Hi everybody.
I need advice on tweeter choosing. Due to my task specific, I need tweeter with the lowest THD in the operating range. I used Fountek NEOCD3.5H before. Excellent speaker with outstanding money value. After Fountek's death I began to use SB26ADC & 27TAC/GB. The best I've found is AMTPRO-4. But it's pretty big thing. They're difficult to use in bookshelf speakers. + they need to be improved with some FIR.
Unfortunately, no manufacturer provides THD measurements in datasheets. HiFiCompass is a great site but not everything is there. I would appreciate any advice on which speakers to look for. Blind buying is expensive and inefficient.
.
I need advice on tweeter choosing. Due to my task specific, I need tweeter with the lowest THD in the operating range. I used Fountek NEOCD3.5H before. Excellent speaker with outstanding money value. After Fountek's death I began to use SB26ADC & 27TAC/GB. The best I've found is AMTPRO-4. But it's pretty big thing. They're difficult to use in bookshelf speakers. + they need to be improved with some FIR.
Unfortunately, no manufacturer provides THD measurements in datasheets. HiFiCompass is a great site but not everything is there. I would appreciate any advice on which speakers to look for. Blind buying is expensive and inefficient.
.
A few years old now, but many current tweeters here still. Commentary and data included.
http://www.zaphaudio.com/tweetermishmash/
http://www.zaphaudio.com/tweetermishmash/
Thank you. Looked at this site. It's very inconvenient to read them in absolute coordinates. But the worst thing is that measurements are made inaccurately. I have some tweeters from his list and I took THD measurements by myself. Also I compare them with HiFiCompass measurements. I don't know how John made these THD measurements, but they're not the same.
From the same page: "Harmonic distortion charts are taken near field at roughly 4". . . . And finally, note - Do Not compare these harmonic distortion plots to others anywhere else, including on this site. The test method was different for this particular set. Make comparisons within this test group only."
Regarding absolute levels:
-30 dB = 3% distortion
-40 dB = 1%
-50 dB = 0.3%
etc.
http://www.sengpielaudio.com/calculator-thd.htm
Regarding absolute levels:
-30 dB = 3% distortion
-40 dB = 1%
-50 dB = 0.3%
etc.
http://www.sengpielaudio.com/calculator-thd.htm
You might also find the HD measurements here useful: https://heissmann-acoustics.de/en/treibertests/
After clicking on a HD plot, be sure to wait for it to cycle through the different levels.
After clicking on a HD plot, be sure to wait for it to cycle through the different levels.
Measurements as Zaphaudio points out don't perfectly correlate to the most pleasing sound. He liked ribbons and they measure so-so at best. I recently went to Audiofest in Seattle and noticed that I liked the treble in precisely 2 different setups: ribbons and the scan speak revelator(which is the ring tweeter). Few if any of the others were pleasing if that helps. Both ribbons and Revelaters(damn spellcheck won't let me spell it correctly. 🙂) are sadly expensive.
The SB26ADC has lower THD (and generally non-linear distortion) than the AMTPRO-4 though it depends on output level. Maybe you are looking for something different than THD or standard non-linear effects?
Note that good compression drivers are often a bit higher in THD (because of second order), but at least an octave above Fs are very low in higher-order non-linear effects. (..those with poly surrounds do better closer to Fs.)
Anyway, if you don't push the driver to loud (and have a suitable high-pass filter) then this soft dome is exceptionally low in THD and non-linear effects:
https://hificompass.com/en/speakers/measurements/scan-speak/scanspeak-d2905970000
The AMT PRO-4 has very low distortion. The problem with it is that it doesn't play very low, it has a slowly rising response, but one that dives down again after about 9kHz. This makes the driver difficult to work with as a "tweeter".
What frequency range do you plan to use? This is important. For example a 3/4" dome (like the Hiquphon) just cannot be crossed very low, not lower than 3-4kHz or distortion will be high. So please give more information on how the tweeter will be used.
What frequency range do you plan to use? This is important. For example a 3/4" dome (like the Hiquphon) just cannot be crossed very low, not lower than 3-4kHz or distortion will be high. So please give more information on how the tweeter will be used.
I think it was Joachim Gearhardt (sp?) that said in a thread a while back that the Viawave was the lowest distortion driver he had tested. Not cheap.
