Amount of hornloading in a Synergy horn

Hi Chris, the 11cm measurement is a timed acoustic measurement from the 4594's mid section to the mounting flange, not simply a mechanical path length estimate.
Mark,

Looking back at measurements done on my SynTripP MEH (multiple entry horn), found that though the HF driver acoustical center is physically further from the measurement mic, it's output arrives .62ms before the BP (band pass) LF output. The combined raw output of the two 10" with the 3" diaphragm HF driver, no crossover used at all results in a relatively flat phase alignment as can be seen below.

It's even possible that Chris' MEH might eliminate or reduce the delay required by reversing the mid/high polarity- without doing the math (which I'd probably hose up :eek: ) it sounds like the path length plus the .5ms LF delay used may be close to one wavelength at the acoustic crossover point.

Art
 

Attachments

  • HF& LF arrival times.png
    HF& LF arrival times.png
    323.2 KB · Views: 283
  • SynTripP Raw Freq & Phase.png
    SynTripP Raw Freq & Phase.png
    171.4 KB · Views: 307
Last edited:
Hey guys, with all due respect, you are making something very simple, very murky.

Diyuser, yes finding a subs acoustic center is very difficult, or better said, relatively difficult to full range / HF drivers.
But it's not difficult in the least, not for a compression driver... not at all.

Patrick, no offense but what you're saying has nothing to do with finding acoustic centers..I don't understand...

Chris, 600Hz, group delay, your plot... ALL are totally immaterial to finding the acoustic path length of your driver.. Please, Just go measure it !!!
 
Last edited:
Hi ya Art,

Ok, what I've come to be dang certain about...;)
...acoustic path lengths are physical geometry...they don't vary due to polarity, crossover topology or order,... etc.
Period.

I think I know this because I've found the acoustic centers sooooo many times for various multi-way drivers/sections.
Finding acoustic centers is simple and uncomplicated....just use a full range signal.

People get confused because they try to find electro-acoustical centers, with a crossover (low pass) in place.
Those findings are not acoustic centers, and lead people to all kinds of alignment errors, because they conflict/trash the cornerstone of timing....which is acoustics.
 
There's no electrical crossover filter delay in the measurements posted for the 4592, above. No "named" crossover filters were used.

I'm honestly not sure what you're worried about.

What do you think phase response is? Group delay?

Chris
 
Last edited:
AES E-Library >> Effects of Acoustic Center Position in Subwoofers

While this paper is pertaining to subwoofers, the same concept applies. The actual acoustic center position is not an arbitrary location ( like the voice coil ) or diaphragm. Only acoustic measurement along with a full duplex sound card will accurately reveal the acoustic center in the far field.

Surprisingly in the paper, all of the methods used ( BEM simulation, spreadsheet prediction, near field scanning of the transducer, and an actual 2 microphone measurement ) placed the acoustic center almost a foot in front of the cabinet.

This can lead to errors in loudspeaker polar measurements if the center of rotation is not the actual acoustic center.

Another good read:

AES E-Library >> On the Movement of a Horn's Acoustic Center

Please accept my apologies diyuser2010 and Patrick,

A bit of wine last evening had me reading your posts far too fast for fair consideration :eek:

I'd love to read the linked AES papers, both look very interesting, particularly the surprising acoustic center in front of the sub ..maybe it's about time to look into membership...???

I always use dual channel FFT to find acoustic centers. and have found once properly located, the center can be 'marked on the box' so to speak.
Subs are always a pain to get repeatable measurements, but I think that is more a function of FFT's linear data spacing, and relatively fewer data points down low trying to find peak energy, than anything else.

That said, using FIR, I have found I way to keep distance variability down to a handful of processing samples. Measure a sub's distance with no LPF in place. Then using a FIR filter with a known number of samples delay, implement whatever LPF you like. The new distance to sub's acoustic center simply becomes very accurately, the first measurement plus the FIR sample delay.

Now, because of the paper you linked I'm going to have to get out the tape measure and see where the sub actually is vs measurement !
Thx :)
 
There's no electrical crossover filter delay in the measurements posted for the 4592, above. No "named" crossover filters were used.

I'm honestly not sure what you're worried about.

What do you think phase response is? Group delay? Chris

Hi Chris, here's my concern...

Everything I can gather from the unity/synergy patents, TD's posts, great threads like 'suitable mid-range', yours, nc535's, Patricks, Bill's, and many others..... ...all say to keep the maximum distance between the mid range ports and the HF acoustic center, within 1/4WL of the crossover frequency (due to the creation of an acoustic low pass filter near said wavelength).

Now I'm trying to do as you and a few others have done, and skip the mids, using a CD that goes low enough to mate with a woofer. But the same 1/4WL rule should apply, shouldn't it?

