it's an option. I love the woodwork on it... that's exactly along the lines of what I hope to do. If yours, do you have a construction page?
Thanks for the input!

 

[ThomasW]

Brian
You don't double the volume of the box for an isobaric design. It's 1/2 that for a single driver. So this box is functioning as if the box were 124L.
Yes it's mine, and yes there is a page, but it's not quite finished. I'll post a link tomorrow. Also there will be a stretch fabric grill cloth to cover the exposed driver.
The last time I checked the XLS series there was something goofy about the T/S parameters, I'd need to look at them again because I can't remember exactially what it was.
Also.... IMHO polycone drivers like the Titanic do not have a stiff enough cone. As a result you get cone breakup when trying to reproduce low bass. Drivers like the ones from Adire and the P.E. DVC clones of the Adire's have Kevlar epoxy bonded to paper. These are some of the stiffest cones I've seen in 40 yrs of speaker building.

 

[Jon Hancock]

Uh, I see the little plastic doodads for clipping on the grille assembly- how do you get it on over the outside woofer?
Seriously, a lot of good suggestions above for woofer selection. For selecting drivers for the rest of the frequency range, and choosing the system configuration, consider a few hopefully useful points:
1). Operate drivers *within* their pistonic range. This is particularly critical in upper midrange and lower highs. Cone breakup and cones designed for gradual decoupling (curvalinear) in my experience don't convey they ultimate resolution in sound quality. In mid/woofer and mid/range drivers, good examples come from Eton, Focal, and Accuton. Look for "wrinkles" in the impedance curve (published or measured)- these are due either to cone or cabinet resonances. Stay away from the frequency bands where they exist.
2) Select driver sizes and crossovers to achieve similar dispersion/directivity over the overal frequency range. Easier to say than to do. Look at off axis plots.
3) If you don't plan on buying a pre-designed kit by someone reputable, expect to spend a moderate amount of money on measurement equipment and softare. CLIO Lite is a good starting point if money is an issue. If that's too exensive, you may want to stay with well designed kits.
4) Room interaction is a critical issue- take it into account in the design if at all possible (woofer and mid woofer height above floor). Best of all, if you want to get a little out on the edge, consider a dipole bass design- it dramatically reduces room interaction in upper and midbass region. With more conventional systems, consider the use of a program like RPG Acoustics Room Optimizer for modeling room and listener locations, and optimizing LF respone considering both boundary mode and room mode interactions. At under $100, it's dirt cheap for the time and aggravation it can save you.
5) Above all, have fun! This is a hobby, not regligion! Sometimes those of us who tend towards the intense forget that a little....
Regards,
Jon

 

[Jeff Rosz]

hello brian,
if the wife wont go for that driver sticking out the side of the box, check out the design of the M&K subs atLink Removed a lil more eye pleasing. good luck and welcome to the forum.

 

[Brian J Dupuis]

if the wife wont go for that driver sticking out the side of the box, check out the design of the M&K subs atwww.mksound.com a lil more eye pleasing

Interesting... I was actually considering purchasing some M&K equipment (before the DIY bug hit), but never realized the particular technology/alignments used. And to be honest, it still doesn't make sense to my newbie self. Is that considered an isobaric design? I thought the volume of air between drivers was meant to be as small as possible, whereas it looks like those designs share the entire volume? I've been reading all the relevant texts (Cookbook, etc.), but I'm suffering from information overload . Thanks for the suggestion.

 

[ThomasW]

Brian
Here's a link to info on the isobaric sub
Link Removed
Just click on the "Rava Too!" link. This is just barebones info so as to get something quickly uploaded for people to look at. Much more detail including CLIO measurements will be posted when available.
Regarding the Peerless XLS drivers, the Qts for both is very low, and the Vas is relatively high. Plus the drivers are pretty darn expensive compared to the competition. Also I've never been a big fan of PR based subs. I've built and heard, what many consider to be the "better" of the PR designs, and find their transient response/bass quality to be lacking...
Yes the M&K subs are isobaric configurations but they are also resistively loaded designs. This probably isn't something a novice would want to try to duplicate
Regards
Thomas
[Edited last by ThomasW on July 13, 2001 at 06:58 AM]

 

[Jack Gilvey]

I thought the volume of air between drivers was meant to be as small as possible, whereas it looks like those designs share the entire volume?

