Crossover frequency does matter

Anything not covered elsewhere.
Message
Author
Bruce Weldy
Posts: 8301
Joined: Fri Nov 27, 2009 11:37 am
Location: New Braunfels, TX

Crossover frequency does matter

#1 Post by Bruce Weldy »

Well, it's a slow Easter Sunday here on the forum.....so, I though I'd start a little discussion.

First, an observation concerning different speakers in the same room. I have run sound in a particular venue here in town several times - pretty evenly split between a rented system that a provider brings in and the one that the band I work with owns.

I was in there last night with the same band on their system and both the owner and his son were gushing about how good the vocals sounded - and they sounded damn good to me too.

The rented system is powered JBL PRX 725 mains. The band system is 25 yr old Peavey mains.

I used both systems with the same band (a really good singer with a great voice). The new JBLs don't hold a candle to the ancient Peaveys. Why? I think it all boils down to crossover points. The JBLs are crossed at 2.4k, the Peaveys are 1.2k. The lower crossover point in the old Peaveys really lets the vocals get clean, crisp, and present. The JBLs are harsh when you hit 'em with any volume...my guess is that it's due to the lack of higher frequencies expected to be covered by the 15 inch drivers.

If crossing that high was a decision made for better sound, then why are the SRX series equipped with better drivers and crossed at 1.2k? Well, obviously it is a matter of price. Cheaper boxes use cheaper HF drivers that can't go as low and that equals crappy sound.

Now, how does this relate to our world? The melded array is crossed at 2k, but really jumps up at the 2.5k mark. That's the reason that Bill requires that you use a low frequency driver that has rising mids - this helps fill that gap. The CD option with a good driver (the NSD2005 or better) lets you cross at 1.2k. The ASD drivers use a 2k crossover like the piezos.

Having used both the melded array and the NSD2005 - while I've always been impressed with the horizontal dispersion of the melded array, the CD horn just flat sounds better. I'm convinced that it's because of the lower crossover point.

I currently stack a CD-equipped OT12 on top of a melded box - they sound good together....kind'a best of both worlds and I'll continue doing this. But personally, if I build more - I will opt for the lower crossover point and give up some of the dispersion.

Now, if I'm comparing the melded array vs. those same JBLs, I'd definitely say that the melded wins because of the rising midrange of the woofer.

In summation - my point here is that the decision to pick the melded over the CD horn (only the better driver - the ASD is crossed at the same place as the melded, so there's no advantage) can be based on a variety of factors. The melded array is a great low-cost tweeter and it sounds good.....hell, it sounds better than most of the off-the-shelf stuff that's out there. But, if you are looking for the best quality, especially in live sound, the CD horn is the hands down winner for clarity - and it competes with the higher-end boxes that cost a lot more and also use a 1.2k crossover.

And if you are going to buy off-the-shelf, then pay attention to the specs and buy a box with a lower crossover point - it will sound better.

Comments?

6 - T39 3012LF
4 - OT12 2512
1 - T24
1 - SLA Pro
2 - XF210


"A system with a few knobs set up by someone who knows what they are doing is always better than one with a lot of knobs set up by someone who doesn't."

Bruce Weldy
Posts: 8301
Joined: Fri Nov 27, 2009 11:37 am
Location: New Braunfels, TX

Re: Crossover frequency does matter

#2 Post by Bruce Weldy »

Oh, and I might add.....last weekend I ran 11 acts over two days at a different venue than the one listed above. This was a QSC system. The horn in those boxes are crossed higher also, BUT - they were three way and the midrange speaker helped to make up the difference to keep the vocals pretty nice.

There's just too much gap in most all 2-way boxes that are crossed at 2.5k.

6 - T39 3012LF
4 - OT12 2512
1 - T24
1 - SLA Pro
2 - XF210


"A system with a few knobs set up by someone who knows what they are doing is always better than one with a lot of knobs set up by someone who doesn't."

User avatar
Chris_Allen
Posts: 3355
Joined: Wed Jul 11, 2007 2:43 pm
Location: Huddersfield, UK

Re: Crossover frequency does matter

#3 Post by Chris_Allen »

Interesting point.