When I go to determining where to put the ports, by laying out the distance to the CD's acoustic center, it seems to me I need to know exactly where that distance begins (apex/HF acoustic center).

So, to even begin designing, I put a mic even with the mounting flange of my CD, and measure. I get a surprising 11cm. I repeat the measurement at greater distances, backing out tape measure mic to driver distances, and repeatedly get the same 11cm. (That's why I ask you to check yours.....)

When I look at 1/4WL for let's say a 475Hz crossover, I get 18cm port to apex distance to work with. I believe I need to subtract 11cm for the CD's path depth, plus another cm or so for a wooden mounting flange, for a total path of 12cm before even getting to the horn walls. Can this 12cm path be ignored or reduced, and in keeping with the 1/4WL rule?

Because if it can't, I'm down to 6cm of room to try and fit the woofers' ports.

It just doesn't look like it can work...and stay within the acoustic low pass design guidelines.

That's my concern.
 
Not that I dont appreciate all th interesting discussions, I do have another question about my topic.

Would there be acoustic coupling between the 2x 15"ers? Or how does this work in a synergy?

The ports would definitely be spaced more than 1/4th wavelength apart, but could maybe be between 1/2 wavelength so coupling between the drivers could still happen within the horn itself, before there is even horn coupling added. Right?
 
Everything I can gather from the unity/synergy patents, TD's posts, great threads like 'suitable mid-range', yours, nc535's, Patricks, Bill's, and many others.....
...all say to keep the maximum distance between the mid range ports and the HF acoustic center, within 1/4WL of the crossover frequency
(due to the creation of an acoustic low pass filter near said wavelength).

When I go to determining where to put the ports, by laying out the distance to the CD's acoustic center, it seems to me I need to know exactly where that distance begins (apex/HF acoustic center).

When I look at 1/4WL for let's say a 475Hz crossover, I get 18cm port to apex distance to work with.

Can this 12cm path be ignored or reduced, and in keeping with the 1/4WL rule?

Because if it can't, I'm down to 6cm of room to try and fit the woofers' ports.

It just doesn't look like it can work...and stay within the acoustic low pass design guidelines.

That's my concern.
Mark,

The relationship of the volume of the enclosed throat chamber, length and area of the cone entry ports determine it’s acoustical low pass. The distance of the cone entry ports from the throat of the horn determine a reflective null, a different concept.

There has been a lot of reference to keeping the multiple horn entry locations within 1/4 wavelength from each other at the crossover point, but the MEH output location and it’s acoustic center regarding time are two different things. Different HF drivers will have different depths, those depths can be time aligned to coincide with the low pass output of the cone drivers entry points. However, if the location of the multiple entry points are not within 1/4 wavelength, off-axis lobing will start to occur due to path (time) length differences, regardless of on axis time alignment.

Consider that the HF “plumbing” on MEH line arrays, or DSL’s Paraline adds somewhere between 8 and 20+ centimeters path length to a HF driver’s acoustical center, at much higher frequencies than you are considering, and perhaps your concern over a few centimeters will be reduced.

Art
 
1)Would there be acoustic coupling between the 2x 15"ers?
2)Or how does this work in a synergy?
3)The ports would definitely be spaced more than 1/4th wavelength apart, but could maybe be between 1/2 wavelength so coupling between the drivers could still happen within the horn itself, before there is even horn coupling added. Right?
Droco,

1) Yes.
2) Probably best to simulate your design in Hornresp to get a graphical representation of "how it works".
3) Axisymetric ports will result in a 6dB on axis increase over a single driver on axis, ports spaced more than 1/4 wavelength apart will reduce off axis response and smoothness.
What is limiting your port location to more than 1/4 wavelength?

Art
 
3) What is limiting? I do mean the distance between the ports of the seperate 15" woofers. Each woofer would be on 1 side (so symmetrical axis wise), so there would be too much distance between them I think. Not sure yet. My horns arrive this friday, so this will be one of the first things I'm gonna check.
 
Member
Joined 2007
Paid Member
In my case I just used a lower order electrical filter on the mids than the tweeter such that they acoustically summed in phase. The taps are within 1/4 lambda of the throat at the crossover frequency. Measurements of the on axis frequency response for both the mids and tweeter were done at the same point in physical space. Its possible the mids/tweeter are delayed by multiple cycles at the crossover point but I doubt it. I don't see how this is super complicated or the stuff about subs acoustic centres is applicable.

To me this is just ensuring that the tweeters -6dB point matches the -6dB point of the mids and the acoustic rolloff is the same order. In my case the tweeter was not rolling off at the crossover point and so needed a higher order filter. This means that the crossover point is not a free design choice and is constrained by the roll off of the mids. That is why I put a lot of effort into increasing this frequency. I believe I could not have increased it further without either the mas of the mids cone been lighter or the mids having a stronger motor.
 