As I understand it, the "clamshell" iso (like Thomas') is just one option. When so installed, the drivers are wired out-of-phase, so they're both going in the same direction.
In case this presents an aesthetic problem, as you indicate it might, you can also front mount one driver conventionally, and another driver firing into the cabinet (wired in-phase), like M&K. If mounted on the back of the box, this other driver can be hidden. Of course, a sub done like Thomas' can also be turned around, so as to present just a solid wood "table" to the viewer.
As you're probably aware, the only advantage of isobaric loading is one of reduced cabinet size, so I'd only recommend it if size is a very big issue. Having said that, such loading would allow you to effectively duplicate the performance of Adire's Rava sub in just over 1 cu.ft. stuffed. It would sound even better if you used the whole 2 cu.ft. you're "alloted", as that would be close to Thomas' lower Qtc alignment.

 

[Brian J Dupuis]

Having said that, such loading would allow you to effectively duplicate the performance of Adire's Rava sub in just over 1 cu.ft. stuffed. It would sound even better if you used the whole 2 cu.ft. you're "alloted", as that would be close to Thomas' lower Qtc alignment.

Hmm... I will have to model it, but you really think I could get some decent performance with half of Thomas' internal volume? From Thomas' sub construction page, it looks like his sub is already stuffed with long fiber wool. I think some modeling time is warranted... I am still very new to modeling and to T/S parameters overall. Therefore, things that may seem intuitive to you experienced folks are jaw-dropping revelations to me . I suppose I still see the Vb that people come up with, like Thomas' 62L as inviolable rules. Must do more homework.
Thanks again for the valuable input, guys.

 

[Jack Gilvey]

Hmm... I will have to model it, but you really think I could get some decent performance with half of Thomas' internal volume?

I believe so. (Thomas knows infintely more about this than I do, and I may be off in my estimations, but anyway...) Adire's Rava complete subwoofer (using a Shiva and an AVA250 amp) has a Vb of approximately 63l, and by all accounts is much more than decent. By using a pair of Shivas in iso, you could have a sub with the same Qtc in half the space, so about 32l. I think it achieves A Qtc of .7 or so with stuffing, which is referred to as "maximally flat".
Thomas' sub starts with the Rava's Vb, but due to his Shiva's iso configuration, "sees" a volume of 124l. A shiva in that size box will have a lower Qtc (well below 6 stuffed) and, as a result, have better transient response and lower extension. This Qtc approaches "transient perfect". Thomas is spoiled by his IB's, so I'd imagine nothing less would suit him at this point.
It really comes down to taste, in the end. While lower Q subs do have the transient advantage, and will dip a bit lower, subs with a Qtc of .7 are seen sometimes seen as more "punchy" and well suited to ht/music double-duty.
And, in comparing these examples, I tend to magnify the differences for clarity. The .7 will not sound "boomy" or ill-defined, and will probably sound "tighter" than any reflex or PR sub you'd care to name.

 

[Brian J Dupuis]

Thomas' sub starts with the Rava's Vb, but due to his Shiva's iso configuration, "sees" a volume of 124l. A shiva in that size box will have a lower Qtc (well below 6 stuffed) and, as a result, have better transient response and lower extension. This Qtc approaches "transient perfect". Thomas is spoiled by his IB's, so I'd imagine nothing less would suit him at this point.

