Z19 Crossover

Zilch

Curve Junkie
Subscriber
Express interest in a variant of the Altec Model 19 crossover I developed nearly three years ago has motivated me to build up a pair. It was originally posted on Lansing Heritage Forum here:

http://www.audioheritage.org/vbulle...-Altec-9844-8B&p=197971&viewfull=1#post197971

Since then, the design has been built and used successfully by Skywave-Rider, Similost, Dc270, and others. Here's my version, from the original schematic:

attachment.php



Shamefully, I myself had no data on how it actually performs other than the SPICE sims originally posted in Skywave's LHF thread, so I mounted a B&C DE250 on an Altec 511A and dialed it in to mate with the LE14A in my test stand:

attachment.php



attachment.php


Yah, that's +/- 1 dB there.

Yah, biampable.


Here is the wavelet transform:

attachment.php



The cumulative spectral decay:

attachment.php



And the full adjustable range of Z19 with this driver and horn:

attachment.php
 

Attachments

  • Z19 Proto Pic.jpg
    Z19 Proto Pic.jpg
    76.9 KB · Views: 4,040
  • DE250 511A Pic.jpg
    DE250 511A Pic.jpg
    58.2 KB · Views: 3,591
  • DE250 511A +.jpg
    DE250 511A +.jpg
    63.2 KB · Views: 3,894
  • DE250 511A LE14A WLT 325.jpg
    DE250 511A LE14A WLT 325.jpg
    64.6 KB · Views: 3,510
  • DE250 511A Min Max.jpg
    DE250 511A Min Max.jpg
    72 KB · Views: 3,766
  • DE250 511A CSD Z19 625.jpg
    DE250 511A CSD Z19 625.jpg
    103.3 KB · Views: 3,546
Last edited:
For comparison, here's the adjustable range of a stock vintage Altec Model 19 crossover with this driver horn combination, and how flat I could adjust it:

attachment.php


attachment.php



Usable range of "stock" Z19 varies with driver/waveguide combination:

attachment.php


attachment.php


attachment.php


[W/BMS spike-killer notch across the driver = 0.15mH, 0.47uF, 2R7.]


attachment.php
 

Attachments

  • DE250 511A M19.jpg
    DE250 511A M19.jpg
    63.9 KB · Views: 3,153
  • DE250 511A M19 Min Max.jpg
    DE250 511A M19 Min Max.jpg
    70.9 KB · Views: 3,194
  • Samples 1C.jpg
    Samples 1C.jpg
    69.1 KB · Views: 3,100
  • Samples 1B.jpg
    Samples 1B.jpg
    68.1 KB · Views: 3,103
  • Samples 1A.jpg
    Samples 1A.jpg
    68.4 KB · Views: 3,125
  • Samples 1.jpg
    Samples 1.jpg
    75.4 KB · Views: 3,158
Last edited:
Is the "final" schematic shown in post # 118 of the LHF post you linked to? The number of caps don't match (but they could be in parallel).

http://www.audioheritage.org/vbulle...-Altec-9844-8B&p=197971&viewfull=1#post197971

So how does the adjustable range look with the B&C DE250 on the QSC 152i waveguide? Is the DE250 in a safe operating range when the mid-range is maxed out?

Is the LF section woofer specific or is it really part of the secret sauce that makes it all work?
 
LF section is woofer-specific, just like in the 'standard' E-wave setup. Tune the LF to the woofer, to match the horn/highpass section specified here.

Regards,
Gordon.
 
Hey Zilch... Do you have a pair of Selenium D220Ti OMF (bolt on drivers) handy? If so, could I poke and prod you into trying them on the Altec 511 horn, with the Z19, and posting THOSE results? I'm wondering if that, with it's slight 20KHz "peaking", might actually work better on the 511 than the DE250 did... might get darn close to 20KHz. I know that the D220Ti/511 combo performed pretty well on the quick-and-dirty setup I did here a while back using a "standard" E-wave crossover... but, I didn't get to measure it much, other than a quick once-over before it went out to its PA gig...

Regards,
Gordon.
 
Is the "final" schematic shown in post # 118 of the LHF post you linked to? The number of caps don't match (but they could be in parallel).
Yes, I built to that schematic using parallel combos for the 21uF (12||9.1) and 8uF (4.7||3.3). From prior experience with the bridged-T portion of the circuit, I let the 6uF be 6.2 uF, but in retrospect, there's enough room on the board to have made that out of a pair (3.3||2.7), as well. I'll do more testing to verify if that could all be simplified, but I wanted a baseline to spec for starters. The parts list I put together for Simi back in January is here:

http://www.audiokarma.org/forums/showthread.php?p=3378958#post3378958

Mine are built with all Solen caps, so it was a bit more expensive than the $80 shown there, more like just over $100 with the input terminals included.

So how does the adjustable range look with the B&C DE250 on the QSC 152i waveguide?
As I did with the PT waveguides, I'll run a selection of drivers on both 152i and 511 for you guys, but that will not include Altecs, as the only one I have here needs rehab.