Droco,

1) Yes.
2) Probably best to simulate your design in Hornresp to get a graphical representation of "how it works".
3) Axisymetric ports will result in a 6dB on axis increase over a single driver on axis, ports spaced more than 1/4 wavelength apart will reduce off axis response and smoothness.
What is limiting your port location to more than 1/4 wavelength?

Art

Also, bit of a problem... I wont be able to get the necessary specs I need on the horn to model it in hornresp. Any way to "measure" them myself?
 
Mark,

The relationship of the volume of the enclosed throat chamber, length and area of the cone entry ports determine it’s acoustical low pass. The distance of the cone entry ports from the throat of the horn determine a reflective null, a different concept.

There has been a lot of reference to keeping the multiple horn entry locations within 1/4 wavelength from each other at the crossover point, but the MEH output location and it’s acoustic center regarding time are two different things. Different HF drivers will have different depths, those depths can be time aligned to coincide with the low pass output of the cone drivers entry points. However, if the location of the multiple entry points are not within 1/4 wavelength, off-axis lobing will start to occur due to path (time) length differences, regardless of on axis time alignment.

Consider that the HF “plumbing” on MEH line arrays, or DSL’s Paraline adds somewhere between 8 and 20+ centimeters path length to a HF driver’s acoustical center, at much higher frequencies than you are considering, and perhaps your concern over a few centimeters will be reduced.

Art

Thank you Art !

Yes, I've been talking about the 1/4WL rule as if it were the low-pass determiner, instead of a notch determiner. My bad. I do understand the location of the ports relative to the horns cross sectional area and flare rate determines the acoustic low-pass, not the distance to horn apex.

The problem with the 4594 seems to be getting the ports out to the necessary cross sectional area for chosen xover freq, and also keeping the 1/4WL notch above the chosen xover freq.

The 11cm acoustic depth of the driver appears to eat up too much of the allowable 1/4WL length,... length that is needing getting out the horn to a sufficient cross sectional area.

I've been looking at a 60x60 horn for a start. If I went wider maybe the sufficient cross sectional area needed would occur close enough to apex to stay within 1/4WL. I think the k402 horn is pretty wide...maybe that's part of why it's working well?

Anyway, real question is: where do you think the horn's apex is ? Should I use the 11cm depth for modeling and layout?

I don't know what to think, or how to incorporate the paraline into my thinking...I'm not sure it applies to what's going on..it may be creating an acoustic image well in front of the drivers. When I said in a previous post that acoustic centers are a function of physical geometry, I should have said apparent acoustic centers. The point I was then trying to make is that apparent acoustic centers themselves don't move from polarity, crossover, etc. Sure, the way an acoustic center interacts with other acoustic centers changes, but the point-in-space it emanates sound from doesn't.

It seems to me the goal of a synergy/meh, is to get the acoustic sources playing together as nicely as possible acoustically before having to finish alignment with various crossovers, EQs, and delays, etc. I guess that's my fixation on knowing where to begin with the CD's acoustic center...

thx again :)
 
If you go active, which seems like the right thing to do with an all-out system like a Synergy horn anyway, then you wouldn't have to worry about the acoustic path of the 4594 (just delay the mids) and could just concentrate on getting the midrange ports right .
 
Member
Joined 2007
Paid Member
Regarding modelling a plastic horn I measured the width with respect to the axial distance in x and y. Used this to work out the internal area at that axial distance. Then plot the area vs. axial distance. This then clearly showed an initial conical flare and then an expansion which I fitted to an exponential flare. I then chose appropriate points to same these flares for the cross sectional areas in hornresp. It helps to use a spreadsheet.
 
Hi kipman725, InOtIn,

Yes, I figure i can use active to do almost whatever.
I'm just trying to accomplish as much acoustically as possible first. Not trying to overthink it..just trying to get the basics right.

if I build, it will be conical, or rather pyramid, so at least the cross sectional area will be easy to figure ...
Nice job kipman725, doing the work to model your plastic horn
 
Oh I see, so you use the tuning distance to figure out the distance relative to the apex? So for 120hz you would take a theoretical 71,5cm..? Looks like that would be to big a distance for the SH50..

Would this be that can be easily calculated, or would I have to go deep into hornresp to figure out how to work this? Modelling a simple vented box I know how to do. Could I just take the backchamber volume and use this to model the woofers in a vented vs sealed configuration?
 
In any case, for PA use configuring the horn to have the best output down to 100hz looks like the best option, as coupling the MEH with a sub @100hz (nicely within 1/4 WL)(for example an Othorn) is pretty easy and would ensure greater output down to 20-40hz without putting too much strain on the woofers.