This just confirms what most of knew already based on HiFiCompass' review. This is a special drive unit. But, now we have some more and different data to give us more insight.
Here's the link to my original review because some of the tables and such won't easily translate here:
Purifi Audio PTT6.5W04-01A 6.5 Inch Woofer Review
Here's the link to my original review because some of the tables and such won't easily translate here:
Purifi Audio PTT6.5W04-01A 6.5 Inch Woofer Review
Welcome back Erin! Thank you for this incredibly detailed measurement data. Superbly done.
One thing I noticed, in your conclusions section, it states:
"An incredibly low 2% THD at 80Hz with 12vRMS (making for about 101dB @ 1 meter) and well below 1% THD above 100Hz."
But looking at the graphs, I see 12V RMS is labeled as 109dB. This is the first time I'm looking at Klippel distortion data, so I'm not sure if this is a typo or I'm misinterpreting.
Also, why 12V? In the testing the driver failed at 14V.
It's kind of unfortunate that the driver failed both distortion testing because of the compression threshold in its breakup region, which is almost surely not used in real world application with proper crossover implementations. I would be really curious to know how loud the driver can get if the distortion test ignored breakup, say 20Hz-3000Hz and 80Hz-3000Hz
Overall, this is better than I'd ever hoped to see in a objective review. Please do more!
One thing I noticed, in your conclusions section, it states:
"An incredibly low 2% THD at 80Hz with 12vRMS (making for about 101dB @ 1 meter) and well below 1% THD above 100Hz."
But looking at the graphs, I see 12V RMS is labeled as 109dB. This is the first time I'm looking at Klippel distortion data, so I'm not sure if this is a typo or I'm misinterpreting.
Also, why 12V? In the testing the driver failed at 14V.
It's kind of unfortunate that the driver failed both distortion testing because of the compression threshold in its breakup region, which is almost surely not used in real world application with proper crossover implementations. I would be really curious to know how loud the driver can get if the distortion test ignored breakup, say 20Hz-3000Hz and 80Hz-3000Hz
Overall, this is better than I'd ever hoped to see in a objective review. Please do more!
Last edited:
erin,
I've meant to do this for some time -- congratulations on running superb tests! I always learn something from them, and they are very thorough and performed with good equipment. The interpretation is wonderful.
Well done!
I've meant to do this for some time -- congratulations on running superb tests! I always learn something from them, and they are very thorough and performed with good equipment. The interpretation is wonderful.
Well done!
Thanks.
No, sir. You are indeed correct. And that's my fault. I write my own scripts to process the data so they always have the same format (and, if I ever find a certain format doesn't fit a speaker I can retroactively quickly re-process the previous tests and keep everything looking the same). I do this so you guys can quickly compare, as opposed to having one set of data with differing y-axes or different points of data.
Anyway, I just fixed that mislabeling in the MTON legends. I had copy/pasted one over from my loudspeaker tests and forgot to update it for driver tests (mic is at a different position for raw drive units vs speaker).
As for the bandwidth used, yes, that is unfortunate. And I actually don't disagree with your proposed method. However, the reasons I chose 80-5k was:
1) My sanity. I don't want to have 5 or 6 different versions of the test. I know me. I'm bound to test one with the wrong template and not realize it until it's shipped back. So, I have 3 distinct bandwidths: subwoofer, mid/woofer/range, tweeter (if I ever test tweeters in this manner). It's not exactly "fair" to every driver type but it provides an easy apples-to-apples without me having countless templates for every size and type of speaker under the sun. Some might want to use a driver as a dedicated midbass while others would use it as a mid/woofer. Etc. And with that said, for me, the main takeaway of this information is the distortion profile with/without the bandpass applied.
2) I want to see what's going on an octave above the typical LPF region because that does still effect what happens below it thanks to the nature of the multitone test itself (it's not just harmonics).
