@SpeakerBob do you have more information about a multiple entry horn with 4x4" mids ? a link to a thread, or document? Building a MEH is outside the scope of this project (we'd like ideally to have something working for a spring event, and we have limited funds), but i'm very interested in them and would like to investigate further these designs.
There's a couple of pics of my 60x40 horn coming together in this thread.
This SPL Runt thread has good info
The SynTripP by Art Welter on this forum has a huge amount of build / design info as well.
Have a search for Bill Waslos spreadsheet calculator for MEH speakers. He has done all the hard work for us.
Rob.
This SPL Runt thread has good info
The SynTripP by Art Welter on this forum has a huge amount of build / design info as well.
Have a search for Bill Waslos spreadsheet calculator for MEH speakers. He has done all the hard work for us.
Rob.
So far the proposed horns have 100+ ° horizontal dispersion. A MEH with 100° dispersion and pattern control down to 520Hz would be at least 1,5m wide. It will also be very heavy. If such a wide dispersion is really needed, a MEH is probably not an option.
The original post did not specify dispersion requirements. The spreadsheet I referred to calculates a 42cm width for 100x60 MEH with horizontal control to 518Hz. Maybe its wrong ?
Rob.
Rob.
The JBL 2380 really has only slightly more than 90° dispersion, 100° +/- 23°.
The SynTripP without the secondary horn extension is only 26.5" (67.3 centimeters) wide, it maintains a -6dB 90° horizontal pattern to 400 Hz.
The JBL 2446/2380 weigh 16 kilos, (36.5 pounds) without a box.
The SynTripP weighs just under 15.4 kilos (34 pounds) complete with a box.
A MEH along the lines of the SynTripP could substitute a pair of 4" for the 10" and still have far more output in the 500Hz range than a 4" diaphragm driver like the 2446.
Art
That dispersion is horrible! Why build that? The 15-20dB valley between 3 and 4kHz under angles is bad for even 80s speakers.The SynTripP without the secondary horn extension is only 26.5" (67.3 centimeters) wide, it maintains a -6dB 90° horizontal pattern to 400 Hz.
The B&C DE750TN weighs just 6.1 kg.
I personally would xo at 6-700Hz but that's not the point. The SynTripP has with the 10" more output at 500Hz but the dispersion will be even worse in the midrange and would add another 10kg (or more). I really doubt it will perform better in distortion in the midrange or decay. It gets a lot more expensive, you need all 6 ch of the DCX even without having subwoofers, you need more amp channels. So what advantages justify those drawbacks?
I would agree that the dispersion pattern of a conical horn is not as consistent as other designs.
Straight sided conical horns are easy to build, and sound decent, and don't "beam" like exponential horns. The straight sides make it easy to mount mid drivers, which makes a virtual point source easy to achieve.
You are exaggerating that polar deviation by a considerable amount.
If the polar response was perfectly consistent, the response would be -6dB at 42.5 degrees off axis compared to the "FPPO" (0 degree) response.
The worst case at ~3.5kHz is ~-12dB, the pattern has narrowed.
Although even response is preferable, having an off-axis dip in the response where hearing should be most sensitive is not the worst.
Personally, I find the response being +4dB 25 degrees off axis (wider dispersion) is a greater defect.
Yes, that 3" diaphragm driver is far lighter than the 4" JBL 2446 the OP is considering.
The 4" diaphragm is capable of around 4.7dB more (relatively) clean output at 600Hz.
I would say a 3" diaphragm on a horn like the JBL 2380 is pushing it's limits at 600Hz.
I personally find the "open" sound of a conical compared to a diffraction horn compensates for the dispersion defects.
The cone midrange allows for far less THD or IMD than a 3" or 4" diaphragm at 500-700Hz.
Yes, adding one more crossover point adds expense.
That said, a MEH can be made passive, negating that expense.
Anyway, I'm not trying to "justify" alternatives, just was explaining that a 1.5 meter wide, heavy MEH was not required for use above a 2x15".
Art
That's not the conical expansion, compare it to non-MEH conical horns/WGs, it seems to be an inherited trait of a MEH. The main problem seems to be the distance of the entries to each other since that creates a runtime distance difference, or in other words, a phase shift that can't be fixed, neither per electronics (dsp) nor per construction - it's the horns builds principle, after all. The runtime difference gets worse the farther apart the entries get, and in turn, of course, with the increase of dispersion angle. The logical solution to reduce the effect is to try to get the drivers entries as close as possible to the throat. Since the 'real estate' there is scarce, the geometry of the horn has to follow the mechanical properties of the drivers as well as the acoustical ones. Not to forget, the entries also worsen the dispersion of the 'earlier' drivers. That lead me to the conclusion, you cannot get both to an acceptable dispersion width and constant directivity and positioning wasn't mechanically possible with a such small horn, it has to be quite a lot bigger.
Straight sided conical horns are easy to build, and sound decent, and don't "beam" like exponential horns. The straight sides make it easy to mount mid drivers, which makes a virtual point source easy to achieve.
You are exaggerating that polar deviation by a considerable amount.
I wasn't exaggerating, but you are right, I've made a serious mistake when I was switching between different measurements of several MEHs, which were 5dB/div, not 3dB. I'm sorry for that, that should not have happened. Still horrible, 13dB and an extremely uneven pattern.
About the "beam".. Please look at this:
Look at how extremely even the beamwidth is and even where it isn't as controlled, hof softly it changes. That's magnitudes better above the measurements you've shown. So that's exactly the opposite of what the claim is. And yes, that's the jbl 2380.
Why avoid ie. diffraction horns like the devil the holy water if you insert your construction with built-in interferences which are way worse in linearity, constancy and dispersion? You cannnot talk this smooth in any way. You can get better dispersion for 10-20 bucks from horns off ebay.
