Something stirs in the Undergrowth

[IslandPink]

Here's the 288H and Yuichi horn driven by the 45SE amp instead of the chip amp.
Not much difference.

Yuichi415_20cm_45SE_Dist.jpg (109.85 KiB) Viewed 15901 times

Oh great, another page throw. every time I post it throws a page !

Edit : after a bit of looking going back and forth across the annoying page break, I think there is a bit of cancellation of 2nd-harmonic : with the 45 SE amp there's less 2nd-harmonic overall.

[chris661]

Mark,

The high distortion across the board worries me. I went through my files and dug out the response & distortion curves for an 18Sound ND1460 on an RCF HF94 horn. They're decent components, and were being tested outdoors at "reasonable" levels from a Powersoft T604 amplifier. Distortion was at -50dB above 800Hz.

Chris

[IslandPink]

Yes, so I measured the Radian 745 Neo Be , I used the Azurahorn AH425 .

Azura745Be_20cm_Dist.jpg (96.9 KiB) Viewed 15857 times

That's much better . It has some 2nd-harmonic, but some of that could be from the amp ( again with 45SE ) .

Here's the frequency response close to the horn :

Azura745Be_20cm_SPL2smoothed.jpg (63.92 KiB) Viewed 15857 times

[IslandPink]

There's an odd and unexpected dip at 2.5kHz. I wasn't expecting this from the driver, so I put the microphone at 1m for another test :
The dip at 2.5kHz has more or less gone, but some other room-related ripples come in.

Azura745Be_1m_SPL2smoothed.jpg (64.71 KiB) Viewed 15856 times

[vinylnvalves]

Mark- what’s the conclusion the Altec’s are knackered? If they are as bad as measured they should not have been listenable too.

[chris661]

Yes, so I measured the Radian 745 Neo Be , I used the Azurahorn AH425 .

Looks more like it.
I'd suggest the other compression drivers need some attention, then. Might be as simple as the alignment being slightly out.

Chris

[vinylnvalves]

I don’t generally look at the distortion tab.. here’s one from a freq sweep of my midrange AMT. Ignore the low freq end that’s not real.

1CDF7B0F-4E5B-4D78-B0B4-461A90E4E594.jpeg (123.23 KiB) Viewed 15830 times

[chris661]

Looks like you've got noise floor above the distortion. Might be worth increasing the drive level a bit, but a good result nonetheless.

Chris

[IslandPink]

Mark- what’s the conclusion the Altec’s are knackered? If they are as bad as measured they should not have been listenable too.

I don't think so, but I haven't tested the other one. I think that's next on the list.
I must admit I'm thoroughly confused.
The 288's have always sounded better in the sub-1kHz range, more colourful and dynamic. the Radians were more refined and had better tone higher up.
The 288's on the Yuichi seemed to be the best compromise. I love the sound on guitars and percussion, usually with the ribbon there too. I did have great results on female vocals at times too. Whether something had changed, i don't know. The whole OB did fall over the other day, but I didn't think anything other than my right ankle had been damaged !
I think in the test the driver is playing a bit louder than I would listen, so the distortion levels could come down.
If you look at the phase, it starts turning above 1kHz with the Radian, the 288 stays flatter a bit lower, maybe this is the origin of the better colour and snap lower down.

I wonder about the tangerine phase plug on the 288H. There were mixed opinions about this on Lynn's Beyond the Ariel thread years ago. The tangerine is undiscriminating about where it passes diaphragm output. The normal circumferential plugs like on the 288C and the Radian are designed to be more selective about where to load or transmit the output. Maybe that is the source of the 3rd and 5th harmonic.
Anyway - test the other driver first.

[chris661]

Well, in the passband the distortion was still -30dB to -40dB, so not particularly easy to detect in listening. Measuring the other 288 will be informative.

Chris

[Retsel]

Just a couple more comments. I researched what Rms is, as I had not heard anything about that. It is tough, for me anyways, to intuitively figure out how that particular set of parameters could cause better microdynamics or better sound at lower sound levels. But it is interesting to learn about a new theory to be aware of. My DX4s have a calculated Rms of 0.76, so it would fall within the good window. Interestingly, my DX65s have an Rms of about 1.0.

