So why not interpose more coils with taps throughout the system to further enhance the "life of the music"? This comment implies that resistors remove said lives, so I guess we should remove all non wire wounds from the system?
Welll there are different choices that designers. In Lynn's case he is not put off by well designed transformers. If you look at his amp preamp designs they use interstage transformers instead of caps as Lynn feel they are more detrimental to the sound. He also places a large premium on linear devices in the signal path
It would seem that using a autoformer instead of caps and resistors when possible is in keeping with other design choices he makes.
Ron, Actually I think right on topic. Today I rigged up TAD2001 on 340Hz Le Cleach horns with just a 4uF cap in series, as a tweeter above my mid horns, and was surprised at how much an overall improvement this was from using a conventional horn tweeter (Fostex T900A). Not a difference in the HF tonality as one usually perceives tweeter swaps, but in the overall dimensionality and scale of the music. Also I found the whole room now contains a more uniform live feel.
I quite agree with Lynn that a 1" CD just doesn't cut the mustard if you want it to do 700 - 20K Hz - but give it a go as a tweeter with a much larger horn than necessary and you might find it really does something quite special.
Of course the point of the expensive 1.4" Be units is to do just this in one go, but for folks with more pedestrian CDs in need of a tweeter I am convinced this is worth trying. I guess it is the same kind of idea as the wide band RAAL ribbon tweeter.
Technically I believe excursion still increases below cut off with a 1st order cross over, so warrantee void - but I guess they survive because of minuscule power in home use.
martin
Using a nom de plume is an old habit from my consulting days when market reality meant that my client's decisions often did not reflect the technical 'best choice'.
It remains my default choice when signing up to a new forum or support site.
I've found transformer based volume controls to sound significantly more vital and enjoyable than resistive solutions, line-level. Cost controls are fine unless you're willing to pay the incremental costs of meaningfully improved performance.
The "why"? Ground behavior maybe? Limitation on out of band noise? Dunno but there are good reasons both technical and subjective to go with iron.
Off the subject a bit you guys. I need a bit of advice.
When designing a crossover for the Radian 165 (high-pass) see this graph. Radian High Frequency Compression Drivers - Radian 465PB - Radian 465PB 1" high frequency compression driver. Radian 465PB 1" high frequency compression driver is available here. Radian 465PB speaker components.
Is it bad practice to implement a 1st order at around 6000 Hz and be done? I know there's a big loss in efficiency but I only want around 98db anyway and the driver should only see a watt or three.
Thanks, Ron
I wouldn't feel good about that. I may not know as much about horns as other posters, but I do know that below horn cutoff most of the diaphragm-loading benefits disappear, and excursion increases very rapidly.
A 1st-order crossover would be (just) barely tolerable with a conventional soft-dome tweeter, but compression drivers are not made for long excursions, and can strike the back of the phase plug if they move too much. Plastic (mylar) diaphragms may survive, but I'd be doubtful about aluminum. You'll hear really loud clicks if that happens, followed by silence.
There's an impact on IM distortion too. If the compression driver is not excursion-controlled below horn cutoff, there will be a lot of below-passband diaphragm motion that does not result in much sound, but pushes the diaphragm out of the linear region of the voice coil. This results in IM sidebands in the working region of the driver, and also throws away headroom.
I would not trust the published Radian curves. They look very heavily smoothed to me, and the test horn is not specified. I would measure the driver with the horn you intend to use, and design the crossover around that. The majority of horns require in-band equalization, and some also require one or more notch filters to remove significant (and audible) peaks.
If you're getting the idea that horn crossovers are more complex than crossovers for direct-radiator soft-dome tweeters that cover the same frequency range, you would be right.
Side note: the AH425 horn under discussion is extremely unusual in not requiring in-band (shelf) equalization. The tradeoff is that it is not a constant-directivity horn, and is rolled-off 30 to 45 degrees off-axis. (If the off-axis response response is equalized flat, the on-axis response will be strongly tipped-up.) This dictates a limited listening window, a supertweeter, or both.
The best of the constant-directivity horns have a well-matched set of response curves over the intended coverage pattern, but must be equalized flat with a passive or active crossover.
Last edited:
Thank you Lynn,
I figured there'd be a pale full of negetives somewhere but thought that maybe, just maybe, because I didn't care about efficiency it may work out. Yes that smoothed graph phenomena is like buying snake oil. I guess that's why there are forums like this.
And then there's the problem of working the horn/cd in with the mid driver.
It's like there could be an niche industry out there for someone to build cross-overs for specific CD/horn/mid combinations. Take several combinations that have proven brilliant, perfect them and offer the services.
