Drive Unit Tests: Scanspeak 7-inch Midwoofer, SB Acoustics Ceramic 6-inch Midwoofer, Purifi 4-inch Midwoofer

[hardisj]

Look, I know we all like for me to share the reviews directly here but 3 different reviews is gonna be a royal pain for me to copy/paste here. So, if you are interested in these drivers' performance, check the following links to my reviews:
ScanSpeak Illuminator 18WU4741T-00 7 Inch Midwoofer Review
SB Acoustics SB17CAC35-4 6 Inch Ceramic Midwoofer Review
Purifi Audio PTT4.0W04-01A 4 Inch Midwoofer Review

 

[HooStat]

Just curious -- what is the use case for the 4" Purifi? I don't pretend to know anything, but would be interested in hearing how some DIY people view the usefulness of this driver.

 

[MZKM]

Just curious -- what is the use case for the 4" Purifi? I don't pretend to know anything, but would be interested in hearing how some DIY people view the usefulness of this driver.

Same.

Unless you have hundreds of watts or will use near-field, it needs to be doubled to increase sensitivity to make it usable. Also, compared to the 6.5”, the upper extension is only a bit higher.

 

[andreasmaaan]

Nicely done again @hardisj

I'm a little disappointed with the SB17CAC's Klippel performance/bass distortion. It's such an exceptional driver in the midrange. I guess for <$100 it's hard to have it all.

Interesting that you say:

One-way linear excursion measured approximately 2.8mm for this test sample. This is very low for a midwoofer and indicates this particular drive unit is best served as a midrange. I’d be leery of pushing this below 80Hz (as a starting point).

I tend to agree, but Harman/Revel obviously doesn't!

 

[Zvu]

It could just be possible that drivers in Revel are not of the shelf SB acoustic.

I often see that presumption but unless someone measures Revel woofers out of the cabinets, we will never know for sure.

 

[andreasmaaan]

It could just be possible that drivers in Revel are not of the shelf SB acoustic.

That's true. I've heard plenty of online gossip concerning these drivers and Revel, but I won't repeat it here.

One thing that does seem quite likely (based on measurements I've seen in various places online) is that the CAC and CDC series are essentially cosmetic updates of the NAC/NBAC series. Performance suggests they use the same motor and suspension, and essentially the same diaphragms, other than that the Alu is coated with a thin ceramic layer (the purpose of which I can't say is clear to me, although I do think they look nice, and the price differences are pretty negligible).

 

[Zvu]

I'm guessing that it is a marketing trick - superiority of CAC over NAC. CAC and NBAC measure the same, but you can choose between three colours and that helps with the sales.

 

[Vini darko]

This is way above my pay grade. Understood almost nothing. I still have a crazy amount to learn about drivers. Thankyou for the review.

 

[Ericglo]

Great job as usual on the testing.

You noted in the SB review that there is a 5" version that might be better as a mid. There is also a 4" version.
https://sbacoustics.com/product-category/family-product/cac-family-product/

Rick used the 4" in his Ceramicas
http://www.selahaudio.com/ceramica
http://www.selahaudio.com/ceramica-kit

 

[briskly]

Just curious -- what is the use case for the 4" Purifi? I don't pretend to know anything, but would be interested in hearing how some DIY people view the usefulness of this driver.

A "modern" home audio midbass driver: low efficiency, low cutoff point/wider bandwidth enabled by high-power electronic amplifiers. The power response of the driver is higher extended in frequency than its larger sibling, but this comes off as a side benefit. Suited for a compact 2-way loudspeaker used within a couple of meters and driven from a high power Class D amplifier.

I tend to agree, but Harman/Revel obviously doesn't!

It would be nice if the nonlinearity didn't exist, but Kms(x) nonlinearity is relatively benign in the grand scheme. A bigger problem is that the Klippel shows suspension asymmetry.
The driver's directivity is slightly funky for midrange duty, possibly from the cone construction. A horn to ease the HF unit's directivity transition and a lower x-over point are probably the best approach.

the purpose of which I can't say is clear to me

A thicker oxide layer to protect the aluminium cone would be a simple one. From Ulrik Schmidt (SB Acoustics):

The surface treatment on the cone is a different process. The SB12PAC cone (as well as the NBAC cone) is anodized, whereas the CAC cones are treated with a plasma process that penetrates slightly deeper into the material (i.e. it turns a little more of the aluminium into a ceramic material) - but it is the same basic cone that is used.

