Polarity, again.

[boputnam]

Whole cabinet phazing is readily accomplished with the AA when you can see the cones move.

I've had difficulty with that, particularly for the MF and HF - seems that much of the energy of the AA get's consumed by the network and makes the results less visible.

I've had great luck with the "Cricket" of Galaxy Audio (http://www.galaxyaudio.com/galaxy/Products4.html#TEST EQUIP), thanks to a non-related-party recommendation from Mike Caldwell (http://www.vintageaudiosales.com/). Results are clear and unambiguous.

[Mr. Widget]

" I've had difficulty with that, particularly for the MF and HF - seems that much of the energy of the AA get's consumed by the network and makes the results less visible. "

Since most systems employ crossovers to filter out the low frequencies from the mid and high frequency drivers the very, very, very low frequency of 0Hz coming out of a battery better be filtered out!

[boputnam]

the very, very, very low frequency of 0Hz coming out of a battery better be filtered out!

That'd be it!

[4313B]

"the very, very, very low frequency of 0Hz"

Some days I just skip reading the funny papers and read this forum instead

Thanks Mr. Widget

[martinleewin]

RG; Glad we can be of amusement to you.
BP & Mr. W.; EXACTLY my point... you may have to bust inside or rummage thru a tech lib to figure these things out for yourselves if you are not sure all drivers in a cab are mutually phased. So then YOU know. But will you remember? Scribble or print the schematic and tack it to the back of the cabinet. Staple an envelope flap, thumbtack a sketch or scribble in crayons as you choose.
Now you will always have that ref info when and where you need it. The next owner/user will also appreciated your forethought to save them from the tedious repetition of your research.
Sometimes I upgrade a commercial speaker cabinet for resale with better drivers, alter the crossover specs, and other stuff. I print a new Wizard Labs spec label, laminate and staple it to the back of the cab so customers know what they are getting and what to do with it. Doing so helps the customer get the hookup right the first time and saves tech support calls to my "factory".
Martin W.
Wizard Labs

[Doctor_Electron]

NOTICE:
This is a Dr_Electron-length post.
Several years ago, I undertook a project to re-wire my pair of 4410's, since the crossovers looked like low-rent rats' nests, and I do not trust sonically anything less than soldered connections inside the cabinets.
Unfortunately, I got in a hurry and did not thouroughly document the wiring during the demolition phase of the project, or save the cheesy original wires.
When ready to put it all back together, I requested and received the tech specs PDF for the 4410L,R from JBL Pro customer service @ Northridge, and immediately noticed a couple of odd things about the wiring diagram (in view of my several previous & successful, pain-free speaker re-wires, & experience working on JBL, etc. components going back to the early 1970's).
First was the apparent anomaly of the 127H woofer & 104H-2 being shown connected "in-phase"; with second-order butterworth networks the mid is normally connected 180 degrees out from the woofer. I again contacted JBL, and was "assured" that they "did things differently on different types of systems".
So I went with their #^!^&*%* advice.
Anomaly # 2 was the interconnection of the mid and high drivers' L-Pads "low" terminals, with no connection to "signal common". Again contacting JBL, I was told that the controls were meant to be "interactive", and that the drawing was correct. However, I distinctly recalled that the mid and high output levels as originally wired could be completely turned off (as in every other of the numerous JBL's I had owned). The "interactive" topology did not allow this, and IMO also caused an unaccectable vagueness in tonal balance no matter the %&*&*!# "interactive" settings.
I wound up connecting the L-Pads' common junction to "signal common", and judged the result to be satisfactory (& typical).
Now, back to the phasing questions: With the 127H & 104H-2 drivers connected in-phase, the sound was so horrendous that over a period of time I thought that my CD player, and/or power amp had gone south.
I recall discussing that can of worms with Robh on the phone at the time. He was addressing issues of trimming the values of the fixed padding resistors in some newly acquired JBL's, which were essentiall identical to the 4410's, save port locations, cabinet geometries, and use of fixed pads. As I recall, we weren't able to help each other much at the time on either set of issues.
After much suffering hearing very sick sound, experimenting, documenting & re-documenting every change, as well as trying repeated CD & amp swapping, I satisfactorily addressed the phasing quandary by rigging up a phase-reversal switch with leads long enough to reach the typical listening position, fed the mid and later the high driver through the switch (wiring routed out through the port(s), and finally resolved the problem(s) through listening evaluation. The results were profound and pleasing.
Final verdict: 127H & 035tiA are connected in-phase; the 104H-2 connected out-of-phase.
HOW ABOUT THEM APPLES?
Of course, now I start a notebook for any project (duh, learned that back in '75 on an R&D job supporting engineers, but got lazy & complacent with experience), and document EVERYTHING!
NOTE: Any such audible phase testing must be done from the listening position, as the sound waves are going through normal 360 degree (peak, null [180 d.], peak) wave propagation as they travel from the baffle to the listener. Evaluating / adjusting driver phase at closer distances does not give a real-world sonic result.