I looked at using radial horns, of which I think Yuchi is a type of, but came away suspecting that they are not optimum. Clearly they can result in a smaller vertical and horizontal footprint, but their mouths are undersized for the low frequencies that they reproduce. This is usually associated with resonances in the lower midrange, where the horn is not long or large enough to support those frequencies. These horns would not maintain pattern control to a very low frequency in the vertical dimension (maybe 1000 hz?). I think that one of the most desirable quality of horns/waveguides for home use, where efficiency is not much of a consideration, is the pattern control, to remove the room as a factor to a low a frequency as possible. Most radial horns are not designed with a roundover at the mouth, which can also cause resonances, HOMs. One thing to try with your horns is to wrap the outside of the horn mouth with some type of sound absorbing fabric, which can help present the sound from reacting to the hard edges of the mouth of your Yuchi horns.

I remember seeing a video of a guy who demonstrated the effect of reducing horn mouth resonances - he spoke through a magazine curled up into a horn-like shape, and then wrapped some fabric at the end of the mouth of that shape and spoke through it again - the improvement in intelligibility was quite obvious, even on you tube. What I intend on doing when I make my final front waveguides is not only create a roundover for the mouth, but also cover the roundover with fabric to further reduce the resonance at the horn/waveguide mouth. Right now my waveguides are about 3 feet long with a very large mouth 2 feet by 3 1/2 feet, but with no roundover nor fabric, to help reduce resonances. My final waveguides will not be that long, but will employ those improvements in the horn mouth.

The supposed advantage of Lowthers, or other such widerange drivers, is the ability to cover many octaves with a single driver. It simplifies the speaker by avoiding a crossover and reduces the issues associated with different drivers covering the full sound spectrum, but being placed at some distance from each other. You really need a deep room to allow the image of horns to come together. But these efficient drivers like Lowthers do utilize breakup in the upper frequencies to reproduce the upper frequencies, which creates problems there. These companies utilize different methods to deal with the issues to better reproduce the higher frequencies, at different levels of success. I don't know if I will end up relying on the Lowthers for the higher frequencies - they do sound quite good on a lot of music despite their problems at higher frequencies. I guess this is one reason why some people go the Unity/Synergy route, to have that coherent sound without resorting to a single driver. If not utilizing the whizzer or other means to reproduce the higher frequencies for midrange, only, you don't need to spend the extra bucks for Lowthers. But the midrange of Lowthers is to die for....

[Scottmoose]

I have news for you, which may provide some interest / perspective. Most soft-dome tweeters rely on a degree of controlled resonance to supply the upper audible frequency range, while all soft-material ring radiators are in resonant mode over their entire bandwidth -it's how they work.

[IslandPink]

Funny, I was talking to Colin about the same subject recently, Scott. I said I have never really liked standard dome tweeters since I heard a good ribbon ; however the beryllium ones he's used recently are a different matter - the first time he demoed them at Owston I knew they were a significant improvement. I suspect they are pistonic up to past 10Hz ?

[IslandPink]

Here's the distortion of the other 288H on the other Yuichi horn. About the same output level, though I wasn't able to control it exactly.
It's not any better, but on the other hand it's not the same either - quite a few differences in the levels of the different orders.

Yuichi415_II_20cm_45SE_Dist.jpg (105.37 KiB) Viewed 15753 times

However when you look at the frequency response, this one is not as good - has a slight level of roll-off in the upper frequencies compared to the first one. So maybe not surprising re. distortion.

Yuichi415_II_20cm_Average1.jpg (69.73 KiB) Viewed 15753 times

One of these drivers was returned to GPA since it had poor FR higher up. It was found to have defective material in the pole piece. Makes me think that perhaps both should have been returned !
However, given that most of the roll-off could be fixed by bringing the tweeter in a little bit lower, it should be fixable in any system.
If the overall efficiency isn't the same either, then likewise a bit less 'padding' would be easy, the horn will be way more efficient than any lower-mid cone.

[chris661]

Funny, I was talking to Colin about the same subject recently, Scott. I said I have never really liked standard dome tweeters since I heard a good ribbon ; however the beryllium ones he's used recently are a different matter - the first time he demoed them at Owston I knew they were a significant improvement. I suspect they are pistonic up to past 10Hz ?

Yeah, beryllium domes will make it well past 20kHz before breakup. I think those particular HF units also have other high-tech features (copper in the motor, linear suspension, back chamber, etc) that help them sound great.

That said, the good aluminium domes will typically make it past 20kHz, too. I don't think I've seen those at Owston, though. The Behringer B2031P speakers that I had for a while did have Al domes that had a breakup peak ~27kHz IIRC. Always sounded fine to me, and Linkwitz liked them in the active version.

Chris