Not sure why the first image is so tiny. I don't do tiny.
Can you manipulate the last chart above to show only the difference between the 2 curves? The result should be essentially a straight line (i.e. no difference) with some deviations above and/or below it.
Last edited:
I did that originally and it was not worth showing.
THe differences were 0.2db or less. Best,
Erik
I did that originally and it was not worth showing.
Best,
Erik
Then one can only conclude that's one hell of a tweeter!
Who's got a Dayton AMT they'd like Erik to test? How will that hold up?
Then one can only conclude that's one hell of a tweeter!
Who's got a Dayton AMT they'd like Erik to test? How will that hold up?
Here's the chart you were asking about. I picked 5kHz as the comparison point. That is, I use the measured 5kHz level that to be the reference level for each chart, and then compare all other frequencies relative to it. It could look a little better if I tweaked the levels up or down. Remember I don't have a very good way of using absolute levels, so instead I just assume the 5kHz level to be "correct." Still, you get the picture, literally.
Sorry for the poor formatting, but here you can see the data I was going from.
The first column is my "desired" level, followed by the measured acoustic level at 5,018Hz.
For each series I calculated the measured level difference at 5,018Hz. I use this difference to adjust each series (besides 70db) by this single number. For instance, for the 102.5 dB series, I subtracted 31.58 dB for each test point.
This would yield the first chart I showed you.
I then take this series and subtract the reference 70 dB level from it.
Sheet Actual F db_diff
70 db 69.91 5,018.1 0
90 db 90.1 5,018.1 20.19
100 db 99.55 5,018.1 29.64
102.5 db 101.49 5,018.1 31.58
Last edited:
Then one can only conclude that's one hell of a tweeter!
Who's got a Dayton AMT they'd like Erik to test? How will that hold up?
Yes, Happy to do it. Please send me the tweeter to test, along with your request written on the outside of a box of 100% Agave Casadores Silver Tequila......
So if I read this right, and using the handy dandy online calculator, the worst distortion is about 55.8 db. 102.5 - 55.8 = 46.7 db.
Converting this to %THD = 0.46% THD around 2kHz. Not bad.
Of course, I'm probably doing something wrong.
Best,
Erik
I just saw the post! I though I'll get e-mail notification when new posts come up, but nothing...
Excellent work, Erik!
This is indeed an excellent tweeter in every respect!
Distortion spectrum stays pretty much the same as at 90dB and that is great!
I could easily say that this type of distortion is controlled by quite linear compliance, without influence of Bl product nonlinearity (if any in AMTs).
We see here that approximately, for every XdB change in level we get ~1.5*X dB rise in THD level (at any reasonable SPL).
That is exactly what I get with true ribbons with homogeneous magnetic field, where distortion comes only from ribbon tightening as the excursion increases.
It's the most natural type of distortion you can have, with no breakups (of cones or domes or suspension) or abrupt Bl product changes involved (VC leaving the gap)!
It's also the most natural distortion rise vs. level that you can have (and it's unavoidable).
All in all, the distortion and compression is a non-issue in this tweeter and it will be inaudible (not distinguishable from natural sound in those departments).
It's a high-end stuff at it's finest! Two thumbs up!
Hi Raal!
Nice to see you are back. Thanks for the compliments but also for the explanations. You write things so quickly it will take me days to understand.
When you invent your next tweeter, please send me a set of five to evaluate fully....
I'm glad you find them so good. I must say subjectively I've been very happy and quite spoiled, so it's really put me off lesser speakers that cost 30x as much.
Best,
Erik
Nice to see you are back. Thanks for the compliments but also for the explanations. You write things so quickly it will take me days to understand.
When you invent your next tweeter, please send me a set of five to evaluate fully....
I'm glad you find them so good. I must say subjectively I've been very happy and quite spoiled, so it's really put me off lesser speakers that cost 30x as much.
Best,
Erik
He, he! I'll keep in mind your wishes to help evaluation!
Explanations don't matter much unless you're into building tweeters...That Mundorf sounds great in a couple of speakers that I listened with it (Kaiser Vivace and Chiara), and that's all that matters!
If you could just mate it with SS Illuminator somehow...very complimentary sound, alive and transparent.
Explanations don't matter much unless you're into building tweeters...That Mundorf sounds great in a couple of speakers that I listened with it (Kaiser Vivace and Chiara), and that's all that matters!
If you could just mate it with SS Illuminator somehow...very complimentary sound, alive and transparent.
Hi Raal,
I've asked but very few have had the chance to actually listen to the Illuminator vs. Revelator. T/S is so close I could just about use an Illuminator as a drop in replacment for the cabinet. What are your thoughts?
Erik
Well, Revelator had always sounded boring to me. All nice, but flat and lifeless.
