“Let our rigorous testing and reviews be your guidelines to A/V equipment – not marketing slogans”
Facebook Youtube Twitter instagram pinterest

SVS PB13-Ultra Subwoofer Measurements and Analysis

By

A few days after the listening evaluation ended I was able to pack the PB13-Ultra out to the test site and take some 2 meter ground plane measurements. The goal was to learn with certainty what a few different aspects of the subwoofers performance was. Some of these being frequency response shape, the effects of the various controls on the response, how the unit maintains bandwidth uniformity when asked for increased output, distortion levels, peak output levels, signal decay, etc…The overall approach to this testing along with the equipment and software used is outlined in the article here.

 See: Powered Subwoofer Testing Outline and Procedures Overview

All measurements of the PB13-Ultra were taken with the sub placed sitting on the ground as per normal use, facing directly at the Earthwork’s M30 microphone which was at a 2 meter distance as measured from the front baffle of the sub and laying directly on the ground. The grill was removed for all tests also. A 100ft length, 10ga power cord connected to a 120v, 30 amp breaker, ac main that usually reads about 125-126v at the outlet, supplied power for the Sledge amplifier. Since the PB13-Ultra has three different modes of operation which can affect its performance in different ways it was decided that we would run through the various measurements in all three modes and compare the results. This ended up being quite a lengthy test session and produced a massive pile of data. Enough for what would normally be three different subs. I have tried to condense this down to the most pertinent and informative data for this review. Otherwise this would turn into one very long article indeed! 

The 20Hz mode of the PB-13 Ultra was tested first with a variety of combinations of the internal settings, offered by the DSP built into the Sledge amplifier, to observe how they modified the response of the subwoofer: Available EQ presets, low pass filter function, parametric EQ, etc. Another reason for tackling this first is to identify the appropriate settings to use as the configuration for the rest of the tests. The DSP interface on the Sledge amplifier made this a very easy and straightforward process. As you will see in the graphs below, there is a huge amount of signal shaping available to tweak the performance.

svs pb13 ultra 20hz xover settings.jpg

  SVS PB13-Ultra: Effect of LPF Setting

svs pb13 ultra 20hz room settings.jpg


SVS PB13-Ultra: Effect of Room Gain Compensation Settings

  

svs pb13 ultra 20hz dsp eq settings.jpg 

SVS PB13-Ultra: Effect of Various PEQ settings on Response

I did not try every combination of parametric EQ settings possible with both bands. I used three octave spaced center frequencies: 31, 63 and 125Hz. Three Q values: 2.0, 7.2 and 14.4 and three boost / cut ranges: +3dB, -6dB and -12dB. I also looked at the results of using the two bands to reinforce each other which are the results at 63Hz. As you can see there are a lot of possibilities for signal shaping here, especially once you combine the room compensation control and a low pass filter into the mix. That’s not to even mention the three different output modes.

svs pb13 ultra mode response comparo.jpg

SVS PB13-Ultra: Basic Response of Each Mode as Tested

The comparison of the basic response from each output mode shows that the PB13-Ultra’s response is useable up till 200Hz if needed and is very flat and extended. The settings for the Ultra were: PEQ, LPF, R.G.C. all disabled and the volume at 0dB for all modes. The appropriate port plugs and DSP setting for each output mode were recorded. The two ported modes roll off very quickly below their tuning range which indicates that they have a high pass filter in line with the usual 24dB/ octave roll off of a bass reflex alignment. This is typical behavior. The sealed mode starts its roll off much higher near 30Hz but it has a much more shallow and gradual decline at about 12dB/octave. It appears that there may be a rumble filter used with the sealed mode as well but it seems to not take effect until below 14Hz or so. Again this is typical. Overall the PB13-Ultra in 20Hz mode meets SVS’s 18-150Hz +/-3dB specification easily. In 15Hz mode it falls within a 15-200Hz +/-3dB window. In sealed mode it is about 24-200Hz within the same window.