Would a large piezo go lower?

and could you cross the SLA Pro at 5KHz? I always wanted to try the SLA Pro with some cheap ribbons.
Built:
6xDR200, 2xT39, 2xT48, 2xJack110, 1xOmni10.5, 1xAutotuba, 1xT18, 1xSLA Pro, 1xW8, 1xW10

Ryan Sober
Posts: 204
Joined: Sat Feb 01, 2014 8:34 pm
Location: Birmingham, AL

Re: Crossover frequency does matter

#4 Post by Ryan Sober »

IMO, you've got it backwards:
The cheap tweeters can't go as low... and because they're cheap, they also sound worse.
More expensive, better-made tweeters can go lower... and because they're better-made, they sound better.

In other words, the crossover frequency isn't the actual reason the cabs sound better... it's the quality of the components.

I'm not sure it really matters which is the cart and which is the horse, though; the cost vs. quality law is still in full effect.

Edit: PS. The PRX735 cabs sound far and away better than the 725... abandoning the middle 15" for an 6.5" mid makes a world of difference.

Bruce Weldy
Posts: 8301
Joined: Fri Nov 27, 2009 11:37 am
Location: New Braunfels, TX

Re: Crossover frequency does matter

#5 Post by Bruce Weldy »

Ryan Sober wrote:IMO, you've got it backwards:
The cheap tweeters can't go as low... and because they're cheap, they also sound worse.
More expensive, better-made tweeters can go lower... and because they're better-made, they sound better.

In other words, the crossover frequency isn't the actual reason the cabs sound better... it's the quality of the components.
I think I'd disagree with that to some degree..... Just because a tweeter doesn't low doesn't mean it's cheap, but rather that it takes less materials, less strength of materials, etc. to not go lower. Granted, there is some junk out there. However, the driver in the PRX 725 costs from $135-160 to replace...that's not cheap for a 1.5 in. diaphragm driver.

I'm sure that JBL 2.5k tweeter is just fine.....from 2.5k and up. It's the missing area from 1.2k to 2.5k that i think makes a box sound better - when it comes from the horn and not the woofer.

Edit: PS. The PRX735 cabs sound far and away better than the 725... abandoning the middle 15" for an 6.5" mid makes a world of difference.
Exactly my point in the second post. If you have a driver that's designed to fill that void, it's going to sound better.....just as the QSC boxes that I referenced.

But here's an interesting note....both the 725 and the 735 use the exact same driver, but the 725 is crossed at 2.4 and the 735 is crossed at 2.0. I would have guessed that they would have crossed the three-way box higher since the mid-driver is there to help. I wonder if they use a much steeper slope in the 735 to protect the driver....

If you step up to the SRX725, you get the 1.2k crossover point with a 4 inch diahragm and a 1.4 inch exit.....for a replacement cost of $1500....each.... :shock:

Oh well......thanks for the discussion. And I think we agree that a better horn is worth the money.

6 - T39 3012LF
4 - OT12 2512
1 - T24
1 - SLA Pro
2 - XF210


"A system with a few knobs set up by someone who knows what they are doing is always better than one with a lot of knobs set up by someone who doesn't."

User avatar
Bill Fitzmaurice
Site Admin
Posts: 28620
Joined: Tue May 02, 2006 5:59 pm

Re: Crossover frequency does matter

#6 Post by Bill Fitzmaurice »

Part of the problem with the JBL et al is using direct radiating fifteens. Whatever they produce above 1.5kHz or so will have weak dispersion. You can improve response with EQ, but not dispersion. The narrow horn throat of our horn loaded fifteens, and tens and twelves for that matter, gives wider dispersion than a direct radiator using the same driver size, allowing them to go to 2kHz.

Bruce Weldy
Posts: 8301
Joined: Fri Nov 27, 2009 11:37 am
Location: New Braunfels, TX

Re: Crossover frequency does matter

#7 Post by Bruce Weldy »

Bill Fitzmaurice wrote:Part of the problem with the JBL et al is using direct radiating fifteens. Whatever they produce above 1.5kHz or so will have weak dispersion. You can improve response with EQ, but not dispersion. The narrow horn throat of our horn loaded fifteens, and tens and twelves for that matter, gives wider dispersion than a direct radiator using the same driver size, allowing them to go to 2kHz.
Good point.