Illuminator is a totally different beast and it sounds as alive and transparent as any top-end full range driver, and I'm not hyperbolizing here.
I cross it with 140-15D 2nd order passive at 2k5.
7" Illum. is peaking at ~6k, but I still didn't want to use 4th order filter, as with 2nd order (and physical time-alignment) you can get the impulse response from a 2-way that looks like it's coming from a single speaker and sound like that in general, only better.
It's been a while since I worked on speakers with Feastrex or Supravox, but that's the same sort of presentation, lively and correct, without ear-bleeding shouting like from Lowther or Fostex.
When mated with the right king of tweeter, immediacy and effortlessness is stunning. It sucks you right in and you're into the core of the music right away.
Illuminator is a totally different beast and it sounds as alive and transparent as any top-end full range driver, and I'm not hyperbolizing here.
I cross it with 140-15D 2nd order passive at 2k5.
7" Illum. is peaking at ~6k, but I still didn't want to use 4th order filter, as with 2nd order (and physical time-alignment) you can get the impulse response from a 2-way that looks like it's coming from a single speaker and sound like that in general, only better.
It's been a while since I worked on speakers with Feastrex or Supravox, but that's the same sort of presentation, lively and correct, without ear-bleeding shouting like from Lowther or Fostex.
When mated with the right king of tweeter, immediacy and effortlessness is stunning. It sucks you right in and you're into the core of the music right away.
Erik,
Just a word of caution, so you don't get disappointed if the same thing happens to you.
I got 20 pieces of 7", 4 Ohm Illuminators and got to measure T-S parameters on just one of them. The Fs wasn't close to the published data, it was 50.6Hz.
After 1 hour of breaking in, at 20Hz and +/-9mm of excursion, the Fs dropped to ~45Hz.
After another 23 hrs, the Fs settled down to 42.6Hz.
I don't know if I stumbled at one-off piece. It doesn't really matter for the project we're doing, so I don't really care and didn't went on with measuring all of them.
These are the T-S parameters that I got:
Qts = 0.464 Total Q
Qes = 0.503 Electrical Q
Qms = 5.945 Mechanical Q
Fs = 42.602 Hertz, Free Air Resonance
Re = 3.2 Ohms, DC resistance
Ls = 79.73u H, series inductance
Lp = 493.3u H, lossy series inductance
Rp = 3.492 Ohms, loss across Lp
Dia = 139m meters, effective
(%shift) 35.9 %, resonance with box
Vas = 24.27 litres, air volume equivalent
mms = 18.59 grams, effective mass
cms = 750.4u m/N, compliance
bl = 5.627 T*m, motor strength
n0 = 358.6m %, max efficiency
SplSens = 87.54 dBSPL max @1W absorbed
(Box Volume) 16 liters
Take these at +/-3% accuracy.
Just a word of caution, so you don't get disappointed if the same thing happens to you.
I got 20 pieces of 7", 4 Ohm Illuminators and got to measure T-S parameters on just one of them. The Fs wasn't close to the published data, it was 50.6Hz.
After 1 hour of breaking in, at 20Hz and +/-9mm of excursion, the Fs dropped to ~45Hz.
After another 23 hrs, the Fs settled down to 42.6Hz.
I don't know if I stumbled at one-off piece. It doesn't really matter for the project we're doing, so I don't really care and didn't went on with measuring all of them.
These are the T-S parameters that I got:
Qts = 0.464 Total Q
Qes = 0.503 Electrical Q
Qms = 5.945 Mechanical Q
Fs = 42.602 Hertz, Free Air Resonance
Re = 3.2 Ohms, DC resistance
Ls = 79.73u H, series inductance
Lp = 493.3u H, lossy series inductance
Rp = 3.492 Ohms, loss across Lp
Dia = 139m meters, effective
(%shift) 35.9 %, resonance with box
Vas = 24.27 litres, air volume equivalent
mms = 18.59 grams, effective mass
cms = 750.4u m/N, compliance
bl = 5.627 T*m, motor strength
n0 = 358.6m %, max efficiency
SplSens = 87.54 dBSPL max @1W absorbed
(Box Volume) 16 liters
Take these at +/-3% accuracy.
Hi Raal,
Oh, I could have told you this.
Their published data doesn't even match itself. Look at the resonant F on the impedance chart. Totally different than spec. Look at the recommended box alignments, no way you can have that Vas and that Q.
Sorry man, you should have talked to me first.
Madisound's published recommended alignments seem to be based on real data.
Best,
Erik
Oh, I could have told you this.
Their published data doesn't even match itself.
Look at the resonant F on the impedance chart. Totally different than spec. Look at the recommended box alignments, no way you can have that Vas and that Q. Sorry man, you should have talked to me first.