THD increase linearly with volume, therefore they must be measured at a certain input power. It would be more correct to do this at some identical SPL,
About SB26ADC vs ATMPRO-4, at the same volume THD in AMTPRO are less then SB, and it is well audible. But
AMTPRO is very difficult to use speaker. Especially if you do it with analog filters only. I use a proprietary FIR filter which, in addition to HPF, linearizes FR/PR. But this speaker is really king size. 🙂
I've known Dmitry Malinovsky for many years and even installed his RT850 in my monitors. I also tested GTR-145 when Dmitry just started making it. These are outstanding speakers, but not the best.
Stanislav makes really great twitters with extremely low THD an 2-10kHz. But they are not so good above 10kHz. I tested T25A-6 and unfortunately it "sounds" at the end of HF. Measurements of T25D-6 look much better but it's overpriced for my purposes. I make commercial product and have to take care of the cost.
I use bi amping with FIR filtering for both speakers. Filters linearize amplitude-frequency and phase-frequency characteristics as much as possible. Thus I can use all woofers features and usually set up crossover at 4,5-5,5kHz. I would be very happy if I had twitter with THD<0.1% at 4-20kHz that would cost no more than $300 per pair.
Nope.
HiFiCompass has both measured.
The Pro-4 is lower in non-linear distortion at lower output levels above 3kHz overall (and 2.2 kHz up nearing a low 88 db).
For reference the Pro-4's 1v41 result is similar to the ADC's 2v83 at that output level (around 88db).
As long as you have a sufficiently "steep" high-pass filter above 2.5 kHz the Pro-4 will likely be lower in non-linear distortion. BUT if you need to crossover at a *lower freq. (and/or you want a more "shallow" high-pass filter) AND you need output higher than about 90 db (even if only for dynamic range) then the SB will be lower in non-linear distortion.
*and this is critical with respect to midrange integration.
The most audible component of non-linear distortion is 5th - the Pro-4 (again, unless suitably high-pass filtered) has very high levels of 5th order distortion as you increase its output below 3 kHz.
If what you actually find audibly objectionable is non-linear distortion between the ADC vs the Pro-4 then you are either listening at lower input levels (which is pretty common) or you have a higher freq. high-pass filter. I wouldn't rule-out the audible differences having nothing to do with non-linear distortion though.. 😉
Last edited:
I agree. I have owned a pair for several years, but never really found a good application for it. It's too large to be an effective tweeter and has a drooping high end.
I finally decided to try it and am building a 4-way project with it now. I will be crossing it over to a large ribbon tweeter, probably around 5k-7k Hz and will use it down to 1.5kHz or so. Within that band it should work very well, and it will be relatively easy to flatten the response using IIR DSP.
You still did not mention how low in frequency you need your tweeter to play...
I forgot to mention that I measured the AMTPRO-4 back in 2017. I did this "nude" by suspending the driver by some wires, with the rear felt removed - that was how I was hoping to use it at the time. I posted the measurements to this forum. You can find them here:
https://www.diyaudio.com/community/...cross-when-used-open-back.289610/post-5009411
The distortion looks fine down to almost 1kHz, at least the low orders I measured with this approach. This was a relatively crude measurement, not like HiFiCompass, etc!
https://www.diyaudio.com/community/...cross-when-used-open-back.289610/post-5009411
The distortion looks fine down to almost 1kHz, at least the low orders I measured with this approach. This was a relatively crude measurement, not like HiFiCompass, etc!
Scan-Speak D2604/8330
In-box measurement (440mm H x 210mm W x 350mm D) cabinet, no roundover)
@1m; 2.83V input. No HPF. Absolute SPL shown.
Made in Denmark tweeter that has excellent low end performance down to 1.3KHz:
Soft dome.
After spending on that money on the Purifi PTT6.5X04-NFC, I need to pinch a few pennies on the tweeter.
I have 27TAC/GB as a metal dome, but not yet had a chance to compare.
RIP Vifa.
In-box measurement (440mm H x 210mm W x 350mm D) cabinet, no roundover)
@1m; 2.83V input. No HPF. Absolute SPL shown.
Made in Denmark tweeter that has excellent low end performance down to 1.3KHz:
Soft dome.
After spending on that money on the Purifi PTT6.5X04-NFC, I need to pinch a few pennies on the tweeter.
I have 27TAC/GB as a metal dome, but not yet had a chance to compare.
RIP Vifa.
Last edited: