Building a custom center speaker

[rdgrimes]

OK It's 6dB down at 80hz. If it works well in the system you want to build this center for then don't go ported. The 115H's sealed are right in the same ballpark. Probably be a bigger box than the PC600 though with a pair 115 in there.



There have been a couple of DIY centers here. It's work so you have to want to do it. You won't be able to use the stock low pass for the "woofers" you will have to tweak it a bit because of the impedance/driver differences.

Rob

Not too concerned about lower impedance due to the 2 woofers, my amp can take it. But what, if anything, would adding 2 woofers to an XO designed for one actually do to the total impedance. Would it also affect the performance of the XO in some other way?

[Robh3606]

But what, if anything, would adding 2 woofers to an XO designed for one actually do to the total impedance. Would it also affect the performance of the XO in some other way?

A picture is worth a thousand words so here is what LEAP tells us. This is the L86 low pass with an 8 ohm resistor and a 4 ohm resistor. The original woofer would be 8 ohms Red and the 115Hs would be 4 ohms Blue. As you can see the curves are quite different and if you didn't tweak the network for a 4 ohm woofers you would end up with a hole in the response at the crossover point.

The second set of curves are with a quick correction using the double the capacitance and halve the inductor value you see on the Internet when you half the impedance. That's the Green curve.

The last curve is adjusting the resistor from 5.1 to 2.4 with the 4 ohm load.

Now keep in mind these are dummy resistive loads. Differences between real reactive loads like different woofers in different enclosures, with their unique impedance peaks and voice coil inductances, can change things quite a bit . They are illustrative only. The real deal may need a fair amount more tweaking than this to get it right.


Rob

[rdgrimes]

A picture is worth a thousand words so here is what LEAP tells us. This is the L86 low pass with an 8 ohm resistor and a 4 ohm resistor. The original woofer would be 8 ohms Red and the 115Hs would be 4 ohms Blue. As you can see the curves are quite different and if you didn't tweak the network for a 4 ohm woofers you would end up with a hole in the response at the crossover point.

The second set of curves are with a quick correction using the double the capacitance and half the inductor value you see on the Internet when you half the impedance. That's the Green curve.

The last curve is adjusting the resistor from 5.1 to 2.4 with the 4 ohm load.

Now keep in mind these are dummy resistive loads. Differences between real reactive loads like different woofers in different enclosures, with their unique impedance peaks and voice coil inductances, can change things quite a bit . They are illustrative only. The real deal may need a fair amount more tweaking than this to get it right.


Rob

I see what you mean, and almost understand it all. It's still within the realm of what a good auto-setup parametric EQ can compensate for though. I am also considering using the XO from a PC600, though it appears that this would also open a can of worms with the different drivers.

It's also not out of the question to use the XO from an L96/L112, since they have the MF/HF adjustments.

[Joe Alesi]

Hello RD,

If I wanted a small/medium centre channel for >80Hz I would consider 2 x 2118 and a 2404 in a d'appolito style cabinet. Crossover and cabinet volumes could be taken from the 4612.

Best
JA

[Robh3606]

2404 or possibly a 2405

2404 definately a 2405 won't play low enough


Rob

[Joe Alesi]

Oppsie.....Good point Rob. I will correct my post before anyone pops a 2405

Best
JA

[duaneage]

Here is a project I did a while back
http://audioheritage.org/vbulletin/s...ght=l20t+tower

Bass was substantial as the bigger box and lower tuning attest to but power handling suffered a bit. The 115 is a higher Qts driver that produces ripple response in vented alignments and does better with smaller box/higher tuning.

The sealed arrangement using two drivers would handle enough power. Even though the rolloff looks bad, the gradual slope makes for better transient response while the sealed alignment could tolerate some EQ boosting, certainly more boost than a vented alignment where out of band signals could bottom the cone.

The delay times are much better for the sealed alignment. I'm considering a dual 115H design this winter (not a center but a stereo pair).

[rdgrimes]

I'm also thinking about a pair of 706G-1 drivers.

[SMKSoundPro]

I have used 4612 for many things, and see where it makes a great center channel speaker!