[boputnam]

Doctor_Electron

Thanks for that.

I have found that many (but not all...) three-way systems are phased with the MF out-of-phase from the LF and HF. Certainly so for the 4312 L/R, but the 4313B has both the MF and HF phased together, and out-of-phase with the LF. So, it all depends, and in the end, it's how it sounds that matters.

The reason I have devoted so much time to this issue, is that building my own cabinets with seemingly "ideal" compliment of transducers was yielding wobbly results - which I came to discover were polarity and phasing related.

I satisfactorily addressed the phasing quandary by rigging up a phase-reversal switch with leads long enough to reach the typical listening position...

THAT has me ROFL - what a clever idea. It is so hard to A/B systems meaningfully - this sounds patentable!

[4313B]

The phase reversal switch is documented in the JBL Loudspeaker Instruction Manual dating back to the 70's.

The phase also depends on where the loudspeaker will be located with respect to the listener. Up front in your face or across the room.

*****

It is interesting to note though that the L80t/L80t3, which is the Home version of the 4410, has all the drivers in phase as well. No wonder that whole series sounded goofy. Dr. Electron has solved the mystery Actually, I did have a discussion with G.T. about that many years ago and there WAS a reason but I can't remember the specifics now, I really didn't care then and I guess the information didn't "take".

[martinleewin]

Dr. Electron (great name!) shirley illustrated my point of documenting the before so that the during is simpler to accomplish for a satisfactory after.

You indicated an opposite phased midrange driver is common in JBL systems. I have not found this to be the case with the (non-JBL) mostly-cone (some piezo tweeter) transducer systems I am more familiar with. Is this because of the the fact that JBL used midrange compression drivers more often than most manufacturers in their speaker systems?

I know that compression vs. cone transducers are significantly faster on transient response. And that pure conehead designers may set their cone transducers rearward as the cone depth shortens in order to align the effective driver propogation planes with respect to equal time domain starting points. A compression driver with horn already accomplishes this (perhaps too much?) by positioning the midrange well to the rear of the cabinet face.

Many designers use an intermediary smaller cone midrange to bridge the time domain gap between the sluggish woofer and the much quicker (and often more forward) tweeter. IMO 2-way systems rarely sound "in phase" through the critical midrange because of the comparative difference in driver speed no matter how good the frequency response is.

Is there something about compression/horn transducers which prompt the designer to set them in opposite phase-- vs. simply mechanically positioning them to an equal planar source position?

And since I know far less about compression driver theory and construction... can the forum members provide links describing compression driver;
operating principles
acoustic properties
construction details (material properties, cross section or cutaway dwg)
care and feeding...
so I could better understand their behavior?

I think this topic was worth visiting again for the knowledge gained.
Martin W.
Wizard Labs

[DavidF]

Watch them Nulls! I can only think of two influences that may cause a horn to act differently than a cone relating to the crossover design. One is the longer path of the source (group delay) and the other the directivity of the sound wave (power response). The reversal of adjacent drivers is more to do with the summation of the response in the area of the crossover and even this effect is limited to certain crossover design models. For instance, a three-way system on a second order Butterworth may cause a huge null at the crossover frequencies. Reversing the phase of the mid reverses the influence of phase cancellation and provides a smoother frequency response. In any event, the reversed-phase mid is by no means universal for all applications. You have to convince me (I am easy to convince) that horn compression drivers are more (or less) responsive than cones in terms of transients. Would it be more appropriate to state that for a given sound pressure level the driver/horn is using far less energy and effort at a significantly reduced distortion level? A cone has the potential to deliver all the bite and sparkle of a horn. Given that, the driver/horn system has the headroom to sustain accurate dynamics at SPL levels that cones can not match. The remaining differences are probably open to endless debate.

David F

[martinleewin]

DavidF, Thanx for the clarification on the midrange opposite phasing being needed to address the crossover phase shift response.

re; "You have to convince me (I am easy to convince) that horn compression drivers are more (or less) responsive than cones in terms of transients. Would it be more appropriate to state that for a given sound pressure level the driver/horn is using far less energy and effort at a significantly reduced distortion level? A cone has the potential to deliver all the bite and sparkle of a horn."

I am not in a position to convince by means of test data. My opinion was based on some limited listening test examples. Perhaps when I get some realtime analysis software to run pulses with high enough data acquisition rates to evaluate some example drivers I would be able to quantify and substantiate that widely held belief.
Your statement " Given that, the driver/horn system has the headroom to sustain accurate dynamics at SPL levels that cones can not match." does support my premise at the loud end of the amplitude range, doesn't it?

