Some driver distortion measurements

Here's a few driver distortion measurements I made today

These aren't terribly scientific, but if they help someone out, cool! I just set my amp to -3dB and measured each driver at the same amplifier level, which isn't necessarily the same SPL level because the impedance of these drivers is all over the map

I used a 15Hz highpass and a 120Hz lowpass, both LR4

I used no baffle; this is just the raw driver

First one up is a measurement of a twelve inch Infinity woofer from the Infinity Prelude speaker. These are readily found on Craigslist for a few hundred bucks. I trashed mine (all of the mids and midbasses had surrounds that were trashed) but was pleasantly surprised by how capable the subwoofers are, despite looking fairly pedestrian
 

Attachments

  • 2023-06-13 12_03_07-REW V5.20.13.jpg
    2023-06-13 12_03_07-REW V5.20.13.jpg
    255.5 KB · Views: 149
Here's a DIYMA R12. This was a super affordable twelve inch woofer made by npdang over at DIY Mobile Audio almost twenty years ago.

Funny story:

I nearly threw these away.
Basically, the sensitivity on these is so low, I thought there was something wrong with them. I've had them in my garage for ages, and even had a few people interested in buying them, but was reluctant to do so because I thought they might be defective.

Whenever I played them full range and gave them a lot of power, they sounded pretty rough. One of the reasons I went to the trouble of measuring the distortion on some of my drivers today, was that I was trying to figure out if these are broken or not.
It looks like they work perfectly fine, they're just so low in sensitivity, if you play them full range and crank the volume, the 2nd and 3rd harmonic distortion will be more audible than the fundamental.

This is something that's vexxed me about low distortion subs, in general: you have to get accustomed to the fact that a sub that's really low distortion in a properly built enclosure will be FELT and not HEARD, and if you're not careful with the low pass you could wind up in a situation where most of what you're hearing is 2nd and 3rd harmonic, particularly if you don't listen at really high volumes.

The overlay shows the Infinity vs the DIYMA, note the Infinity is 10dB more sensitive above 100Hz. VAS on the DIYMA is something ridiculous, it works in a box of around half a cubic foot and the woofer weighs a TON because they had to use a ton of motor force to drive the VAS down. I had the Diymas in a 15Hz tapped horn back in the day and it nearly killed me moving the subwoofer. Thing was like a refrigerator:

https://www.diyaudio.com/community/threads/night-of-the-living-bassheads.168737/
 

Attachments

  • 2023-06-13 12_42_14-REW V5.20.13.jpg
    2023-06-13 12_42_14-REW V5.20.13.jpg
    239.7 KB · Views: 94
  • 2023-06-13 12_21_00-Overlays.jpg
    2023-06-13 12_21_00-Overlays.jpg
    246 KB · Views: 103
  • diyma12small-1.jpg
    diyma12small-1.jpg
    58.5 KB · Views: 91
Here's an Alpine SRW-843D

I'm a big fan of the Alpine subs of this era; Alpine was basically offering subs with split gaps and shorting rings and underhung motors at prices that weren't much higher than the cheap Skar Audio stuff that's so ubiquitous on Amazon
Naturally, 90% of the car audio buyers couldn't care less about low distortion, and Alpine wound up getting gobbled up in some kind of weird merger:

https://www.reuters.com/article/alps-alpine-ma-oasis-idUSL3N2101IA

As far as the measurements go, a few thoughts:
1) I think the high distortion is deceptive here. When running the sweep, I can hear something buzzing, but I think it may be the frame of the sub itself. So two bucks worth of mortite could go a long way here. The Alpine weighs about 1/4 as much as the Diyma

2) We can see in the measurement that there's only so much bass you can squeeze out of an eight. The Alpine response looks a lot like the Infinity response, but about 8dB lower. The Alpine's VAS is just twelve liters: https://www.alpine.co.uk/fileadmin/...car_audio_manuals/SWR-823D/OM_SWR-823D_EN.pdf

The designer of the Diyma R12 told me that it took some inspiration from the Peerless XXLS subs. Due to this, it's worth comparing the BL^2/RE of the two. The BL^2/RE for the Peerless XXLS is 44.77 and for the Alpine it's 31.88

Assuming the R12 specs are similar (I still need to measure that), the R12 is definitely a better value. They were selling for around $100 when released. XXLS costs $216: https://www.digikey.com/en/products/detail/peerless-by-tymphany/XXLS-P835017/6557551

Which is actually a really good price for that kind of performance, all things considered
 

Attachments

  • 2023-06-13 13_21_40-Overlays.jpg
    2023-06-13 13_21_40-Overlays.jpg
    303.6 KB · Views: 56
  • 2023-06-13 13_25_36-REW V5.20.13.jpg
    2023-06-13 13_25_36-REW V5.20.13.jpg
    260.8 KB · Views: 50
Here's one that looks really terrible at first glance. Cerwin Vega V 12S.