Is the DE250 in a safe operating range when the mid-range is maxed out?
I would not run the DE250 at 1.2 kHz with the mid control at "Max," as there is limited HF comp at that setting. HF comp works by midband attenuation, typically -10 dB or more, meaning the compression driver is only seeing 10% power (at most) at the low end. Thus, in this application, used as intended, DE250 is likely "safe." Geddes runs DE250 down to ~800 Hz in Summa.

Similarly, the little Celestion neodymium CDX1-1731 has a recommended minimum crossover frequency of 2.2kHz. I don't know how low Duke runs them HF comped in his AudioKinesis Prisms, and it would require further analysis for an informed design, but they sound mighty nice on the JBL PT waveguide at 1.2 kHz. It's obvious from the curves I posted why he chose that driver over D220Ti.

Is the LF section woofer specific or is it really part of the secret sauce that makes it all work?
What Gordon said -- that's the lowpass for the 416 woofer in Altec Model 19, and it should be further optimized for alternative woofers, as required.... :yes:
 
Last edited:
I have built several (8) of these (near same) dual section networks, LF being the option. This is an excellent network for the larger 15" Econowave models where dialing in the horn/driver to match the woofer & room becomes magic. Fact is it is the only network I consider for many two way build ups, like the recent ones on my Avatar. The best combination I have found was the larger QSC wave guides, used with the Radian 475 PB driver- ulrea smooth down to 1.2k and up!!

I would highly endorce this network for most any large application! The Selenium 220 tends to get pretty harsh below 1.5k to me- especially with power applied! but with this network it did manage to take the rough sonic edge off that and keep the upper end freq. better blended. Another advantage is better balancing for those odd room anomolies that want to suck up or hammer given frequencies.

While it is more costly to build- its benefirs are well, well worth it!!
DC
 
Yes mine were built on the stripped 19 metal plates- you commented how good they looked at one time!
Thanks again for that!
DC
 
Yes mine were built on the stripped 19 metal plates- you commented how good they looked at one time!
That was quite an impressive build: Z19 in the proverbial 5-pound bag. ;)

Now that you've shown it can be done, my next attempt will take that approach, once I get the performance variables sorted out here.... :thmbsp:
 
Last edited:
Samples 2 - 152i

attachment.php


attachment.php


attachment.php


attachment.php
 

Attachments

  • Samples 2C.jpg
    Samples 2C.jpg
    67.1 KB · Views: 2,935
  • Samples 2B.jpg
    Samples 2B.jpg
    68.5 KB · Views: 2,924
  • Samples 2A.jpg
    Samples 2A.jpg
    68.9 KB · Views: 3,015
  • Samples 2.jpg
    Samples 2.jpg
    76.7 KB · Views: 2,946
Samples 3 - 511A

attachment.php


attachment.php


attachment.php


attachment.php


Bonus -- BMS 4552nd:

attachment.php
 

Attachments

  • Samples 3.jpg
    Samples 3.jpg
    77.1 KB · Views: 2,902
  • Samples 3A.jpg
    Samples 3A.jpg
    69 KB · Views: 2,968
  • Samples 3B.jpg
    Samples 3B.jpg
    69 KB · Views: 2,955
  • Samples 3C.jpg
    Samples 3C.jpg
    68 KB · Views: 2,883
  • Samples 3D.jpg
    Samples 3D.jpg
    67.2 KB · Views: 2,875
Last edited:
Looks like the 152i/D250 combo wins the honking contest. Any thought on what values to change to take out the honk? I'll try to build and test it if someone will point me in the right direction.
 
Looks like the 152i/D250 combo wins the honking contest. Any thought on what values to change to take out the honk? I'll try to build and test it if someone will point me in the right direction.
As DC suggests, I know of no other passive crossover topology having this broad applicability across such a range of driver/waveguide combinations. Add in the Altecs for which it was originally developed some 30+ years ago, and its utility is obvious.

In measuring these samples, my criterion was flat within +/- 2.5 dB. DE250/152i appears among the flattest responses across the widest SPL range (12 dB) with it, actually. The peak at 2 kHz suggests that the core 3rd-order highpass (C2, L2, C3) is somewhat underdamped with that combination, which could be easily resolved in PCD for a final design, if desired:

attachment.php


attachment.php


The advantage, as DC also suggests, is that system power response is adjustable for different room conditions and deployments; the only other crossover design offering this capability of which I am aware is by Tannoy....
 
Last edited:
Looks like the 152i/D250 combo wins the honking contest. Any thought on what values to change to take out the honk? I'll try to build and test it if someone will point me in the right direction.

I can tell you the QSC waveguide and the DE 250 are an excellent combination.

Russellc
 
The advantage, as DC also suggests, is that system power response is adjustable for different room conditions and deployments; the only other crossover design offering this capability of which I am aware is by Tannoy....

What about JBL's 4430 monitor? (albeit designed specifically for the 2344 horn)

Russellc
 
I can tell you the QSC waveguide and the DE 250 are an excellent combination.

Russellc

Yes, I have them running right now paired up with a kappalite 3015. I'm using the "LE14" variant of the eWave crossover. I'm trying to work up a better crossover but adventures with PCD and HOLMImpulse have not been very successful (most likely due to operator error). The LE14 crossover sounds good enough that I can't find the motivation to devote an entire weekend or whatever it takes to figure it out.
 
Back
Top Bottom