Sub: 1 large or 2 small drivers ?

[Bigun]

back to the design...

staying with two drivers I like the push-pull approach. Here's an interesting write-up:

http://www.teresaudio.com/haven/subs/subs.html

But it's, well, it's simply, well, not pretty. With the driver showing it's magnet hanging out there. Should be possible to tidy this up. But maybe I'm creating an audible disaster with this idea ?

 

[planet10]

staying with two drivers I like the push-pull approach. Here's an interesting write-up:

Inspired by his buddies TL here:
http://t-linespeakers.org/projects/steve/index.html

In both cases the line really is too small.

dave

 

[Bigun]

Wow, a small world (how do patent lawyers keep up). This TL design is the only alternative I'm looking at to a sealed box - I ruled out a BR because, well, it's boring (!). The sealed box is the easier option no doubt and more flexible in terms of design. I also see a majority of opinion that it provides the best quality of bass ('tighter' and response fall off not as steep). But I remember somewhere that you switched your SDX7 experiments away from a TL to a sealed box. I wonder if this would still have been your decision if starting with a larger/different driver ?

Anyhow, I don't like seeing the gizzards of my speakers hanging out the sides - do you see anything 'wrong' with the option I sketched up to recess the backwards speaker inside the box so that it don't hang out ?

 

[tinitus]

The sealed box is the easier option no doubt and more flexible in terms of design.

Ehh, not really, on the contrary...its an old misunderstanding

 

[marchel]

Personally , I dont even consider the sdx7 a sub, BUt the application of that driver for bass duty is a personal taste.

So chosing between the sdx10 and 2 sdx7 is difficult cause personal pref. comes into play.

But if we change the example to choosing bet. 2 12's and an 18, which have roughly same cone area. I'd choose the 2 12's cause it is much easier to find and cheaper to find 2 12's with XMAX of more that 12mm than an 18 with the same Xmax. But then again , personal preference come in to play when choosing what's best for you.

 

[Bigun]

I think the advantage of the SDX7 was allowing relatively high XO, but this isn't a factor in my application. I am leaning towards the use of two TRIO8's or two SDX10's.

I guess the debate (for me) is over, I'm going to use two drivers.

Instead, I'm trying to choose between a sealed box and folded TL (as in the PMC subs). I stand corrected on assuming the sealed box to be simple, but relative to the TL where I have no experience or simulation software the sealed box is much lower risk. The tall pipe version (un-folded) of the TL doesn't fit my application that well (not to mention low WAF).

This thread caught my interest: http://www.diyaudio.com/forums/showthread.php?s=&threadid=98478&highlight=

But for a folded TL I have no idea how long the line should be for my drivers or how the length should change if I stuff it. This is why the sealed box is looking more attractive.

Then theres' the issue of whether drivers face sideways to front-back. I personally like to see a driver facing front instead of the plain front of the box.

 

[marchel]

HI,

If I were to make a sub, I'd go sealed cause it is more practical, But if you're toying with the idea of a TL, The lenght should be 1/4 the wavelength of the subs.

The formula:

1130= aprox. speed of sound in ft/sec.

So...

1130 divided by (FS of the subs) divided by 4, then you'll get the lenght
of the TL.

In the case of the trio 8.

1130/27/4= 10.46 ft. Which is not too short.

Stuffing the TL between 1.25 to 1.5 pounds per cubic feet should reduce the speed of sound travelling inside the line to about 70%, Which would effectively make the lenght of the TL about 3/4 the wavelenght of the fs of the sub. stuffing is a matter of trial and error though, there are no concrete guidelines about it.

Which brings us to the next question , Would you think that a large 2 trio8 in TL sub is better than the smaller 2 trio10 in a sealed box.

I'd definitely choose 2 10's in a sealed box than a much larger 2 8's in a TL. This is also why TL is not a very popular box for subs either commercial or DIY.

 

[Bigun]

That's an interesting question - I thought that the main reason there were few commercial TL subs was cost. A rigid cube (often ported) makes for a small and cheaper sub for the wants of most HT people. The PMC sub's are TL and based on the PMC FB1's I have the base is very deep and smooth ???

If I stick with the 2 TRIO8's in a sealed box there's still the question of best tuning. With lower Qtc I can push the resonant frequency down. I assume I want to avoid too high a resonant frequency with the risk of significant driver excursions. The disadvantage I hear is that lower tuning reduces driver loading - no free lunch.

Here's my analysis:

 

[tinitus]

http://www.tymphany.com/830952

This 12" seems to work very well in closed...nice smooth response too

 

[GM]

1130 divided by (FS of the subs) divided by 4, then you'll get the lenght
of the TL.

Stuffing the TL between 1.25 to 1.5 pounds per cubic feet should reduce the speed of sound travelling inside the line to about 70%.........

Greets!

Apparently you're not 'up to speed' WRT current TL, etc. design theory: http://www.quarter-wave.com/

GM

 

[planet10]

But I remember somewhere that you switched your SDX7 experiments away from a TL to a sealed box. I wonder if this would still have been your decision if starting with a larger/different driver ?

As marchel points out SDX7 is really a woofer. My use revolves around using it up fairly high, and its mid-bass/low midrange performance is key.