pb13 15hz waterfall.jpg

SVS PB13-Ultra: Waterfall Decay (15Hz Mode)

svs pb13 ultra group delay comparison 800.jpg

SVS PB13-Ultra: Group Delay All 3 Modes Compared

In the group delay and waterfall decay plots we are looking for delayed energy being stored or resonating. I have only included the waterfall for one operation mode, 15Hz. The 20Hz waterfall looks the same except that the moderate ringing near the port tune is centered a little higher. The sealed mode does not exhibit this and uniformly and quickly decays at all frequencies. In the group delay chart the blue and red curves are representing 1 cycle and 1.5 cycles of delay respectively, which are generally held as the thresholds of audibility. However there is still some debate on the subject. We can see that the sealed mode offers the best performance in this metric and that the 2 ported modes have some increased delay near the tuning frequencies. This is typical of resonant alignments. Also EQ and high pass filters can increase this somewhat as well. Still the PB13-Ultra is below 1 cycle down till about 23Hz and just barely breaks 1.5cycles at 20Hz, which is pretty good performance and very likely to be inaudible at such low frequencies. I certainly didn’t note anything during the listening tests with Hellboy II which has plenty of content in that range. In general lower is better in this test but as the frequencies go down to 20Hz and below the audibility of this is debatable.

svs pb13 ultra 20hz power compression.jpg

SVS PB13-Ultra: Long Term Power Compression (20Hz Mode)

 pb13 15hz power compression.jpg

SVS PB13-Ultra: Long Term Power Compression (15Hz Mode)

pb13 sealed power compression.jpg

SVS PB13-Ultra: Long Term Power Compression (Sealed Mode)

What is being shown in these measurement graphs is the ability of the subwoofer to maintain its basic response shape and track the increase in output volume asked of it. The signal is an ascending sine wave sweep which is 24 seconds in length. The sweeps are conducted back to back with little time other than a few seconds for the driver voice coil to cool in between. The starting level is with the units output referenced to 90dB at 50Hz at 2 meters ground plane. Each subsequent sweep is increased by 5dB until the unit being tested shows severe compression, or exhibits audible distress noise. In some cases a 2 or 3dB higher sweep level will be attempted. This is a very demanding test. The PB13-Ultra exhibits good control and tracking of the signal up until the 105dB sweep. During the 110dB sweeps there starts to be some port compression in evidence near the tuning frequencies and perhaps some additional amount due to voice coil heating since ported subs do not cool themselves well at tuning. The sealed mode has some limiting going on presumably to keep the driver from over driving itself below 25Hz. (Bass reflex alignments have a large reduction in excursion near the tuning, while sealed alignments rely solely on the drivers displacement for output.) The sealed mode exhibits severe compression and limiting once the sweep level is pushed up to an 113dB nominal level. The 15Hz ported mode completed a nominally 115dB sweep but also was clearly into the limiter and compressor by then. In 20Hz mode the 115dB sweep did not show as much compression as in 15Hz mode so an even higher 118dB sweep was tried with virtually no improvement in the output, which clearly shows that it was being severely clamped at that point. Eventually every system runs into its output limits. For the PB13-Ultra these are quite high.

 svs pb13 ultra mode compression comparo.jpg

SVS PB13-Ultra: 110dB Power Compression Comparison (20Hz, 15Hz and Sealed Mode)

In the comparison of compression exhibited by the different operating modes during their respective 110dB nominal sine sweeps you can see that above 30Hz the results are basically unchanged. Below that point is where most of the response and changes in loading on the driver will occur and that is where we see the biggest difference between them. The two ported modes exhibit the most obvious effect centered at their tuning frequencies where port compression occurs and voice coil heating is high. The sealed alignment is probably engaging the limiter circuit already at this sweep level. Overall the PB13-Ultra delivers a solid performance here in all 3 modes. Above 20Hz the amount of compression is less than 2 dB even at this very high output level. 