But even directly on-axis, the 15s in those boxes don't seem to fill the gap. Not bashing JBL, just using them as an example as most all manufacturers are building these ridiculous double 15 boxes with a horn that can't come close to keeping up.

JBL used to build a 2x12 box (4732) with a 2 inch exit horn. I guess that just doesn't "look" big enough for the average consumer.....but that's perfect to use with subs.

6 - T39 3012LF
4 - OT12 2512
1 - T24
1 - SLA Pro
2 - XF210


"A system with a few knobs set up by someone who knows what they are doing is always better than one with a lot of knobs set up by someone who doesn't."

Grant Bunter
Posts: 6915
Joined: Mon Nov 28, 2011 11:12 am
Location: Ilfracombe Queensland Australia
Contact:

Re: Crossover frequency does matter

#8 Post by Grant Bunter »

Bruce,
Thanks!! This thread has got my mind going, as I intend to built some CD DR250's in the near future. Got a box full of goodies from Leland a while back to that end.

I know your thread doesn't particularly revolve around piezo's, but they are a great place to start.
Why? Because a positive enquiring mind might wonder how a bunch of piezo's ie the melded array or flat array, can have a 2K corner frequency crossover, with a HF array made out of piezo's, that, as far as the manufacturer states, has a useable frequency response of 3K and up.
I can only assume that adding the piezo components raises sensitivity to the point where there is adequate 2K response, in combination with the rising response of the mid driver.
Funnily enough, one guy who has used my system has dropped out 800Hz, and only 800Hz, outdoors, and loved that (with Beta10a loaded DR250's), and that doesn't figure into crossover point whatsoever.

Move on to the 2K corner frequency crossover for dual ASD's in those cabs it applies to, and, wait a sec, resonant frequency of the ASD is 1.5K, but useable from 2.5K up.
Again, rising response in mid driver must be covering the gap.

Dual NSD's aren't an option in the DR250 due to space restrictions, but, given what you say, I might be better off looking for some 1.2K crossover point suitable CD's, perhaps bolt on to save some space...
Built:
DR 250: x 2 melded array, 2x CD horn, March 2012 plans.
T39's: 4 x 20" KL3010LF , 2 x 28" 3012LF.
WH8: x 6 with melded array wired series/parallel.
Bunter's Audio and Lighting "like"s would be most appreciated...

User avatar
Bill Fitzmaurice
Site Admin
Posts: 28620
Joined: Tue May 02, 2006 5:59 pm

Re: Crossover frequency does matter

#9 Post by Bill Fitzmaurice »

Grant Bunter wrote:a positive enquiring mind might wonder how a bunch of piezo's ie the melded array or flat array, can have a 2K corner frequency crossover, with a HF array made out of piezo's, that, as far as the manufacturer states, has a useable frequency response of 3K and up.
The crossover.

Bruce Weldy
Posts: 8301
Joined: Fri Nov 27, 2009 11:37 am
Location: New Braunfels, TX

Re: Crossover frequency does matter

#10 Post by Bruce Weldy »

Bill Fitzmaurice wrote:
Grant Bunter wrote:a positive enquiring mind might wonder how a bunch of piezo's ie the melded array or flat array, can have a 2K corner frequency crossover, with a HF array made out of piezo's, that, as far as the manufacturer states, has a useable frequency response of 3K and up.
The crossover.
So, does that mean that the manufacturers rate the frequency response limits based on a slope of say 12db/octave, but you can get more usable frequency by using a much steeper curve and setting the crossover point lower?

I've always wondered the same as Grant ..... i.e. The NSD driver is rated to 1.5k using a 12db slope, but the passive crossover in your designs is at 1.2k.

6 - T39 3012LF
4 - OT12 2512
1 - T24
1 - SLA Pro
2 - XF210


"A system with a few knobs set up by someone who knows what they are doing is always better than one with a lot of knobs set up by someone who doesn't."