I think I'm starting to get the pattern through my thick skull... what I'm reading into this (uh-oh) is that any additional Vb on top of the 32l for the iso config lowers Qtc and moves more towards the "transient perfect" side and adds a bit of extension to the mix. I'm sure there's a point of diminishing returns here, so you don't want to overdo things, but anywhere between the 32l and 64l is the "sweet spot" for a Qtc between .7 and .6 and between good extension and better extension. Am I "getting it", or am I making invalid jumps in logic?
After visiting Thomas' page, I see what you mean about "nothing less would suit him." I shudder to think what his house must sound like... I can't fathom what it must be like to have useable output at 9Hz(!) . Though what you would find to actually generate such tones (aside from test equipment), I must admit I have no idea. Still, that's bragging rights for you.
(As an aside, when I was in college many years ago, I was discussing audio systems with someone. This fellow felt he had to brag.
"My system will play down to zero hertz," says he.
"Same here," says I, pointing at my little JBL 4410s. "See?! There they go! Just listen to that extension!" )

 

[Jack Gilvey]

quote: I'm sure there's a point of diminishing returns here, so you don't want to overdo things, but anywhere between the 32l and 64l is the "sweet spot" for a Qtc between .7 and .6 and between good extension and better extension. Am I "getting it", or am I making invalid jumps in logic?[/quote]
No, I think that's fair. A Qtc of .707, or "maximally" flat", will have the lowest F3 for a given driver, and good transient reponse. A Qtc of .577 (edit: actually 0.5), or "transient perfect", will have a higher F3, but a shallower rolloff, allowing more deep bass (and , of course, the transient superiority that the name implies). Some prefer higher (some car subs) and some prefer lower (Thomas' IBs). I think 6-7 is a good range, though. Some consider a very low Q (below 5) to be "overdamped".
One thing I'm not particularly well-versed in is the effect of voice-coil heating when a sub is in operation. As I understand it, though, this phenomenon ,which occurs in medium to high-power operation, and can actually raise the Qts of a driver. Designing to a lower Q would be a way to avoid the sub rising above 7 in operation. (I think I've really simplified this, or just gotten it plain wrong

).
I think that transient response really begins to suffer much above Q=.707, so you can see how beginning with a driver with a high Qts leaves little chance of winding up with a good sounding sub.
[Edited last by Jack Gilvey on July 13, 2001 at 09:55 AM]
[Edited last by Jack Gilvey on July 14, 2001 at 07:17 AM]

 

[Brian J Dupuis]

I think that transient response really begins to suffer much above Q=.707, so you can see how beginning with a driver with a high Qts leaves little chance of winding up with a good sounding sub.

Very interesting information. I thank you for your insights. When I get home this evening I'm going to play around with the parameters for an iso Shiva design and see what I can come up with. This thread has wandered wide and far, but there's been a lot of great information and useable tips. Thanks to all for sharing your experiences.
I will most likely make my decision on the entire design this weekend and get the parts on order. I'm currently reading the post that you recommended on debates of sub material (nice timing ), as I had previously resigned myself to MDF. I haven't ruled it out yet, but I also haven't gotten to the bottom of the thread. I do not have a whole lot of woodworking experience, so this entire project is a new enterprise for me... one that I hope to find enjoyable. I will most likely not post additional questions (unless I absolutely run into a wall), and will provide some project pics to share with the community as you guys have. Well, unless it turns out so poorly that I must avoid ridicule

.
Thanks again!

 

[Jon Hancock]

One last grace note on the melody....
Q is a reliabile indicator of transient response. Critically damped transient response occurs at a Q of 0.5. At this Q, at the Fb in the box, anechoic response is down 6 dB. With typical room gain and proper placement, response in room may be flat. This gives response that is comparable in tonal clarity and transient definition to IB sub, and pretty close to the same extension, without any extra "punch" that isn't in the original recording. If you want the most "ideal" response, this is the way to get it. Naturally, there are times one makes compromises in enclosure size and tuning due to other considerations, like space, but in my mind, this is the principle advantage of the isobarik, using the expense of an additional driver to provide something close to an ideal response profile in a small enclosure.
Regards,
Jon