Madisound's published recommended alignments seem to be based on real data.
Best,
Erik
Take for instance the driver I used in the box we're talking about.
I used the 18W/4531-G00 7" Revelator. If I take their spec at face value:
http://www.scan-speak.dk/datasheet/pdf/18w-4531g00.pdf
Look at the impedance chart there and compare with claimed Fs.
Another good data point for you is the Klang Tong Nada kit alignment, which IS for an Illuminator. That's what I used for the Revelator / Mundorf kit we've been discussing. Rather than rely on the manufacturer, I cheated and used it since the Qts and Vas were so similar and I totally hit it right. but they didn't measure anywhere near spec.
Best,
Erik
I used the 18W/4531-G00 7" Revelator. If I take their spec at face value:
http://www.scan-speak.dk/datasheet/pdf/18w-4531g00.pdf
Look at the impedance chart there and compare with claimed Fs.
Another good data point for you is the Klang Tong Nada kit alignment, which IS for an Illuminator. That's what I used for the Revelator / Mundorf kit we've been discussing. Rather than rely on the manufacturer, I cheated and used it since the Qts and Vas were so similar and I totally hit it right.
but they didn't measure anywhere near spec. Best,
Erik
Last edited:
I strongly suspect that when these measurements were published an intern was involved, and they've just never bothered to correct it. I'm sure you have contacts at Madisound, they measure most of their drivers, and you being a valued supplier of theirs they would probably offer to share with you their real data.
They're really nice guys, but won't usually share their measurements with us mere customers.
Best,
Erik
They're really nice guys, but won't usually share their measurements with us mere customers.
Best,
Erik
Also see this previous discussion:
http://www.diyaudio.com/forums/multi-way/281185-dayton-dats-v2-new-ss-7-4-ohm.html
Here is spec vs. measured for the Revelator 7", 4 Ohm 18W/4531-G00. calculated using added mass of 16.7 grams
Fs 33, 48.5 Hz
Qts 0.35, 0.61
Qes 0.38, 0.68
Le 0.3, 0.168 H (10k)
Vas 1.48, 1.09 (ft3)
Vas 41,9, 30.9
Re 3.4, 3.51
Min Z 4.3, 3.8
http://www.diyaudio.com/forums/multi-way/281185-dayton-dats-v2-new-ss-7-4-ohm.html
Here is spec vs. measured for the Revelator 7", 4 Ohm 18W/4531-G00. calculated using added mass of 16.7 grams
Fs 33, 48.5 Hz
Qts 0.35, 0.61
Qes 0.38, 0.68
Le 0.3, 0.168 H (10k)
Vas 1.48, 1.09 (ft3)
Vas 41,9, 30.9
Re 3.4, 3.51
Min Z 4.3, 3.8
Yes, guys at Madisound are fantastic people!
Yeah, I know about the discrepancy in impedance peak and the Fs number.
IIRC, a long time ago (>15years) I think I read in some of the datasheets of some old Scan Speaks that they measure the acoustical response of the speaker in 100l box.
Dunno if that's true (never bothered to ask) but ever since, I attributed the discrepancy to the speaker being measured in 100l box (and never actually calculated whether that fits the difference).
Whatever, they were including a sidenote in their datasheet that referred to the measurement conditions, but not any more...
In any case, since I never expect anyone to trust my measurements, I apply the same logic to the stuff I want to get, so I always just get a piece and see what I have to work with, then take it from there...it is what it is and I'll try to get the max out of it, no biggie.
Funny, but I never did that with SS Ill. I listened to it in a few very fine speakers and when I needed something like that, I got 20 without previous testing...ts, ts, ts.
Guess I'm lucky that I need them for 0.6 cu.ft. closed box! I could still vent that, too, we'll see...
Yeah, I know about the discrepancy in impedance peak and the Fs number.
IIRC, a long time ago (>15years) I think I read in some of the datasheets of some old Scan Speaks that they measure the acoustical response of the speaker in 100l box.
Dunno if that's true (never bothered to ask) but ever since, I attributed the discrepancy to the speaker being measured in 100l box (and never actually calculated whether that fits the difference).
Whatever, they were including a sidenote in their datasheet that referred to the measurement conditions, but not any more...
In any case, since I never expect anyone to trust my measurements, I apply the same logic to the stuff I want to get, so I always just get a piece and see what I have to work with, then take it from there...it is what it is and I'll try to get the max out of it, no biggie.
Funny, but I never did that with SS Ill. I listened to it in a few very fine speakers and when I needed something like that, I got 20 without previous testing...ts, ts, ts.
Guess I'm lucky that I need them for 0.6 cu.ft. closed box! I could still vent that, too, we'll see...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.