Martin W.

[DavidF]

Martin,

Your statement " Given that, the driver/horn system has the headroom to sustain accurate dynamics at SPL levels that cones can not match." does support my premise at the loud end of the amplitude range, doesn't it?

Yes, and then again, no. Think this through for a minute. Take a Klipsch woofer in reflex cabinet compared to installed on a folded horn. I would expect the transient characteristics to be the same, save around the resonance frequency. However, for the same sound levels, the horn loading is going to significantly reduce the excursion requirements. Now if we push a baffle-mounted cone driver to the point where its breakup causes audible distortion and compression effects all the while the horn-loaded driver, at the same SPL, will be operating at much lower distortion levels. Therefore, the horn-loaded driver may seem to be displaying better transient characteristics. This is not due to better transient characteristics but rather cleaner and more accurate output through the horn mouth. In my modest space, I would not be able to tolerate testing this upper limit “horn benefit” to test the theory. Now if you start off on focused wave propagation mathematical models, you will quickly leave me behind.

What do you think?
David F

[martinleewin]

Now, David, you are bringing up particular design elements and extreme usage which will distort a typical listening picture.
1. A Klipsch-type speaker, with its long folded wave path, introduces substantial delay for the reflex loaded driver(s). This is another design characteristic with its own pros and cons best left for another discussion. Not that that will stop me, though.
re: "Take a Klipsch woofer in reflex cabinet compared to installed on a folded horn." I accept your generous offer to take it off your hands. How about if we split the shipping charges?
"I would expect the transient characteristics to be the same, save around the resonance frequency."
Measured at the driver--perhaps, but no data presented as support. Not having any real design knowledge or measurement data, I must surmise your point is based on the cabinet's air pressure load within the resonant band will inhibit the driver's transient ability. Measured at the listening point--add aroundabout 10 mS delay for a folded horn design. A systematic complication of another sort. Now you need to employ Boseian engineering calculations. I (personally) find this design to be great for increasing driver "space" and low end response while causing multitudes of phase errors within the cabinet's range when not all the drivers are similarly delayed. My ear is particularly sensitive to phase errors induced by artificially extended wavepaths mixed with direct paths. My ear will "buy into it" at longer wavelengths and then objects to the mix when it approaches the midband and above.
"However, for the same sound levels, the horn loading is going to significantly reduce the excursion requirements."
Absolutely. Horn-loading EITHER type driver increases efficiency by decreasing excursion requirements WITHIN the effective HORN BW.
2. Basing an evaluation upon results obtained at maximum safe operating limits is neither prudent nor a fair representation of a speaker system's typical usage and best acoustic performance. I can not afford to damage that many voice coils to test a theory. Something I feel is best left to "pro" audio designers and road concert "sound" engineers to determine the maximum SPL vs. distortion so close to the end of life point. Would anyone care to donate some specimens for evaluation? You can have them back when we're done.
Where were we? Oh, yeah...
Transient response is often a secondary design consideration compared to efficiency and spectral compensation. Good transients alone do not guarantee "good sound". But better-appreciated designs will exhibit TRANSIENT characteristics at least equal to their FREQUENCY response characteristics. Then getting multidriver systems to work together (in or out of PHASE as a whatever works option) and then getting the DELAY times to coincide in a cohesive fashion... and you've got the perfect? Compromise.
I don't blame speaker designers for their limitations. I blame Newton and his darned First Law of Motion. It has, so far, kept anyone from designing the perfect flat, fast, and (inherently) in-phase single driver capable of full audio bandwidth.
I suppose that is why guitarists can be so much happier with their speakers-- a la a D130 or Celestion Greenback vs. all the rest. That is because we only care about three or four octaves; which is more than the chords some players know. + we expect DISTORTION and often demand more! Count me as a Celestion fan because their distortion is available in greater amounts and varieties (though SWR had to order a "special" [clean] version for their acoustic guitar amps) as catalog items. Stick 'em in a box with five or six sides, plug into your amp of choice, and play! It ain't no big deal. Want it louder? Just add more speakers and boxes and amps. If the sound sucks, it is probably not the speaker's fault.
Clean and full range speakers are for keyboardists and PA guys to worry about. Home audio enthusiasts, and especially Audiophiles, suffer from the same afflictions when they fail to achieve sonic Nirvana.
Martin W.
Wizard Labs

[martinleewin]

Well, if you're going to go by the INSTRUCTIONS (to find whatever works)... should that be a bass or tenor male vocalist?

As far as "Utilize the phase condition which yields the most natural reproduction."... I thought this board was G-rated! Did Jim Lansing study at the same school as Masters & Johnson and Dr. Ruth? Sorry, I misread "condition" as "position". Never mind.
They call me Mellow Yellow,
Martin