 

[andreasmaaan]

It would be nice if the nonlinearity didn't exist, but Kms(x) nonlinearity is relatively benign in the grand scheme. A bigger problem is that the Klippel shows suspension asymmetry.
The driver's directivity is slightly funky for midrange duty, possibly from the cone construction. A horn to ease the HF unit's directivity transition and a lower x-over point are probably the best approach.

TBH, if I were going to use a driver from this series as a midrange, I think I'd just use the SB15CAC/NAC/NBAC instead.

A thicker oxide layer to protect the aluminium cone would be a simple one. From Ulrik Schmidt (SB Acoustics):

Thanks. I'm surprised that anodisation is deemed inadequate to protect the cone in the first place, but I don't know a lot about this topic.

 

[Dennis Murphy]

Nicely done again @hardisj

I'm a little disappointed with the SB17CAC's Klippel performance/bass distortion. It's such an exceptional driver in the midrange. I guess for <$100 it's hard to have it all.

Interesting that you say:


I tend to agree, but Harman/Revel obviously doesn't!

This is pretty puzzling to me. I've worked with dozens of the 17CAC's and have never run into distortion issues using the standard 90 and 96 dB one meter tests or comparing it with the Scan 8545-01. In addition, the Fs always clocks in at about 36 Hz after break-in, never anywhere near as high as the reported 44 Hz. I have to wonder whether there was sufficient break-in before testing, or whether there was something out of spec with this unit. (The factory claims 29.5 Hz, which is a fantasy.) Finally, this driver, along with its essentially identical black twin, has been Klippel tested elsewhere and did very well. Of course, I may have missed something because I don't read German.

 

[bigjacko]

Thank you so much Erin for testing those drivers, those are what really interest me because if I want to DIY I better get something worth my time. It is quite dissapointing that Illuminator measures so well on Klippel but the distortion is not comparable to SB17CAC and Purifi 6.5".

I think SB17CAC can be used as a mid woofer crossover from 800Hz to 200Hz for someone that wants lowest possible distortion. If crossover at 200 Hz the excursion of the driver would be low. Given the price at just $85 plus a low woofer they might be cheaper than purifi 6.5". The THD for SB17CAC and purifi 6.5" are very similar, SB17CAC got higher third order and purifi got higher second order compare to each other. I think we need to study some papers about the masking effect to determine which one has better THD at mid range. Also wondering if the SB15CAC has comparable THD to SB17CAC, does anyone has measurement or comparison of both?

 

[hardisj]

The driver was sufficiently broken in. I test and re-test T/S parameters until they “settle”. It is hit with a lot of pink noise to get the coil moving for a while. Allowed to cool. Tested. And if Fs moves more than what I consider out of bounds of measurement “error”, I repeat the process until the results settle.

This is hardly the first time I’ve had an SB Acoustics driver clock in at half it’s “linear Xmax” provided in their spec sheet. And won’t be the last, I am certain. You can find other SB Acoustics tests of mine that show this trend. Par for the course when it comes to SB’s “linear Xmax” ratings. Edit: Matter of fact, I think I had this discussion with @andreasmaaan recently and provided a couple examples of SB drivers not measuring up to their stated spec.

 

[bigjacko]

This is pretty puzzling to me. I've worked with dozens of the 17CAC's and have never run into distortion issues using the standard 90 and 96 dB one meter tests or comparing it with the Scan 8545-01. In addition, the Fs always clocks in at about 36 Hz after break-in, never anywhere near as high as the reported 44 Hz. I have to wonder whether there was sufficient break-in before testing, or whether there was something out of spec with this unit. (The factory claims 29.5 Hz, which is a fantasy.) Finally, this driver, along with its essentially identical black twin, has been Klippel tested elsewhere and did very well. Of course, I may have missed something because I don't read German.

Maybe it's just him being picky because it has so good mid distortion and also compared to purifi 6.5". If it compare to other drivers I think it still performs well even at subbass. Distortion at subbass is hard to hear because of masking effect and equal loudness contour so maybe that's why you did not notice distortion. At science forum there are too much nerd chasing SINAD like me.

 

[Wombat]

That's true. I've heard plenty of online gossip concerning these drivers and Revel, but I won't repeat it here.

