|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
71.222.126.239
In Reply to: RE: He already has that one posted by Cask05 on July 01, 2015 at 01:26:27
Chris, what driver does the 110dB constant directivity tractrix use? All the drivers I'm familiar with roll off (if not boosted by beaming from the horn) toward HF, requiring the entire response to be flattened to the level of the top corner (usually around 15kHz). What drivers are 110dB SPL/m at 15kHz?
Follow Ups:
TAD TD-4002s. They're magnificent on this horn. I've had them for six years running on the conical/tractrix horns and they have been the highlight of my day (besides my family, of course). It takes two very mild PEQs and two LOSLV filters (required for controlled directivity horns) to get the response flat from 400-19K Hz with +/-2 dB on-axis. Off axis performance follows the polar plot that I linked above.I've heard a hastily EQed/crossed BMS 4592ND on the same horn in Hope, and it was very interesting--but it needed further active crossover settings work to get things dialed in.
The LF drivers- well, I'd like to defer discussing that for a while until I get things sorted a bit. I'll be ready to talk about that after some tests are completed and I get to look at relative performance.
By the way, thanks for the good software and spreadsheet that you've provided--you're a real asset to the DIY community, you know.
I just finished reading the online article of the interview with Bruce Edgar, and I can see that things have changed rather dramatically since he began to build his horns--not the least of which are the tools. It made me stop and think for a while about how fortunate we all are for the work that a few people have donated to other DIYers. Thanks again.
Chris
"As far as the ear can tell, consistently clean and spacious bass can be reproduced only by a driver unit coupled to a horn-type acoustic transformer..."; Jack Dinsdale, May 1974
Edits: 07/01/15
Isn't maximum possible something like 112dBSPL/m?
Granted, directivity would help some. Have you measured this? Most drivers I've seen are down about 10dB down at the high end, from their "midband" sensitivity.
That's about right for this driver/horn combination without any sort of gain correction...about ~113 dBSPL/m peak. I've measured it and have anechoic measurements.Corrected on-axis SPL/Phase response using the filters mentioned.
Chris
"As far as the ear can tell, consistently clean and spacious bass can be reproduced only by a driver unit coupled to a horn-type acoustic transformer..."; Jack Dinsdale, May 1974
Edits: 07/02/15
No gain corrections? It looks like TAD doesn't publish response curves of the 4002, but here is one curve I found measured by Radian (with TAD's and their replacement diaphragm).
Something doesn't add up...
Not to diss the driver or the horn, just trying to figure out physical limits.
Hi Bill
A plane wave tube response for a compression driver is essentially the same as a truly constant directivity horn except for level.
In this case, there are no drivers which have 110dB 1w sensitivity up high because the mass roll off begins around 2K for most 1 inch exit drivers and about an octave higher for drivers like the TAD.
The only way to get flat response on axis without compensation is by using a horn who's directivty narrows at a rate that compliments the fall off in electro-acoustic efficiency.
Conversely, if one examines the impedance curve for a "perfect" 50% efficient driver (in the frequency range it is 50%), one see's a resistive impedance that is 2X the Rdc, half the power is lost as heat (Rdc) while the other half is radiated. At the upper end of the frequency range, the mass controlled roll off becomes sharper due to the inductance which raises the impedance and the sum often gives that nice curve as the mass break is added to the inductive roll off gradually.
For an omni directional point source, 109 dB 1w 1m is 50% electro-acoustic efficiency and 112 would be 100% so "up high" figures in this neighborhood or higher certainly include directivity for example a horn with a directivity index of 10 increase the on axis level by 10dB over the driver on a CD horn.
The part of the horn that governs the directivity up high is near the throat or even inside the driver and can be seen by Don Keele's pattern loss formula. If one knows the horn wall angle and dimension (inches) inside the horn or at it's mouth, one can figure out where the pattern loss is. That thumb rule is for frequency = 10^6 / (angle X dimension) or for the dimension where pattern control is lost (inches)= 10^6/ (angle X frequency).
Working through a curved wall horn, one can follow why they narrow up high and why a CD horn must have a lower spl on axis.
Best,
Tom
Thanks Tom, that's roughly how I understood it.
In my playing with diy Synergyish horns, I never found it feasible to really get > 100dBSPL flat responses because of the mass rolloff issue.
CASK05,
Thanks, that clears it up. I know the TAD is good, but didn't think it was (un-EQd) that good!
I agree that separate amps for each driver (type) is a better approach, though it does complicate things and makes transporting a speaker (and getting it working again after) a lot more difficult. These days, it might even be cheaper with the high cost of copper and the low cost of class D.
Perhaps my response was confusing (due in part to the inability to format pictures within the message here). Sorry for any misunderstanding.
The response that I posted was after gain corrections. I do have anechoic response for the driver and horn, but I'm not at liberty to post it here. I did confirm your question, however.
The output of this driver/horn has never been an issue, in fact it is very easy to drive.
A comment: there seems to be a tacitly stated mindset on pursuing extremely low power amplifiers driving reactive and resistive passive crossover networks, then driving both extremely high efficiency MF/HF drivers and lower efficiency woofers by padding down or by increasing input impedance to the higher efficiency driver portions of the passive networks. This is not how I do it and I don't recommend doing it that way--it really doesn't sound very good, IMHO.
Rather, active biamping with a high fidelity active digital crossover and direct driver control separates concerns and eliminates the need for adding losses in the signal path and worrying about absolute efficiency across the entire band. This enables the use of a single high fidelity MF/HF driver without attending issues of FM/AM and compression distortion of other cone-type drivers in 3-way systems--3-way systems that are typically designed with sheer output SPL in mind, i.e., not typically a goal of home sound reproduction systems.