Apex Jr was blowing these out for a few years; IIRC I paid less than $100 for four!

The frame is stamped steel, there's no pole vent, voice coil is tiny, magnet is tiny, it just looks cheap cheap cheap

but...

1) it's pretty darn sensitive. Yes, I know sensitivity is nearly irrelevant for subwoofers, but if I had to pick between the $200 Alpine and the $20 Cerwin Vega, I'd probably take the CV

2) Because it's so darn cheap, and it's 2nd harmonic distortion is 10dB+ louder than 3rd harmonic, this is a really really great candidate for a style of box that went out of style almost 30 years ago: isobaric. Isobaric should knock the 2nd harmonic down to nearly nothing, and it doubles our power handling. Yes, we lose 3dB in sensitivity, but we have sensitivity to burn here.
When looking at the distortion graph, note that the distortion trace looks scarier than it really is, because the overall sensitivity of the CV is highest among all the woofers I tested. IE, YES the distortion curve is about 12dB higher, but the sensitivity is ALSO about 10dB higher

Big question is if I went isobaric, would the sensitivity go down 3dB while dropping the distortion curve by 10dB?

I have four of these and there's one way to find out...
 

Attachments

  • maxresdefault (1).jpg
    maxresdefault (1).jpg
    114.7 KB · Views: 69
  • 2023-06-13 13_54_02-REW V5.20.13.jpg
    2023-06-13 13_54_02-REW V5.20.13.jpg
    251.5 KB · Views: 68
  • 2023-06-13 13_50_12-Overlays.jpg
    2023-06-13 13_50_12-Overlays.jpg
    369.8 KB · Views: 72
Member
Joined 2005
Paid Member
Thanks for the measurements. When I’m in a PC I will look at the graphs in detail,

Any particular reason you put in the LP/HP filters?

Sometimes an advantage of doing raw, is just to see where the limits are.
Some people like to use their woofers as subs and subs as woofers, etc
 
As I understand it, electronic low pass filters don't reduce distortion.

So if the driver is intended for subwoofer duty (which it is) and you measure it with the intended low pass, you should see the distortion.

IE, if you have a bunch of 3rd harmonic distortion at 180Hz, even with the low pass filter in place, the source of it should be the fundamental at 60Hz.
 
As I understand it, electronic low pass filters don't reduce distortion.

So if the driver is intended for subwoofer duty (which it is) and you measure it with the intended low pass, you should see the distortion.

IE, if you have a bunch of 3rd harmonic distortion at 180Hz, even with the low pass filter in place, the source of it should be the fundamental at 60Hz.

That's not the way that distortion is displayed in these measurements. If there's a peak in the 3rd order at 180Hz then it means the fundamental of 180Hz produced an overtone at 540Hz.

Measuring subwoofers without a low-pass in place is useful for seeing how capable the motors and cones really are. Some subs you can use up to 1kHz, some only 100Hz. This could be very important for different designs where lots of speakers, these days, like to use a subwoofer as the bottom of a three way and need it to be capable up to around 500Hz. I can't find any pictures of the prelude woofer but given its pedigree this could work well up higher Vs something like the Cerwin.
 
Member
Joined 2005
Paid Member
As I understand it, electronic low pass filters don't reduce distortion.

So if the driver is intended for subwoofer duty (which it is) and you measure it with the intended low pass, you should see the distortion.

IE, if you have a bunch of 3rd harmonic distortion at 180Hz, even with the low pass filter in place, the source of it should be the fundamental at 60Hz.
That is true, that it doesn’t reduce distortion.

Note. Though that when

“Plot harmonics at the harmonic frequency” is selected, REW changes

“the way the harmonic orders for sweep measurement distortion are plotted - rather than plotting the distortion at the frequency of the fundamental it is plotted at the frequency of the harmonic, so (for example) the 2nd order harmonic distortion level for 1 kHz would be plotted at 2 kHz, where the distortion occurs.”

Suppose you want to measure a subwoofer up to 120 Hz (the THX/Dolby LFE standard, you’d want to measure out to 360Hz to see the HD3.

Reference:
https://www.roomeqwizard.com/help/help_en-GB/html/graph_distortion.html
 
Last edited:
In Post #5, I toyed with the idea of going isobaric with the cheap Cerwin Vega woofers. My thought process was that it's a really low tech speaker with high distortion, and isobaric might improve that. So let's see!

Attached is a comparison of the Cerwin Vega by itself (a 4ohm load) and isobaric, with the voice coils in parallel (a 2ohm load.) In the comparison, you can see the output level is basically identical. There's a variance of a fraction of one decibel. The iso pair has more output in the octave between 200 and 400Hz, I'm not entirely sure why that is.

The second and third attachments are a measurement of distortion. The 2nd attachment is conventional, the 3rd attachment is isobaric. Going isobaric doesn't eliminate distortion entirely, but it definitely reduces it:
  • At 20Hz, 2nd harmonic distortion is reduced from 41.3% to 15%. That's a 63.7% reduction in 2nd harmonic.
  • At 20Hz, 3rd harmonic distortion is reduced from 10.9% to 2.65%. That a 75.7% reduction in 3rd harmonic.
  • At 30Hz, 2nd harmonic distortion is reduced from 11.6% to 7.2%. That's a 38% reduction in 2nd harmonic.
  • At 30Hz, 3rd harmonic distortion is reduced from 1.8% to 1%. That a 44% reduction in 3rd harmonic.
  • At 40Hz, 2nd harmonic distortion increased from 1.8% to 3.2%.
  • At 40Hz, 3rd harmonic distortion increased from 0.4% to 0.8%
  • At 60Hz, 2nd harmonic distortion is reduced from 5% to 1.25%. That's a 75% reduction in 2nd harmonic.
  • At 60Hz, 3rd harmonic distortion increased from 0.5% to 1.0%.
  • At 80Hz, 2nd harmonic distortion increased from 3.2% to 4%.
  • At 80Hz, 3rd harmonic distortion is reduced from 0.6% to 1.8%.
I'm no expert on driver design. My hunch on why some of the numbers went up, is that distortion levels in the range of 1-4% probably can't be measured accurately with my crude setup. Basically, if I repeated the test with more power, I could get the SPL levels up to a point where the distortion is easier to measure. As noted in previous posts in this thread, I just set my amplifier to about 95% of it's maximum gain and measured distortion in the nearfield. That's why some of these SPL numbers look crazy high, the mic is less than five inches from the cone.

It's nice being able to compare isobaric back-to-back with conventional mounting. It's noticeably cleaner.
Low distortion subs haven't been a big hit in the car audio world, IMHO. Companies like Pioneer, Alpine and Orion all offered them, ten years ago, but discontinued them. In the isobaric set up, I can kinda see why - once distortion is reduced sufficiently, you really don't "hear" the bass very much. Even with less than 100 watts on tap, the isobaric setup was able to cause things in my room to shake. IE, I could hear things being shaken by the force of the sub, but the bass itself wasn't particularly audible, especially below 60Hz:

400px-Lindos4.svg.png


My intention here isn't to discourage people from building low distortion subs. Just pointing out that it's not going to sound like a conventional sub, and you'll want to invest quite a bit of effort to be sure that your cabinet walls aren't generating harmonic distortion on their own, and that the subwoofer is isolate from the room so that the force of the sub doesn't generate rattles among the items in your room. If you stuck one of these isobaric subs in the trunk of a car, there's a pretty good chance that the sound of the deck lid rattling would be quite a bit more audible than the subwoofer itself.
 

Attachments

  • 2023-06-14 23_50_46-Overlays.jpg
    2023-06-14 23_50_46-Overlays.jpg
    259.5 KB · Views: 33
  • 2023-06-14 23_52_14-Overlays.jpg
    2023-06-14 23_52_14-Overlays.jpg
    197.6 KB · Views: 23
  • 2023-06-14 23_54_12-Overlays.jpg
    2023-06-14 23_54_12-Overlays.jpg
    218.6 KB · Views: 27
  • Like
Reactions: 1 user
Attached are the Cerwin Vegas V 12s specs as published by Apex Jr back in the day, and also my measurement that I did today

My DATS always seems to wind up with a QTS figure that's quite a bit higher than published specs. Not sure if that's because this woofer came straight out of the box and hasn't been broken in at all, or due to age, or both.

My VAS measurement also looks to be on the high side, I was expecting something around 3-4 cubic feet.

If I wasn't so impatient, I'd burn these in for 24 hours and see if the QTS and VAS will go down
 

Attachments

  • 2023-06-15 00_24_29-Overlays.jpg
    2023-06-15 00_24_29-Overlays.jpg
    234.8 KB · Views: 43
  • 2023-06-14 20_51_41-Untitled Project - DATS.jpg
    2023-06-14 20_51_41-Untitled Project - DATS.jpg
    30.7 KB · Views: 34