Great progress. How does the CD with holes for Delatlites compare with CD horn before you cut the holes? I wonder what the large dips up to 300 Hz are - room modes or rear wall cancellation? Are your speakers pushed up against the back wall as shown in photos?
Nice work!
Nice work!
The comp sounds very similar in the horn, I think it also measures quite the same. Maybe I can try to seal the holes with couple of layers packing tape and measure before/after...Great progress. How does the CD with holes for Delatlites compare with CD horn before you cut the holes? I wonder what the large dips up to 300 Hz are - room modes or rear wall cancellation? Are your speakers pushed up against the back wall as shown in photos?
Nice work!
Yes the speakers are pushed against the wall. Rear wall reflection with the horn's mouth at ~1m away from the wall, should cause null to somewhere around 165-175Hz (1/2wl from mouth to wall) - just like seen in the response? (If the horn would direct those frequencies to listener in much greater extent, the calcellation would not occur.)
Some pictures from yesterday:
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
I love how the Deltalites look hanging off the sides of the horn. It is almost a crime to hide them in a rear chamber box. Why not go open back as originally thought for a cardioid effect?
Yes they look nice, I have to listen to both options.
I have listened to them some more this whole morning and I'm amazed how well the two bands - as-is, without xo or other signal prosessing - integrate to each other! They sounds like one radiator even if you listen to them head inside the horn 😀. The sounds just floats in the horn in front of you, one can't pin point it's originating location. The feeling is quite the same if you listen to door sized full range electrostatic speaker right in front of it.
The mic is Galaxy S4's, but you get the idea.
https://www.youtube.com/watch?v=Q2Ttmx9e2R8
It's hard to stop listening to them and do some work regarding the back chambers 😀.
Here's the freq response with the Deltalite ports open (lighter trace) and closed with a packaging tape (darker trace). Open ports seem to worsen the 3kHz region resonance, everything else is pretty much the same. I cannot hear big differences either way with music.
Smoothing the comp's throat transition and rounding the inner edges of the ports should make some difference.
Smoothing the comp's throat transition and rounding the inner edges of the ports should make some difference.
An externally hosted image should be here but it was not working when we last tested it.
That is neat when they integrate this flawlessly - it was just meant to be. 🙂
Just think how much better they will sound with proper XO. I had a similar experience when I couple a PRV 5MR450 with a six driver slot loaded OB. The natural XO of 340 Hz due to OB and woofer cutoff magically worked.
That is a great result on the holes not affecting the CD response that much. Very nice. Smoothing the transition should get rid of a lot of those resonances and improve the HF reach I think.
Last edited:
Hmmm, this is the FR of HF compression driver, right? (only down to 500 Hz or so)
When doing this measurement, are those Deltalite open or shorted?
When doing this measurement, are those Deltalite open or shorted?
I listened to Philip Glass's Solo Piano and different piano notes did not jump out/resonate, which is a very good sign. Piano is one of the best instruments to detect resonances since it's spectrum is quite pure. The horn sounds mostly uncolored to my ears, and it outputs uniform balance (to my ears) for almost it's whole coverage. The balance does not chamge much if listened off-axis at different angles up or down, until the listener is near the edge of the horn's coverage.
I also think it should reach a bit lower, they went to 400Hz in the preliminary test when the horns were just the frames. Deltalites were open circuit in both cases. They should act as a resonators only near their Fs, right(?), but I will see what shorting does. I will also measure the response at horn's mouth to rule out the possible room interaction.
They work better with back chambers, unfortunately. Even more clearer and more bass. With back chambers they can be driven without any electrical crossover, except some low highpass to protect the low xmax cones. The front chamber gives very good acoustical low pass for my needs.
Measurement distance about 60 centimeters.
Solo Deltalites:
Deltalites + 2446J:
A bit funny thing that ear does not protest the 3kHz resonance. It spikes somewhat differently depending on the measurement angle. Maybe the polar response evens things out for the ear.
Measurement distance about 60 centimeters.
Solo Deltalites:
An externally hosted image should be here but it was not working when we last tested it.
Deltalites + 2446J:
An externally hosted image should be here but it was not working when we last tested it.
A bit funny thing that ear does not protest the 3kHz resonance. It spikes somewhat differently depending on the measurement angle. Maybe the polar response evens things out for the ear.
The 3 kHz peak is more than 10 dB, which should sound twice as loud, especially in that region, where hearing sensitivity normally is near it's highest. You might want to check "that ear" with a decent set of headphones and this test:
Free hearing test on line – Equal loudness contours and audiometry
Noise induced hearing loss is usually greatest in the 4 kHz range, I wish I would have used hearing protection back in the good old days...
As well as smoothing transitions, you may want to experiment with some perforated screen over the 15" exits, it probably would reduce the upper 15" response (good for reducing breakup and THD) while smoothing the 2446 peaks a bit. That said, fixing the throat transition will probably make a lot more difference than perforated screen will.
If you can approximate the Hughes Quadratic Throat transition (basically a simplification of Earl Geddes OS throat) the midrange should smooth out quite a bit.