You could even build a flat wide enclosure for it.

link: http://www.jblproservice.com/pdf/The...ies/46120K.pdf

Its actually a woof/mid/tweeter design.

(There are 8 of them at the Bush Company driving hard above the strippers and 4 2245's in 4718 cabs above the meatrack I put in 1984. What a summer! Who knew that's where I would meet my ......)

Scotty.

[rdgrimes]

I'm working on this project now, found a pair of abused L86 to use for parts.

I have a pair of 706G-1, and a pair of 115H. I'll prolly go with the 115H because the white cone is too hard to resist. I'm choosing an L96 XO in hopes of getting a better voice match with the rest of the L system. (150A, 112, 96). So the drivers will be 044 and LE5-12 aligned vertical with the 115Hs vertically centered on the sides.

It all measures out to fit snugly in the baffle of the L86. I hope to use the rear plate from one L86 as the baffle, but may wind up making a new one. The L86 cab has about 2.5x the cubic space as the L15, so I'm planning to start with no port but that can change.

My biggest question mark remains the XO, and whether to do anything with the L96 XO to accommodate the 2 LF drivers vs one. I assume the way to go is with the 115H in parallel, but what happens if they're in series?

My assumption is that my PEQ receiver setup will EQ these things pretty well for a flatter response, but with the L96 XO I'll also have the pots for adjustments to HF/MF.

The entire system will be fed with 400W/ch.

[duaneage]

I'm working on this project now, found a pair of abused L86 to use for parts.

I have a pair of 706G-1, and a pair of 115H. I'll prolly go with the 115H because the white cone is too hard to resist. I'm choosing an L96 XO in hopes of getting a better voice match with the rest of the L system. (150A, 112, 96). So the drivers will be 044 and LE5-12 aligned vertical with the 115Hs vertically centered on the sides.

It all measures out to fit snugly in the baffle of the L86. I hope to use the rear plate from one L86 as the baffle, but may wind up making a new one. The L86 cab has about 2.5x the cubic space as the L15, so I'm planning to start with no port but that can change.

My biggest question mark remains the XO, and whether to do anything with the L96 XO to accommodate the 2 LF drivers vs one. I assume the way to go is with the 115H in parallel, but what happens if they're in series?

My assumption is that my PEQ receiver setup will EQ these things pretty well for a flatter response, but with the L96 XO I'll also have the pots for adjustments to HF/MF.

The entire system will be fed with 400W/ch.

If you double up the woofers start with reducing the lowpass component values by one half. Crossovers using dual LF drivers tend to be a bit picky about lobes and response fields. 3rd order networks seem to work the best. The physical distance matters as does the MF crossover point.


This may sound trivial and time consuming but if you intend to have a center channel that provides voice without being to focused it bears some testing. Test the results by moving around the speaker at different angles and see how it responds off axis.

[rdgrimes]

If you double up the woofers start with reducing the lowpass component values by one half. Crossovers using dual LF drivers tend to be a bit picky about lobes and response fields. 3rd order networks seem to work the best. The physical distance matters as does the MF crossover point.

Huh?

[duaneage]

You said you were using two woofers with the mid and high between them, and a L96 XO. The L96 XO expects a different impedance for the woofers, probably around 8 ohms, whereas the two woofers you have will probably be 3-4 ohms. This will drastically alter the crossover for the woofers from what the single 12 had in the L96.

The L96 network is a good starting point, certainly for the midrange and the tweeter, but some tweaking is needed for it to work properly. unless, that is, you're not concerned with little things like off axis response or lobing.

Joe D'Appolito did early development work on the MTM arrangement popular with center channels and he advocated 3rd order networks with a crossover point chosen that puts the wavelengths complementary to each other. The acoustic center of the two drivers is between them but not using a steep network can cause some harsh side effects.

Use white noise or pink noise and move off axis from the speaker to see how the tone changes.

[rdgrimes]

Project is nearing completion. It's turning out pretty decent for a seat-of-the-pants custom build. All that's really left is some testing, but whatever it winds up sounding like, that's as far as it's going to go.

All vintage materials, save for grill pegs. The baffle is the rear plate from another L86. The cab was pretty beat up, but wound up being passable.

[grumpy]

Nice salvage/reuse project!