You are right, the 4 dB peak is way more offensive to the ear, I completely agree! A dip is always better to listen at than a peak. But: The peak only appears at one narrow angle while the ridge goes over almost all used dispersion angle. And you can't even it out by EQing it because then you'll get a huge bump on axis!
Then use the B&C DE1095TN, that got a 4" VC and weighs even much less, just 2,14kg. Still cheaper than the MEH drivers (below 400€).
Yes, at 500-700. I proposed a higher xo, so that doesn't apply. Moreover, the problem is the distortion in the midrange, not the low-mids. And with 115dB/1W the B&C DE1095TN makes 10dB more at 1W (not 2,83V!) at 700Hz,. The distortion is mostly H2 while the cones produce on the upper end a lot more H3&H5. And then look at the decay. In the midrange, the 2" driver looks excellent, that can't be said about the MEH (please post a waterfall decay of the MEH).
No, it doesn't negate that expense. The parts (exp. chokes) became incredibly expensive, it cost more than the addition of more channels. Plus, you can't delay the different ways to align the phase, which dictates the position of the entries, even if it cripples the dispersion more than otherwise a MEH needed.
The phase shift of the SynTripP's low frequency and high frequency drivers combined with no DSP used at all (raw response) is under 90 degrees from 250Hz up.
The phase response could easily be "fixed" using FIR DSP.
I never made any claim that the SynTripP had better polar response than the JBL 2380 Bi Radial diffraction throat horn.
That said, "magnitudes" would be 20dB or more, while the worst deviations are only ~+3 and -6 in two 1/3 octave regions over a range of more than 5 octaves, as can more easily be seen in this chart using the orange on axis trace -6dB compared to the 42.5 degree off axis trace.
Anyway, I'm very much aware of the many defects in the 10 year old SynTripP design, and have recommended other designs with better dispersion and higher output to many in the years since I posted the plans.
I've never seen a waterfall decay of the JBL 2446/2380 combination, and never made one for the SynTripP cabinets, which I sold around 5 years ago.
I'd certainly prefer the lighter B&C DE1095TN over the old JBL 2446, but the OP is making that choice and has limited funds.
Art
The JBL 2446 are used on eBay ~110€ per piece more expensive than the B&C DE1095TN NEW! Anyway, the MEH isn't cheaper either.
That doesn't look raw, that looks better than a driver alone and the fr suggests there are phase problems. I'd like to see the crossover plan.
The JBL isn't very impressive in the waterfall, I'd prefer the B&C any time. The JBL were used ofc, maybe they had a rough life or aftermarket diaphragms, who knows. Anything with the JBL label (vintage or not but vintage even more) sells for crazy money.
Why did you sell them if they are that good?
interesting discussion but obviously way beyond my knowledge, i'll try to learn something from it.
@weltersys what other newer designs are you refering to?
@ICG as for the jbl, i found them at a good price... I'd prefer to go for newer drivers, but the b&c is over budget at 400e a piece.
@weltersys what other newer designs are you refering to?
@ICG as for the jbl, i found them at a good price... I'd prefer to go for newer drivers, but the b&c is over budget at 400e a piece.
Peter Morris PM90 is a good performer, it uses the RCF HF950 horn (1.4" entrance).
https://soundforums.net/community/threads/new-diy-mid-high-90deg-aka-pm90.11601/
The larger B&C ME464 horn also uses a 1.4" entrance, has quite good polar response and can be used as low as 400 Hz.
The STPLF+HF magnitude and phase graph were of the raw 10" low frequency and 3" diaphragm HF drivers driven on the SynTripP horn with it secondary horn extension laying horizontal with no crossover, measured on axis, ground plane on October 30, 2014.
You can (barely..) see the DriveRack PA DSP settings used in the polar plots in post #9 of the thread:
I had retired from the live sound business and was not using them.
Art
On this picture
I can see the actual phase. And the actual FR of the drivers. So your upper picture is actually the dsp corrected one, nothing of it is raw, just as I suspected.
I should have marked the sarcasm.
The RCF HF950 is a diffraction horn too. I wonder why you deemed diffraction horns to be that horrible, but the PM90 is suddenly performing so well. I think the anti-diffraction is a widespread prejudice, well cared and repeated in the forums.
Don't know which "upper picture" you are writing about.
The polar charts did use the DSP mentioned, but others did not.
The montage you posted in #35 of three separate measurements of the LF and HF (no crossover, no DSP) raw response at four meters height is a different measurement from the upper one in post #30, the combined LF & HF raw response of the cabinet laying horizontal, measured on axis, ground plane on October 30, 2014.
I'm aware the RCF HF950 is a diffraction horn. The PM90, designed ~ a year after the SynTripP, has better upper polar response and far more (almost an order of magnitude..) output potential.
I have never deemed diffraction horns to be "horrible", nor have deemed MEH to be without issues.
After owning and listening to literally hundreds of diffraction horns, (many of which I built), I have judged them on their merits for the particular application.
That said, I prefer avoiding diffraction if the goals for the application do not require it.
Cheers,
Art
Last edited:
The first one in Post #30
Yes, exactly! Combining truncated measurements is everything BUT RAW! As in, the polar opposite! I was instantly extremely suspicious and I was right.
You can't snippet it together because the diffent measurements do not contain the phase differences to each other anymore! That's basic principle #1 for measuring phase! The same goes for the fr, that's not raw either. Both suggest practically perfect phase, no need for a delay and perfect conditions for a passive crossover - which is both dishonest and plain wrong!
The PM90 is a lot better but got issues too but that's not the topic here.
I appreciate that. To be true to that, what's the advantages of the MEH again over the diffraction horn? Or a 2x10" top (horn loaded or not)? I mean, the headroom argument already proved non existent, mind you.