And with the XBL, it does perform like a larger woofer within its limits, but when it hits its limits it is dramatic (same as the FR125) -- it hits its stops in a sudden and disturbing manner. In a TL with the driver unloading below its tuning i ended up stuffing it till aperiodic. (Note: i knew from the get-go that my TL was compromised in terms of length and volume)

The sealed box gives better control below cut-off, allows a bit of LF boost if you are willing to trade extension for max level. It also meshes better with any room gain The SDX7 also works well in a surprisingly small sealed box which fits in with the designs we have applied it to.

Whether i'd do a TL depends on the driver, the application and the size of the box i can live with -- i have a long term plot to install a pair of TLs with push-push 15s for low bass... they will take up much of the 3' x 12' x 12' volume of the top of my workshop.

When i was 1st exploring push-push, push-pull, Nick McKinney said something that stuck... if you have a woofer with good linearity, push-pull buys you very little and comes with quite a few negatives -- especially if you have vented pole pieces. This certainly covers off the XBL woofers.

dave

 

[planet10]

Marchel,

I see GM has spanked you a bit already, but for the benefit od everyone else, a couple pervasive myths need addressing.

The length should be 1/4 the wavelength of the subs.

The length of a TL depends on its shape. A constant cross-section line should be ~ 1/4 Fs (tuning a bit higher has advantages and one has to consider end correction). If the line tapers /flares gettng smaller towards the terminus, then the line should be shorter, if larger towards the terminus it needs to be longer.

Stuffing the TL between 1.25 to 1.5 pounds per cubic feet should reduce the speed of sound travelling inside the line to about 70%, Which would effectively make the lenght of the TL about 3/4 the wavelenght of the fs of the sub. stuffing is a matter of trial and error though, there are no concrete guidelines about it.

The damping DOES NOT significantly alter the speed of sound. I think early TL designers were assigning a symptom (shorter possible length) to the wrong cause (taper allowed for shorter lines, not damping)

Further, at these kind of stuffing densities the line becomes aperiodic and that it looks like a TL becomes largly cosmetic.

dave

 

[planet10]

the size of the box i can live with

Here is a single SD12 TL. A dual SDX10, loosely based on the Vas & Fs differences would be a tiny bit shorter and about 20% larger cross-section. TRIO8 would be a bit shorter yet again, but about the same cross-section.

http://p10hifi.net/planet10/CSS/SD12-TL-map-0v5.pdf

Get your faux marble painting kit out...

dave

 

[Bigun]

http://www.tymphany.com/830952
This 12" seems to work very well in closed...nice smooth response too

It does look nice, but the plan for a design based on two would come in a bit more expensive based on this driver, but I do like the higher sensitivity.

Dave,

That single fold-TL looks nice, but it's too tall, I'd need to fold it over again. I guess that would get it down to roughly one metre tall. But I don't think I need to aim for such a low tuning, as nice as it would be. A box more in line with the PMC-clone idea, perhaps 20% larger, would have the higher WAF. You did a nice sketch of something like this a couple of years ago:

http://www.diyaudio.com/forums/attachment.php?s=&postid=1165679&stamp=1174615919

would this have much of an advantage over a dipole sealed box ?

 

[planet10]

That single fold-TL looks nice, but it's too tall, I'd need to fold it over again. I guess that would get it down to roughly one metre tall. But I don't think I need to aim for such a low tuning, as nice as it would be.

Folding it again would make it deeper... shortening the line would mean increasing the cross-section. The volume of the TL is fixed.

would this have much of an advantage over a dipole sealed box ?

Bipole sealed box? (acknowledging that at LF it isn't a bipole, but an omin-pole just as if it had 1, 3, 4 ... drivers instead)

Depends on the driver, and whether you want a smaller box.

In most real rooms/WAF situations, with the SD12, the sealed is going to win every time. For example contrast the single SD12 TL with this quick-n-nasty visualization of a sealed push-push sub.

Note that at the frequencies involved it makes no difference whether the drivers are side firing or fore & aft.

dave

 

[Bigun]

Yep, the bog TL isn't a good option unless there's a real benefit from trying to fake it in a folded box.

I've determined that the best shape, taking into account WAF and location is the PMC-clone so I now just have to decide between a) drivers on the sides, b) drivers on the front-back. And it seems that this is mostly a cosmetic decision. In that case the only remaining decision is whether push-push or push-pull. The push-push is certainly simpler, but then I don't mind the challenge.

Here's the rough-up of driver on the front (which I've shown push-pull but could be push-push)

 

[Bigun]

and the rough-up of driver on the side, where push-push allows a narrower front aspect.

I do have to finalize Vb, and I'm wondering if these things are better with handles, or do you just stick your fingers into the drivers for a good grip

 

[Bigun]

push-pull buys you very little and comes with quite a few negatives -- especially if you have vented pole pieces. This certainly covers off the XBL woofers.
dave

What are the negatives regarding vented pole pieces ? (is it keeping them clean ?)

 

[planet10]

What are the negatives regarding vented pole pieces ? (is it keeping them clean ?)

They are noisy & can have significant oil can rsonances.

dave

 

[Bigun]

Guys,

If we go with the push-push and I pick a volume of around 50 L my box design starts to come into being. Drivers in parallel gives me the 4 ohm load suitable for a Bash-300 which I can neatly accommodate on the back. What's the recommended level of box stuffing - and how much would I then reduce the designed physical volume by to keep the effective volume at 50L ?

Also, what's not clear to me is the influence of the box self-resonant frequency (of just over 45Hz) on the performance of the sub at even lower frequencies ?

p.s. Dave - I saw a Krell sub which has a similar shape to your 3-d diagram above. They have two 15" drivers with a substantial integrated 2.6kW amp !