User avatar
Bill Fitzmaurice
Site Admin
Posts: 28620
Joined: Tue May 02, 2006 5:59 pm

Re: Crossover frequency does matter

#11 Post by Bill Fitzmaurice »

Bruce Weldy wrote: So, does that mean that the manufacturers rate the frequency response limits based on a slope of say 12db/octave, but you can get more usable frequency by using a much steeper curve and setting the crossover point lower?
Correct.

Bruce Weldy
Posts: 8301
Joined: Fri Nov 27, 2009 11:37 am
Location: New Braunfels, TX

Re: Crossover frequency does matter

#12 Post by Bruce Weldy »

Bill Fitzmaurice wrote:
Bruce Weldy wrote: So, does that mean that the manufacturers rate the frequency response limits based on a slope of say 12db/octave, but you can get more usable frequency by using a much steeper curve and setting the crossover point lower?
Correct.
Well damn! Got me a little book learnin' today.

Thanks.

6 - T39 3012LF
4 - OT12 2512
1 - T24
1 - SLA Pro
2 - XF210


"A system with a few knobs set up by someone who knows what they are doing is always better than one with a lot of knobs set up by someone who doesn't."

Grant Bunter
Posts: 6915
Joined: Mon Nov 28, 2011 11:12 am
Location: Ilfracombe Queensland Australia
Contact:

Re: Crossover frequency does matter

#13 Post by Grant Bunter »

Bill Fitzmaurice wrote:
Grant Bunter wrote:a positive enquiring mind might wonder how a bunch of piezo's ie the melded array or flat array, can have a 2K corner frequency crossover, with a HF array made out of piezo's, that, as far as the manufacturer states, has a useable frequency response of 3K and up.
The crossover.
Righto. So, following Bruce's posts as well, you thought, with the GW 1016's, something like "OK, frequency response with a -12dB slope at 3K is -6dB at 1.5K, so, if I instead do a third order LR crossover with a 2K corner frequency (18db/octave, or -9dB at 1K), that's about the same, and I'll have some fill in that region with a 18db/octave falloff from the mid driver to help fill the gap" and it worked, that's how you got there?

Is that how it went at the time?
If so, hats off to you, that's kinda a giant leap for any kinda speaker without a spec chart...
Built:
DR 250: x 2 melded array, 2x CD horn, March 2012 plans.
T39's: 4 x 20" KL3010LF , 2 x 28" 3012LF.
WH8: x 6 with melded array wired series/parallel.
Bunter's Audio and Lighting "like"s would be most appreciated...

User avatar
Bill Fitzmaurice
Site Admin
Posts: 28620
Joined: Tue May 02, 2006 5:59 pm

Re: Crossover frequency does matter

#14 Post by Bill Fitzmaurice »

Grant Bunter wrote: So, following Bruce's posts as well, you thought, with the GW 1016's, something like "OK, frequency response with a -12dB slope at 3K is -6dB at 1.5K, so, if I instead do a third order LR crossover with a 2K corner frequency (18db/octave, or -9dB at 1K), that's about the same, and I'll have some fill in that region with a 18db/octave falloff from the mid driver to help fill the gap" and it worked, that's how you got there?
Is that how it went at the time?
Not even close, piezo breath. :roll:
Piezos are usually used with no crossover. They can because their impedance goes up as frequency goes down, so there comes a point where the rising impedance alone reduces their voltage input for adequate element protection. That's what the manufacturer response spec is based on. They also cannot be used with a traditional crossover, because they are a capacitive load, not resistive.
The 30 ohm resistor in the crossover we use provides the necessary resistive load, while the capacitor and inductor alter the signal transfer function, increasing sensitivity in the 2kHz region, while reducing the voltage input to the element below 2kHz for increased protection.

Grant Bunter
Posts: 6915
Joined: Mon Nov 28, 2011 11:12 am
Location: Ilfracombe Queensland Australia
Contact:

Re: Crossover frequency does matter

#15 Post by Grant Bunter »

Thanks for that...
Built:
DR 250: x 2 melded array, 2x CD horn, March 2012 plans.
T39's: 4 x 20" KL3010LF , 2 x 28" 3012LF.
WH8: x 6 with melded array wired series/parallel.
Bunter's Audio and Lighting "like"s would be most appreciated...

Post Reply