Wow, I certainly appreciate the attention paid to this thread and the advice. Based on what I've read so far, I believe I'll try integrating the MG3A tweets into my design. If it sounds good, then great, if not I'll start over.
I think I've narrowed my driver chioces down to a few that I can afford for this. I just left Madisound's site and Peerless, Hivi Research, vifa and fostex all seem to offer what I'm looking for. The 3" Peerless drivers seem to be drawing me in, but I'm not that great at interperating T/S parameters yet, so I'm sort of blind right now. As before, any suggestions are appreciated.
I think I've narrowed my driver chioces down to a few that I can afford for this. I just left Madisound's site and Peerless, Hivi Research, vifa and fostex all seem to offer what I'm looking for. The 3" Peerless drivers seem to be drawing me in, but I'm not that great at interperating T/S parameters yet, so I'm sort of blind right now. As before, any suggestions are appreciated.
adason said:
and You? 😉
perhaps You can discuss line arrays and the problem of audibility of comb filtering with tryonziess, as he "probably" made one 😉
so, adason, will You? 😉
best,
graaf
Hi tryonziess,
this discussion is about "line array with coaxial drivers" as far as I know
if the fularange drivers are very small and placed as close as possible, than the issue is small
however, if coaxial drivers are used, tweeters for each individual driver will be forced to be spaced much further
in my opinion it woud be much better to simply remove the tweeters from the coaxial and make separate line array of tweeters next to midrange array
Roger Russell was building line arrays with separate tweeters tightly packed all his life. I do not know why he build his last one from fulrange drivers. Maybe he is loosing hearing at old age? /his array was criticized for poor bass response, of topic/ His array and Kunzes would both benefit from having one tweeter in the middle of an array, my opinion. I sure "experts" like "graaf" will not agree. I built once an array using computer grade hc full range speakers, if I remember correctly, eight per side. Did not have proper hights till I placed tweeter in the middle.
Anyway, all I wanted to say is that line array with coaxials is not a good idea. Line array itself well done is good idea. Out of my three complete speakers, one is open baffle system, one is horn loaded, one is line array. Its line array which is most natural sounding, has lowest distortion, has best soundstage and so on...
xlr8, go to zalytron web to see line arrays...
Probably for cost reasons & relative simplicity more than anything else. With much more limited resources than he had at McKintosh, it will be both cheaper & simpler to use FR drivers & Eq the response at either end than use twice as many drivers & have to design an XO for them into the bargin. HF & LF in both the prototype & production versions need a major boost. Sans Eq the LF won't be up to much, though it should be much better with. Still not ideal though IMO -if FR drivers are demanded, I'd much rather do a focused array, where comb-filtering & HF attenuation simply isn't an issue. Something on these lines: http://www.geocities.com/dmitrynizh/labaffles.htm
Just to add a modicum of personal experience - I've listened to the Jordan 4 unit line arrays a few times. They sound fine, the only anomaly occurring when listening close to and standing with your ears significantly above the line array, when HF tails off a little due to interference effects. Ted's system and room arrangement means you're listening close to the speakers - standing farther away might give a wider vertical spread and less severe effect. As it is, I thought it more tolerable than the left-right sweet spot of some speakers.
But this is a comparatively short array, the whole thing only a foot high. I've heard a six foot array using Bandor drivers but thought that sounded less coherent than the shorter 7th Veil Bandor array on dem at the time. Mind you, the sight of umpteen little drivers producing bass and hardly moving was remarkable to behold!
I would have thought it was worth having a go, especially with smaller drivers. Room reflections and sounds may swamp any theoretical disadvantages anyway.
But this is a comparatively short array, the whole thing only a foot high. I've heard a six foot array using Bandor drivers but thought that sounded less coherent than the shorter 7th Veil Bandor array on dem at the time. Mind you, the sight of umpteen little drivers producing bass and hardly moving was remarkable to behold!
I would have thought it was worth having a go, especially with smaller drivers. Room reflections and sounds may swamp any theoretical disadvantages anyway.
I suspect the the reason the short arrays work acceptably is because of their comaratively diminutive length. Some HF will still be lost though the lobing of the outputs though -we can't break the laws of physics. The more you stick in, the worse the problem is likely to be, as you experienced with the 6ft jobs. Personally I'd love to do a 25 driver focused array with the smaller Jordan or Bandor units, but I'd have to rob a bank first.
that's excelent link Scottmoose, I know that site
similar speaker was constructed for NY Noise
http://www.enjoythemusic.com/nynoise/
similar speaker was constructed for NY Noise
http://www.enjoythemusic.com/nynoise/
The Line Source Speaker & its Applications, Paul Taylor
http://p10hifi.net/planet10/TLS/downloads/taylor-line-array.pdf
http://p10hifi.net/planet10/TLS/downloads/taylor-line-array.pdf
Attachments
adason said:
o yeah! 😀
Roger Russell has impaired hearing, especially in the bass, hence this rightly criticized bass response, I see, and You have poor memory "was it 8"?