One thing that does seem quite likely (based on measurements I've seen in various places online) is that the CAC and CDC series are essentially cosmetic updates of the NAC/NBAC series. Performance suggests they use the same motor and suspension, and essentially the same diaphragms, other than that the Alu is coated with a thin ceramic layer (the purpose of which I can't say is clear to me, although I do think they look nice, and the price differences are pretty negligible).

The coating: Mass tuning or constrained layer damping? Maybe to bleed unwanted surface waves to the surround.

 

[Lorenzo74]

Same.

Unless you have hundreds of watts or will use near-field, it needs to be doubled to increase sensitivity to make it usable. Also, compared to the 6.5”, the upper extension is only a bit higher.

Will be in active loudspeaker like kii three or D&D. Sensibility is of lower importance if you have plenty of clean class D watts.
the high excursion and low distortion are instead the key parameter to allow to play loud in the range of 100-2000kHz then you already handover to the tweeter.

if you look at purifi lineup it’s spot on to replicate a kii three without the need of the bxt.

then it definitely can be use with success in superconpact active 2 way design with a passive radiator or reflex tuned around 45Hz.
The low Fs allow clean use from maybe 70-80 and the long throw plus EQ will give you excellent bass for the size.
Would be best if purifi will comment on 4” best destination.
my Best
L.
ps Fantastic Review!

 

[Dennis Murphy]

The driver was sufficiently broken in. I test and re-test T/S parameters until they “settle”. It is hit with a lot of pink noise to get the coil moving for a while. Allowed to cool. Tested. And if Fs moves more than what I consider out of bounds of measurement “error”, I repeat the process until the results settle.

This is hardly the first time I’ve had an SB Acoustics driver clock in at half it’s “linear Xmax” provided in their spec sheet. And won’t be the last, I am certain. You can find other SB Acoustics tests of mine that show this trend. Par for the course when it comes to SB’s “linear Xmax” ratings. Edit: Matter of fact, I think I had this discussion with @andreasmaaan recently and provided a couple examples of SB drivers not measuring up to their stated spec.

Well, I can state as a fact that the typical Fs is 36 Hz, so that part doesn't compute, at least. I'll send your distortion results to my representative at the factory and see what he has to say. I'm tracing down the other klippel test to run it through a translator and see if I can find anything relevant that I missed.

 

[jhaider]

Nicely done again @hardisj

I'm a little disappointed with the SB17CAC's Klippel performance/bass distortion. It's such an exceptional driver in the midrange. I guess for <$100 it's hard to have it all.

The xmax delta from published is consistent with Erin's earlier measurement of the 8" variant on his old "Medleys Musings" site, and better than an earlier SB17 model he measured.

https://web.archive.org/web/20170707195922/http://medleysmusings.com/sb17nrxc35-4
https://web.archive.org/web/20170702005154/http://medleysmusings.com/sb23nacs45-8/

I'm curious what others think of the suspension giving out before the motor does. On one hand, it might make for a more linear driver when bass managed, right? On the other, if not bass managed that sounds like a recipe for nasty noises as the motor pushes the cone further than the suspension wants to go.

Just curious -- what is the use case for the 4" Purifi? I don't pretend to know anything, but would be interested in hearing how some DIY people view the usefulness of this driver.

I wonder the same thing. The market price for 4" mini-monitors is set at ~1000 USD per pair (Neumann KH 80 DSP, Genelec 8020). You can't fit this woofer, or the similarly pricey ScanSpeak Illuminator, in that budget. In the abstract I guess one could use it to attempt to challenge JBL 705 for most bass in smallest cabinet, but it still would fall short. At this size Purifi's ceramic magnet also takes up a lot of cabinet volume compared to the neo magnets on ScanSpeak Illuminator or JBL 725G. That is a strange place to cut costs, unless they fear supply constraints on neo.

PS: ScanSpeak Illuminators are the only cone drivers that look as cool as the old Aurasound woofers. In a cabinet that matters less, but if I were do to an open baffle speaker I would try to source another Aurasound NS18, use the Illuminator 7" and 4" drivers, along with a cool looking tweeter.

 

[Dennis Murphy]

Since (1) the only real problem area appears to be linear suspension travel, and (2) there's an immense amount of data presented, it wold help me (for one) in understanding the results to have stated in one place exactly how linear travel is determined. What is the drive level? What is the frequency range? Is there any standard protocol for this? I've never given much thought to specs like this before, but now that I am, I'm pretty confused.