Time Delay Crosovers

[donaldpatten]

I would like to collaborate with any one building crossovers.
I have three 605A 's just back from Great Plains for Re Mag & Re Coning.They Did a surperb job . I have built passive time delay crossovers (really a work in progress ) . My Listening Panel just heard them and said they sounded INCREDIBLE .Any one interested ?

[homme_de_terre]

Hi Donald,

I have a pair of 604-8Ks and have been pondering time-alignment since I bought them in 2001. Due to their ferrite magnet structure the z-axis distance between the LF and the HF voice coil should be around 2" less than that of the Alnico 604/605, which is a good thing I suppose.
For a couple of reasons I don't want to go active, so, with me are preaching to the converted here in my choir
Question is how can I contribute? I do know the basics of filter design including the required math, I can tell the hot end of a soldering iron and I can take MLS measurements with CoolEdit plus the Aurora plug-ins.

Perhaps it's possible to come up with sort of a modular design that works for Alnico, ferrite, multicell and Mantaray horns, which need the usual CD horn eq.
What do you think?

Manfred

[dlp4341]

Manfred,

I have an eMac computer with a Scope and FTT signals and analysis. I also have music software to generate tones etc. The hardware you would need includes an omni microphone (electret condenser) with a power supply and preamp.

My test signal for time delay is a 1 kHz tone edited for only 1/2 of a cycle. This pulse is followed by 0.2 seconds of silence and repeated many times. I burned this to a CD and put my CD player on ?track repeat?. The microphone is suspended 12 inches (30 centimeters) on axis in front of the speaker. The preamp output is connected to the eMac input for Scope time delay and FFT frequency response graphs.

The software I use is SignalScope, SignalSuite, Peak 4 (G4+), and Toast 6 Titanium.

My 605A s are 16 ohm (12.5 DCR) on the woofer and 8 ohm on the HF driver. The impedances of the speaker change the crossover components. What are the DC resistances of your speakers?

Don

[Altec19]

Just as a point of reference, are you familiar with Urei studio monitors? Urei produced one of the first, if not *the* first, speaker systems to incorporate E.M. Long's theoretical work on time alignment. I believe all Urei monitors were time aligned and the early models used Altec 604 drivers. The Urei 811A happens to use the 604-8K. Here is a link to JBL Pro Service and the Urei manuals. Some (maybe all) include schematics:

Link to Urei

For what it's worth.

[homme_de_terre]

Don,

I do have a fairly good electret condenser capsule(Sennheiser) for measurment purposes. I yet need to come up with an appropriate body for it.
LF DCR is 6.5 Ohm, HF is 5.8 Ohm. Do you think that voice coil inductance is negligible?

Altec Fan,

I grabbed the Urei 811A manual and it does contain the schematic. Very cool, thanks for the hint.
I'd like to model the Urei crossover in SPICE, I still need to figure out a sufficient speaker model for that purpose. I'm sure voice coil inductance is significant as the woofer's impedance graph shows around 18 Ohm at 1.5 kHz.
Maybe BobR could chime in?

Manfred

[joyspring]

Here are a couple invaluable resources:

Fink, Dennis G., "Time Offset and Crossover Design," [i]J.Audio Eng. Soc.[i], vol. 28, pp.601-611 (1980 Sep.).

Garde, Peter, "All-Pass Crossover Systems," [i]J.Audio Eng. Soc.[i], vol. 28, pp.575-584 (1980 Sep.).

The UREI network schematic exhibits a single-pole high-pass and cascaded Bessel low/all-pass filters to generate delay, a trap (depending on series) and HF equalisation; not a simple filter to design.

The Time-Align(tm) technique uses a proprietary real-time system to quickly realise results.

To be honest, it would be much simpler (and better) to bi-amplify and use an active crossover that incorporates time delay in the low-pass section (i.e. - Rane AC22) and use active (parametric) equalisation for HF and MF gremlins.

The use of MLSSA or similar PC-based measurement system to validate your results will definitely speed the process as well.

best of luck,
BobR

[dlp4341]

Manfred
If as Altec Fan suggests the 811 is the same as your 604s use the 811A schematic. Start the LF input @ L1 ( omitt PLs & Fuses)
Include L1, C1,L2,C2, L3 ,C3 .These make up the Time delay section. L4 C4 R10 are optional. So : 3 coils , 3 resisors and 5 caps and you have the Time Delay Low Pass .
If you have a Mac, Software & Test Mic you can test it.

Altec Fan
I have UREI 809s and knew Ed Long when he worked in Chicago ( 1970s ? ) and have followed his work .
I just wanted to share my 605A crossover work to anyone that always wanted time delay option.I got my 605s in 1963 , and have loving cared for them .

BobR
Thak you for your input. I have used seperates ( amps pre tuner xover ... ); Live is much simpler now . Passive xover 5.1 Recerver, sources and my beloved 605s .
I could never afford all the test gear . Since we got an eMac and software I am amazed at what I can test and measure. Almost free.

[Altec19]

I pulled the front crossover module out of an 811A and poked a camera in the hole. Here's what it saw:

Link to photo

If I'm not mistaken that looks a lot like a 604-8K! Urei would have labeled it an 801AA after their modifications but electrically and magnetically its a 604-8K. It sounds better though, thanks to the Urei crossover.

[homme_de_terre]

BobR

thanks for your statement, I wouldn't have called you, hadn't I anticipated your honesty
In the first place I'm driven by curiosity, means I want to find what time-alignment can do for imaging. There's been much commotion about Jim Thiel's Coherent Source speakers, never heard one and they certainly are never going to replace my beloved Altecs. Actually, I do own a Rane AC22 that I bought on eBay 2 years ago, I just don't want to invest in an appropriate 2nd amp before I'm convinced of the benefits of time alignment. So, copying the Urei low pass section for evaluation purposes seems to be a good start.

Don, Altec Fan,
I've begun studying the Urei schematic; actually the low pass still is a mystery to me. I could easily identify the trap though, its f0 seems to be ~900Hz. I wonder what the cause of this apparent bump in the frequency response is. I measured the 604's frequency response in our company's anechoic chamber and noticed a approx. 4dB low Q bump at around 1.2 kHz and that's independent of the HF attenuator's position. That makes me believe that the woofer is the culprit. Could it be some standing wave caused by reflections of the horn body's back? If so, perhaps it can be cured by the application of a layer of thick felt.
And yes, the 811A definitely sports a 604-8K, thanks for the picture. Cabinet sure is on the small side and I suppose for home application mating it with a sub would be appropriate. I've pondered that too, cause my rear-loaded horn cabs are extremely huge and they do have their immanemt flaws - comb filter effects.
I'll go hunting for the required coils and caps and give it a shot. I'll keep you guys posted and thanks for all the suggestions.

Manfred

[dlp4341]

Manfred , BobR
The imaging is what you get with time delay . We hear instruments that were lost in the mud . I have a friend with a pair of Thiel's and they are superb .Next time you measure frequency response , also measure the woofer bare foot ( without a crossover ) and you will see you'r bump up and then a rolloff in response .
A possible source of capacitors is Solen of France .I think the coils are # 18 AWG ( 1.0 mm Dia ) .BobR got me to look at my old AES Journals , in Vol 24 number 5 ,June '76 :I found Thiel published a paper on how to make Air Core Inductors . Thiel recommends the coil dimensions to be : spool dia = 2 * height
and the outside dia = 4 * height . This gives the coil a square cross section with a maximum of L for the least wire .
Don