😉it is so nice that You can hear well, so well that You can even hear things You haven't heard! congratulations!
therefore I totally agree with You! what else can I do?
focused arrays, tweeter lines
"people greatly delight in tortuous paths"
oh well, anyway, as it has been said I am no expert 😀 🙂
happy experimenting!
best,
graaf
I've made a meager attempt this morning to calculate the load using 20 3" Peerless 830986 full range drivers. If I'm doing this correcly, then I've figured out that 4 sections of 5 drivers each wired in series gives a total of 40ohms per section. These 4 sections wired together in parallel should give me a load of around 10 ohms. Am I close in my thinking on this? Or am I out in left field? Did the same sort of calculation and came up with a nominal impedence of 8ohms for 16 drivers as opossed to 20. Never attempted to calculate so many drivers before.
Aside from that, I really like the simplicity of the open baffle designs you folks have posted. I also like the idea of the focused array, but don't see how I could get the MG3A tweets to mate without possibly running into other problems. I believe my fist attempt at this will be an open baffle vertical line array with the 60" ribbons. I believe this route would be cheap, give me a good idea of what I'm dealing with in the process and ultimately provide me with some direction for future projects like this. I'm sure I'll need more help in the future.😉
Aside from that, I really like the simplicity of the open baffle designs you folks have posted. I also like the idea of the focused array, but don't see how I could get the MG3A tweets to mate without possibly running into other problems. I believe my fist attempt at this will be an open baffle vertical line array with the 60" ribbons. I believe this route would be cheap, give me a good idea of what I'm dealing with in the process and ultimately provide me with some direction for future projects like this. I'm sure I'll need more help in the future.😉
graaf said:o yeah! 😀
Roger Russell has impaired hearing, especially in the bass, hence this rightly criticized bass response, I see, and You have poor memory "was it 8"?
it is so nice that You can hear well, so well that You can even hear things You haven't heard! congratulations!
therefore I totally agree with You! what else can I do?
focused arrays, tweeter lines
"people greatly delight in tortuous paths"
oh well, anyway, as it has been said I am no expert 😀 🙂
happy experimenting!
What Adason will have been primarily refering to WRT hearing is that Roger is not as young as he used to be, ergo, his own HF hearing will not be what it was, and he will probably be satisfied with a little less at the top end than some other people will. Hardly a crime, either way.
Sticking with the IDS25, because it's a popular example, it uses heavy Eq at both ends, in the bass to provide the desired LF gain, and at the top end to correct for HF losses caused by output lobing. His own words, not mine in point 10 of the FAQs: http://www.ids25.com/FAQ's.htm#10. Why is an equalizer required
Fair play, & this solves the SPL issues at either end -unfortunately, it does little about phase problems associated with this arrangement (see the above image Dave posted, a measure done by one of the godfathers of the array concept). No, it won't be a problem for everyone, but it should still be made clear in advance so an informed choice may be made to suit individual requirements.
I'd hardly call a focused array 'tortuous.' It's only a little harder to build than a straight cabinet, and circumvents the combing & its attendant HF losses / phase issues associated with a straight line of FR units (not within its operating window anyway). Personally, I'd call avoiding problems & removing the requirement for Eq less tortuous, not more. If you want a quote in response, try Einstein's 'everything should be as simple as possible but not simpler.' IMO, and that of many others, a straight array of FR drivers is simpler than it should be for its intended purpose unless you are willing to accept the need for major Eq, and that there will be some phase issues.
xlr8 said:
With the ribbon tweeters, you won't need a focused array. Just make sure you cross over low enough. 🙂
xlr8 said:
wire them all in series! 😀
amplifiers like high impedance, the higher the impedance the less current and it means less distortion, Nelson Pass' opinion , not mine
I am not an expert, I have no opinions
but Nelson Pass is a "diyaudio approved" 😉 expert without a slightest doubt! 😀
yeah I know, "one cannot wire drivers in series, it affects Qes and so on blah blah" 😉
I am no expert of course
but isn't Martin J. King a "diyaudio approved" 😉 expert?
here You are:
kelticwizard said:
but in case I am wrong please tell us adason, are Pass and King experts or just "experts"?