I assume that you saw the resulting phase curve using IIR filters only - no phase correction was used except by adjusting digital delays for the one crossover point at 420 Hz. Other benefits also ensue, like the reduction of one crossover network over typical 3-way implementations, which is a really big deal for fidelity.
Additionally, if one focuses on some intermediate measure of merit, like reducing required input amplifier power in the above (not-recommended) configuration, it actually sub-optimizes the system-level goal of sound fidelity by focusing instead on that intermediate measure of merit to the detriment of system-level performance. I don't recommend doing it that way.
But I will acknowledge that way of thinking is a mind set that has developed over many years within the home entertainment community and I would add, largely by those that may not be consciously aware of the resulting effects of making those choices on output sound fidelity.
Since we're here actually blue-skying the design of the greater part of a sound reproduction system, I thought it would be prudent to explicitly state these observations. YMMV.
Chris
"As far as the ear can tell, consistently clean and spacious bass can be reproduced only by a driver unit coupled to a horn-type acoustic transformer..."; Jack Dinsdale, May 1974
"A comment: there seems to be a tacitly stated mindset on pursuing extremely low power amplifiers driving reactive and resistive passive crossover networks, then driving both extremely high efficiency MF/HF drivers and lower efficiency woofers by padding down or by increasing input impedance to the higher efficiency driver portions of the passive networks."
That's not me is it? Cause I stated in my previous post I am using an analog active crossover. That long post has to do with co-entrant Synergy horn, which is a very different beast from an acoustic horn like the K402.I was just listing off to Tom all the ways I tried to skin that cat.
Then come to find out Tom isn't talking about that kind of horn. IMHO the Fig. 7 horn has much more in common with the K-402 than the SH-50.
"Rather, active biamping with a high fidelity active digital crossover"
I can't go with you on that one. I like DSD, and you can't manipulate a DSD file in the digital realm. It is what it is, so I treat my DSD dac like any other analog source, and manipulate the analog signal in the analog realm. I could get more sophisticated for a lot more money. Just consider it one way to skin that digital cat.
Don't like passive crossovers? Guess what. The coax driver in the SM-60M has a mechanical/passive component crossover, and one amp feeds both drivers. You may find yourself using a passive crossover if you use this horn.
Looking at some coax drivers on parts express, some of them are bi-ampable. I too would prefer that.
Big speakers and little amps blew my mind!
I think of the K402 (without the extra ports cut through yet) as an improvement on the double flare conical horns that Mr. Danley uses. Small differences in geometry lead to big differences in performance when it comes to horns. with the spherical (or tractrix) mouth being that big difference in this case. I've posted the polars of this horn (employed without close arrays for home use, i.e., not in close commercial arrays like the SH-50, et al.). The horn is shown above.Actually, you can edit DSD files in the digital domain: http://www.superaudiocenter.com/images/Sonoma32.pdf
"I can't go with you on that one."
I don't argue anymore, but rather invite to listen to a digital active crossover of quality: and I'm not talking about a $250 DCX-2496 or an even cheaper miniDSP here, but rather a more recently developed 24/96 active digital crossover of higher quality than the first generation 24 bit/48 kHz units. Clearly, you haven't heard one of these newer units if you make that argument.
And you're right...I don't see any use for passive crossovers:
"Crossovers may be implemented either as passive RLC networks, as active filters with operational amplifier circuits or with DSP engines and software. The only excuse for passive crossovers is their low cost. Their behavior changes with the signal level dependent dynamics of the drivers. They block the power amplifier from taking maximum control over the voice coil motion. They are a waste of time, if accuracy of reproduction is the goal." Siegfried Linkwitz - October 2009.
http://www.linkwitzlab.com/crossovers.htmDigital actives enable digital delays - something that you don't get with analog actives. When using horns or multi-channel loudspeaker arrays, i.e., 5.1, 7..1, etc., digital delays are an enabler.
Chris
"As far as the ear can tell, consistently clean and spacious bass can be reproduced only by a driver unit coupled to a horn-type acoustic transformer..."; Jack Dinsdale, May 1974
Edits: 07/05/15 07/05/15
I did mean to tell you the FR/phase plot looked excellent, and I' sure it sounds excellent too. I hope you are not going to cut any holes in your 402s!
Wow, you mean I only have to purchase a recoding studio grade workstation to be able to play my hi-rez files? And it's only $16,000. I almost fell over myself running for my checkbook. What am I, Abbey Road Studios?
I was thinking more along the line of HQ Player. Then I could use the 24/192 files I have, manipulate them, convert them on the fly into DSD, blah, blah, blah. It would probably take three dacs. I have looked into the Twisted Pear Buffalo Dac III. I have also looked at the DEQX.
To me 24/96 sounds as good as most vinyl, and 24/192 sounds better. Why would I want to downsample to 24/96? So I can correct a .25ms delay in my PSE-144 horn? I'm not sweating it. In fact, I'm not sweating the 3.5ms delay between my bass bin and my mid horn at 350hz.
Believe me, if a company made a three way 20bit ladder dac active crossover, then I would be the first in line. Until then I have to compromise.
I also prefer low level crossovers to high level ones. However, I still recognize that they still have their uses. I would never use one to pad down a horn to match a direct radiator, but I would use one to join and pad down a 112db horn to a 110db horn.
-----------------------------------------------
Big speakers and little amps blew my mind!
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: