MIDTOP 15" + 8" + 1" 3 way

Leurosound, some of your your facts about the Peter Morris midtop is slightly off.
Crossover is actually around 700Hz. The CD is a very effecient coaxial 1,4 capable of high outputs even at low frequencies, and the parallel walls are intentional and important for pattern control in the crossover region. Throat adaption is actually not critical and the front cavity and the slot shaped throat actually provides a natural sloping above 800Hz while maintaining low frequency output.
With that being said I do not consider the Peter Morris midtop to be the speaker to end all speakers. Peter says so himself; it's an optimized speaker on a stick, not the perfect speaker.
I'm happy to see someone having a go on their own. Would you care to share some of your thoughts on how you came up with this design?
It's a beautiful speaker BTW!
Best regards, Fred

I think the 900Hz crossover problem is confined to the very near field and (on-axis) summation should be generally good.

Crossing horns around lower cutoff often means wide or uncontrolled directivity causing audible integration problems off-axis. Your lowmid horn being large, coaxial and expanding in all directions might allow the upper mid to have slightly wider pattern though.

I Think playing around with slopes and crossovers with regards to both on - and off axis respons is critical in such a configuration. What's the coverage of the various horns?

I'm quit curious about your findings on the upper mid horn. It's quite unusual for a horn to be simply useless until almost an octave above cutoff. As it's mainly a matter of loading at this part of the horns operating range it's very dependent on the driver.

Best regards, Fred

Hi Leru

Have you seen this white paper of coaxial horn from JBL pro? There may be some useful information there regarding the crossover frequency and what kind of measuring results to expect. Page 5 onward.

https://www.jblpro.com/ProductAttachments/PD_HF_v1no29TechNote.pdf

and fine woodwork and execution of the idea by the way

best regards, Andy

15+8+1, always sound good on paper, then in large venue, under pressure, doesn't really make it.

@ Leru, with a 12dB BuutterworthLP on the 15" overlaying the 15's second order rolloff with a bit of EQ you may be able to achieve an LR24 LP, then use a LR24 HP on the 8" to complement it - may just work, directivity is always going to be a little off and not easy working with limited overlap of bandwidth.

In the end I switched to 6ND410, gaining a lot and getting a better energy balanced system.
Being probabily less finicky, I lowered the cutoff to 650 hz lower order filter without any H axis noticeable dip or problems. H axis stays very balanced, getting well linearly decreasing level by the spectrum in 45° H-offaxis (had 3dB scale in Smaart so guess it was around from 6 to 10dB relative difference from lower band to top-higher, still in good linear slope)
Vertical keeps staying with some unavoidable problems off axis, but with the sound 2,7m tall, floor stacked, from 2m to further listenings it's very good for the ears.
It sounds very loud and still quite well focused from 2 meter away to further. Stereo image is very good, but still not high pro level (me dumb as I made 1 baffle cutout of 15 larger, and I made passive xovers in different layouts for save time. While space was not too much and I did the best for lower inductor crosstalk, possible they inducts differently as the pattern of components positions is different.) Arrayed them behave really good, and there's almost 0 combing moving left to right. All of the results are good, fi sure more than good for the Soundsystem use I'm involving the boxes in.
But the time I started learning ABEC3, discovering some of the leakages of this design, so I stopped trying to improve as them are still very good to the ears.


Edited by LeruSound - 01 November 2017 at 10:53pm