As for why 12v, that's because I started doing HD testing first and added the MTON testing later. The MTON testing has the built-in limits. The HD testing does not. Since I am not always in the garage for the duration of the test(s), I chose 12v as my "safe" point. It gives me the standard 2.83v sensitivity through (typically) 100-102dB output @ 1 meter equivalent. And, similar to my response about MTON testing, the HD testing gives me the trend of distortion. What's happening as more power is applied. But the HD testing isn't really intended to be a "stress test". That's what MTON testing is for. The MTON testing gives me more of an idea of the actual maximum SPL based on a standard. I hope you understand.
The cliff's notes version are: to keep things easily repeatable, to keep drivers from operating above failure when/if I am not in the test room at the moment and to keep me sane. 😀
Thanks. And I will!
Last edited:
Thanks. I try really hard to generate this data and present it as thoroughly as I possibly can. It's a chore sometimes. But I work hard at it. So I really appreciate hearing when people have nice things to say. Not that it boosts an ego... it's just nice to know that it's helpful and I'm not providing crap that no one cares about.
Great. It's a 2020 midwoofer, setting new standards for this size midwoofer; for sure.
From Purifi?
Spoiler alert?
From Purifi?
Spoiler alert?
Thanks for the explanation. That makes sense.
The new values makes more sense. I was thinking wow that's some really crazy performance, 110+ dB with low distortion and low compression? That's 500W+ input power into the drivers. They're even better at handling current than I thought! But that's definitely too good to be true.
The revised results look more realistic, but now I'm even more curious how well these drivers handle high power, since the excellent BL symmetry and inductance means the BL(i) and Le(i) performance is class leading, which would be most interesting to see how it handles 100W+ input.
You know, one way to cure sanity is to change all midwoofer's test bandwidth to 80-3000Hz 😱😀 But you're right, distortion performance one octave above LPF is important.
Excellent Review! 🙂
..off-topic:
-just finished watching your Bose 901 review. I think that "preferred" in-room power curve (..the typical down-ward sloping higher freq. response in-room) is derived from Toole & Olive's research (..and here applied to Klippel's measurement process).
This general in-room preference is really only applicable to speakers with similar directive responses.
As you go more "omi" at higher freq.s (which is difficult to do accurately/w/out significant combing artifacts), a "flat" response (or more flat in-room) tend's to sound better. Increase directivity at higher freq.s (like most designs), and treble typically sounds louder relative to the average in-room. (..and yes, this is counter-intuitive.) Note that some of Harman's horn designs (JBL) using this subjective-result data tend to have even greater pressure-loss in-room further down in freq. (..often having a lower power response result in that 500-2 kHz region relative to some of their other more typical designs).
..off-topic:
-just finished watching your Bose 901 review. I think that "preferred" in-room power curve (..the typical down-ward sloping higher freq. response in-room) is derived from Toole & Olive's research (..and here applied to Klippel's measurement process).
This general in-room preference is really only applicable to speakers with similar directive responses.
As you go more "omi" at higher freq.s (which is difficult to do accurately/w/out significant combing artifacts), a "flat" response (or more flat in-room) tend's to sound better. Increase directivity at higher freq.s (like most designs), and treble typically sounds louder relative to the average in-room. (..and yes, this is counter-intuitive.) Note that some of Harman's horn designs (JBL) using this subjective-result data tend to have even greater pressure-loss in-room further down in freq. (..often having a lower power response result in that 500-2 kHz region relative to some of their other more typical designs).
Update
After some feedback, some thinking and a whole lot of re-testing, I have revised my tests to include better distortion measurements. I updated the Purifi results last night.
I have changed the following:
HD results displayed in dB scale (relative to fundamental)
IMD - new
Multitone (max SPL) bandwidth changed to represent a more realistic range and tightened thresholds
After some feedback, some thinking and a whole lot of re-testing, I have revised my tests to include better distortion measurements. I updated the Purifi results last night.
I have changed the following:
HD results displayed in dB scale (relative to fundamental)
IMD - new
Multitone (max SPL) bandwidth changed to represent a more realistic range and tightened thresholds