I suppose that means going the Pi Speakers rout.
I'm sure your deadly against but what have you to say about DSP eq/xover?
Thanks again,
Ron
I figured there'd be a pale full of negetives somewhere but thought that maybe, just maybe, because I didn't care about efficiency it may work out. Yes that smoothed graph phenomena is like buying snake oil. I guess that's why there are forums like this.
And then there's the problem of working the horn/cd in with the mid driver.
It's like there could be an niche industry out there for someone to build cross-overs for specific CD/horn/mid combinations. Take several combinations that have proven brilliant, perfect them and offer the services.
I suppose that means going the Pi Speakers rout.
I'm sure your deadly against but what have you to say about DSP eq/xover?
Thanks again,
Ron
GoTo do a multiway horn system with active 1st order HP / LP crossovers, but then I guess they build their drivers to handle it. Low power, big excursion. They do their own active cross over unit so could perhaps include extra roll offs below the horn cut offs. I once built horns for a GoTo set up and tested it out here in house, with the GoTo cross over - it really came across as a very well integrated system. Emulating this with a normal CD (as I have just been doing) is fraught with expensive hazards and I shall probably desist. Though with 24dB down at the horn Fc I am probably OK. My mid horn is a LM555 on a 200/160Hz horn (WE copy) which is also set up 1st order HP, and rolls of naturally at 5K Hz. The 597a would be the natural complement.
With the LTO rig I am trying to stay as close to Lynn's spec as I can. I even have a transformer coupled pp amp newly completed. Interesting to compare the 3 way horn system in the light of discussions on here, but it is not by any means the same thing.
With the LTO rig I am trying to stay as close to Lynn's spec as I can. I even have a transformer coupled pp amp newly completed. Interesting to compare the 3 way horn system in the light of discussions on here, but it is not by any means the same thing.
... when used with the 745AL
http://www.diyaudio.com/forums/multi-way/100392-beyond-ariel-970.html#post3699913
green=Al, yellow=Be
Hello pos
I would say that the Be needs some EQ. I don't know why many think the Be drivers are more extended up top. They are much cleaner but tend to roll off earlier than there Al of Ti cousins. Even the Truextent Be replacements do that as well.
I don't understand the issue with in band passive EQ asside from simplification of the crossover. Works well and sounds just fine.
Rob
I made the measurements quoted by POS. The trouble with frequency response measurements is that although they tell us how much output there is at each frequency, they can't tell us how much of that output is spurious. Filling in the upper octaves with chaotic breakup and resonance might make the graph look flatter, but it doesn't result in a natural-sounding top end.
Despite the flatness of the AL/PB in comparison to the BE/ND, sonically it was no contest. The AL/PB was a bit opaque-sounding, and had what I would call a "grayish" coloration, with some tizz at the top. The BE/ND was more transparent. Tone colors of woodwinds and strings were quite natural and appeared against a more-silent background.
The BE/ND is definitely missing something at the top. Lynn and I both have Lazy Ribbons on hand. I certainly plan to use mine, but I am interested in trying the equalization approach as well. Actually, the unequalized BE/ND is a nice option to have. It sounds great with some recordings that are otherwise not so pleasant.
I couldn't hear the glitch at 3k, plain to see as it is on the graph.
Despite the flatness of the AL/PB in comparison to the BE/ND, sonically it was no contest. The AL/PB was a bit opaque-sounding, and had what I would call a "grayish" coloration, with some tizz at the top. The BE/ND was more transparent. Tone colors of woodwinds and strings were quite natural and appeared against a more-silent background.
The BE/ND is definitely missing something at the top. Lynn and I both have Lazy Ribbons on hand. I certainly plan to use mine, but I am interested in trying the equalization approach as well. Actually, the unequalized BE/ND is a nice option to have. It sounds great with some recordings that are otherwise not so pleasant.
I couldn't hear the glitch at 3k, plain to see as it is on the graph.
From an empiricists point of view, I'd be interested to see the same measurements with the diaphragms swapped, AL on ND and BE on PB. I have a sneaking suspicion that the 3k dip (not breakup, that was typo) is part of the ND body not the phragm.
I believe you about that sound, was just surprised by the measurements.
I believe you about that sound, was just surprised by the measurements.
For the record, I use a basically 1st order HP in my systems. I don't see a problem with it for home use. Compression drivers on horns have so much headroom in home use that the standard ideas (from Pro usage) on "damage control" simply don't apply. The compression driver is padded down by about 10 dB and then HP'd - resulting in at least 30-40 dB reduction below resonance The excursion is never an issue.