Art
Attachments
Last edited:
This is an edit to my last post, after re-visiting some of the prior posts:
As well as smoothing transitions, you may want to experiment with some perforated screen over the 15" exits, it probably would reduce the upper 15" response (good for reducing breakup and THD) while smoothing the 2446 peaks a bit. The difference the port holes made is pretty large, as can be seen comparing the response prior to the holes in post 171 to the ones recently with the packing tape (which still shows diaphragmatic problems). Moving the port holes to the corners and elongating them as in post 194 would be best, but the perforated screen would be an easy test.
If you can approximate the Hughes Quadratic Throat transition (basically a simplification of Earl Geddes OS throat) the midrange should smooth out nicely.
Art
As well as smoothing transitions, you may want to experiment with some perforated screen over the 15" exits, it probably would reduce the upper 15" response (good for reducing breakup and THD) while smoothing the 2446 peaks a bit. The difference the port holes made is pretty large, as can be seen comparing the response prior to the holes in post 171 to the ones recently with the packing tape (which still shows diaphragmatic problems). Moving the port holes to the corners and elongating them as in post 194 would be best, but the perforated screen would be an easy test.
If you can approximate the Hughes Quadratic Throat transition (basically a simplification of Earl Geddes OS throat) the midrange should smooth out nicely.
Art
Last edited:
You want the mids to have an Fb of about 500hz (sealed) or an FS of about 500hz (open)
There aren't too many midranges with an FS of 500hz, so in order to go "open back", you really have to use some ultra small midranges.
That's the tricky thing with dipole synergy horns; there aren't many mids that will work properly, and the smaller the better
Agreed.
Here's some tables and illustrations on how to approximate oblate spheroidal:
http://www.diyaudio.com/forums/multi-way/257159-how-reduce-diffraction-your-conical-horn.html
Where do you get the 500Hz from? The tuning should be near the horn's lower cutoff around 100Hz, not 500Hz (?).
You can only get about two octaves out of each driver. In an SH-50 that looks like this:
- tweeter: 1.5khz - 6khz
- midrange: 375hz - 1.5khz
- woofer : 94hz - 375hz
The tweeter goes an extra octave and a half due to the phase plug. If you used a phase plug on the midrange or the woofer you could widen the bandwidth. I tried that once, it works. The woofer goes lower than the expected bandwidth because the woofer is basically in a dual reflex bandpass box. Basically you get wider bandwidth than you would if you just utilized the radiation from the front of the cone. You can only do this trick with the woofers, not the midranges. (If you did it with the midranges the radiation from the midranges would screw up the woofers.)
If you used an open back midbass with an FS of 100hz I would anticipate a bandwidth of approximately 75hz - 300hz. If you're seeing output above 300hz there's a pretty good chance you're measuring the radiation from the *back* not the *front* of the cone. To confirm, cover the back with wool or fiberglass and re-measure.
Last edited:
Hi Patrick Bateman,
Looks like Legis isn't building a Synergy horn, but just using the basic shape of a 4-sided pyramid for a conical horn. He is only using two woofers and one compression horn driver (I think?).
Regards,
Looks like Legis isn't building a Synergy horn, but just using the basic shape of a 4-sided pyramid for a conical horn. He is only using two woofers and one compression horn driver (I think?).
Regards,
Last edited:
Well it is a 2 way synergy sort of. With the sealed backs on the woofers the sim predicted a bandwidth of about 100 Hz to 1kHz. Nearly 3.5 octaves but not sure if that is panning out as it looks like the woofers are falling off at 600Hz. When Legis started out the idea was just a mid bass to bass V horn. The CD idea and pyramid came later. He works really fast - one of the few builders who builds almost as fast as I can model in AkAbak. 🙂 the results look very promising so far and I commend Legis on a effort on such a grand scale.
Last edited:
Not "sort of" it is a two way Synergy horn.Danley Sound Labs makes plenty of two way Synergy horns, I have been promoting them for some time, as they simplify the crossover considerably.
The response curve prior to the 15" driver holes in post 171 show very little difference to "post hole", I had not noticed the scales were 10 dB in the prior test, and only 5 dB in the latter test. With a little throat smoothing (and possibly perforated metal grill over the 15" ports) and some basic EQ (passive or DSP), Legis' build will be as near as smooth as any Synergy built thus far.
Given how much effort went into slots in the corners of the horn and angled frustums, it is surprising how well the big 11cm holes right near the throat turned out. Is it just luck or maybe using the big JBL compression driver is not as susceptible to near throat imperfections as a 1 in CD?
When I said sort of a Synergy I meant that it doesn't have the standard double angle expansion but is just straight and I have never seen a Synergy use a 15in as a mid range before.
I had proposed a 2 way with a 6.5in woofer to cover a similar range before but did not realize putting the port near the throat could work without messing up the CD output.
When I said sort of a Synergy I meant that it doesn't have the standard double angle expansion but is just straight and I have never seen a Synergy use a 15in as a mid range before.
I had proposed a 2 way with a 6.5in woofer to cover a similar range before but did not realize putting the port near the throat could work without messing up the CD output.
- Status
- Not open for further replies.