😉best,
graaf
phase problems? can You elaborate on this a little please 🙂Scottmoose said:
are they comparable to phase problems caused by midbass(midrange)/tweeter crossover? and by positioning the midbass and tweeter lines laterally side-by-side?
yes, I agree, I was saying in more general terms, line arrays with many drivers are less or more 'tortuous' for me. I am not an amateur cabinet-maker, just a music lover 🙂Scottmoose said:
cutting 20 holes for midbasses an the 10 for tweeters and this twice for two channels is especially tortuous for me
Though I agree that not for everybody 🙂
my favourite quotation from Albert is:Scottmoose said:
most people at audio internet fora don't understand that
and instead of trying new things they keep saying: "it is a disaster", "this cannot sound", "real experts blah blah" and so on
"the earth cannot revolve around the sun", "the stones cannot fall from the sky" and so on
sad but true
Scottmoose said:
but isn't it that the more the array is focused the less "line source"-like it is?
And I think that equalization is not a problem, it is perfectly normal thing
I wonder why people somewhat fear equalization? What's the point?
perhaps they don't want to be accused of being not "true audiophile"?
it appears that most people think "equalization" and imagine a "crappy graphic equalizer"
I understand that but I think it is wrong attitude resulting from prejudices
Scottmoose said:
but crossover means phase problems, doesn't it?
BTW I would like to thank You Scottmoose! 🙂
I appreciate Your informative and kind responses very much, they are exemplary
for me this is how an interesting and elegant discussion should look alike 😀
best,
graaf
graaf said:
Re the first, I'd suggest you take a look at the diagram Dave posted, and read through the paper it comes from, along with Jim Griffin's, and those put out by JBL. Note that the off-axis response will suffer earlier than the on-axis (in fact, it will narrow the latter). XO's are a different issue; they have their attendant problems (which is why I don't generally like them in the telephone band), albeit usually less severe than those you could expect from, say, a vertical line of FR units with the outputs lobing. YMMV as always. As for twin lines, it's going to depend on the layout & if you designed the spacing etc. right.
graaf said:
I know the feeling -although if you're into arrays, by definition, your 'serious' about audio. As a rule, IMO, the more effort & care you put in, the more you get out; if something takes a while, so be it. Doesn't appeal to everyone of course. 😉
graaf said:
In all honesty, I don't believe this applies to 'most people.' Some, yes, but that applies to most any walk of life, as well as 'net forums.
Having said that, it's worth keeping in mind that speakers do operate within certain well-known physical laws, and if a person tries something knowing in advance that it violates one or more of these (extent immaterial) then it's a case of 'they brought it on themselves' if it doesn't turn out to work as well as they hoped.
graaf said:
Interesting question. A focused array is certainly somewhat different to a straight nearfield array, as I mentioned before. It will have a smaller sweet-spot for example. That said, if we have off-axis phase problems, the usable sweet spot isn't automatically going to be a great deal larger anyway, so it's a case of damned if you do, damned if you don't, and which compromise you personally happen to favour. Remember too that the output from the ends of a straight array has further to travel than the centre before it reaches the listener's ears, hense the frequent use of power-tapering in such designs (see Jim Griffin's paper on the subject). A concave curve will have the same effect, if you design it right, time-aligning the drivers with the listener.
graaf said:
As it happens, I agree with you in certain respects. Eq (especially digital) can be very effective & useful. However, you also have to go careful, it's not a panacea. Not everyone can afford, or has the facilities to use it, and some can be intrusive. Personally, my take is if you can avoid the necessity for it, so much the better (assuming you don't make too many compromises elsewhere). It's all about getting the right balance.
graaf said:
Indeed they can, although again, it's a case of balancing the evils, and selecting where you'd prefer there to be a compromisein the XO or with the driver outputs? As I say, I normally dislike XOs in the midband, but it depends on which compromise is least likely to drive you up the wall.
If you're reffering to the mid/tweeter piont, what would be an appropriate starting piont? Personally, I was thinking around 3.5k since the Peerless drivers I'd be using are able to operate well beyond this. My thinking was to try and keep the crossover piont as far away as possible from the range of the human voice...just something I was reading about, but I enjoy everyone's opinion.
xlr8 said:
Ideally the XO point is determined by the midbass driver spacing and each driver has response on either side of that. If you can get the Peerless mounted with 3" c-c spacing then the absolute max XO is 1 wavelength = 13560/3 = 4.5 kHz. Half that is better. Unless you use 2" drivers you are landing in the critical range in a line-array.
How low do the magnepan ribbons go? I'd be tempted to start by using the XO already in the maggies.
Also keep in mind that the sonics will probably benefit from having the maggie ribbons cover as much of the range as possible.
dave
planet10 said:
within their own Xmax limits, of course - and the stock XO point would probably be a good guess as to where that might be
- Status
- Not open for further replies.