SVS PB13 ULTRA MAX LONG TERM COMPARISON.PNG

SVS PB13-Ultra: Maximum Sweep Level Comparison (20Hz, 15Hz and Sealed Mode)

Looking at the comparison of the maximum level long term sweeps completed in each mode, in 20Hz configuration the PB13-Ultra has the highest long term output everywhere above about 18Hz and it is a substantial amount of output indeed. Remember that this is a 2 meter ground plane measurement. If it were a 1 meter measurement you would add 6dB across the board and the overall output would be in the 115-120dB range from 20-120Hz with great bandwidth uniformity. The 15Hz and sealed modes offer greater ultimate extension but at the cost of overall output above 18Hz. Throughout this testing the PB13-Ultra was poised and composed. There was a bit of port noise in the 15Hz mode at the highest sweep level and a lesser amount in the 20Hz mode. I doubt either would be noticed in real world circumstances with masking content. I never noticed any at all during the listening sessions and it was driven quite hard at times. The sealed mode did exhibit just a bit of a doubling sound below 25Hz as the driver was driven very hard on the final 2 sweeps. If I had to guess the driver was producing well over 2” of peak to peak excursion at that point. Other than that the SVS was steady as a rock and never faltered or shut down. It never even really sounded distressed, at least not compared to other units that I have heard run through this test, only producing the slight amount of noise noted above. Its protection circuits work very well and are tuned to allow as much output as reliably possible while keeping the sound from deteriorating completely. I prefer this approach as it allows the unit to maintain maximum output and dynamic tracking ability with perhaps a bit more distortion in the low bass frequencies, as opposed to severely clamping the output and dynamics in order to ensure that the sound is always pristine no matter the input.

SVS PB13 ULTRA THD COMPARISON 105DB.PNG

SVS PB13-Ultra: 105db Sweep THD Comparison (20Hz, 15Hz and Sealed Mode)

The sealed operation mode has higher levels of distortion during its 105dB sweep than the 2 ported modes and is also exhibiting lower output levels for the most part. It does offer higher output with lower distortion in the very deepest bass frequencies but I consider the performance above 20Hz of much greater importance. Still in sealed operation the PB13-Ultra stays under 10% THD until below 24Hz at this drive level. In its ported operation modes the PB13-Ultra is exhibiting exemplary distortion performance even at very loud levels.

SVS PB13 20HZ CEA2010 PASS CHART.PNG

SVS PB13-Ultra: CEA2010 2 meter Groundplane RMS Results (20Hz Mode)

PB13 ULTRA 15HZ CEA2010 CHART PASS.png

SVS PB13-Ultra: CEA2010 2 meter Groundplane RMS Results (15Hz Mode)

PB13 ULTRA SEALED CEA2010 CHART PASS.PNG

SVS PB13-Ultra: CEA2010 2 Meter Groundplane RMS Results (Sealed Mode)

SVS PB13 ULTRA CEA2010 COMPARISON.PNG

SVS PB13-Ultra: CEA2010 2 Meter Groundplane RMS Comparison (20Hz, 15Hz, Sealed Mode)

CEA2010 Results

Using the CEA2010 test procedure the PB13-Ultra exhibits impressive output capabilities. The sealed mode offers the least output and highest deep bass distortion, most notably below the 31.5Hz band, but it is still pretty capable in that configuration. The 2 ported modes offer a lot of output and are strong down to the 16Hz band (105.1dB for the 15Hz configuration). In the sealed and 15Hz ported mode the PB13-Ultra was even able to record a passing CEA2010 result at 12.5Hz which while at a much lower output level, is still notable. The 20Hz mode offers impressive bandwidth linearity fitting within what amounts to a 7dB window from 20-125Hz. This is a very good performance that many subs would be hard pressed to match. Most will exhibit a more tilted response profile with decreasing output at lower frequencies as they are either EQ’d heavily to boost the low end or lack the necessary port area and driver displacement needed to maintain deep extension and loud output levels. Looking at the overall distortion performance it is quite good with low percentages maintained until down into the very deep bass frequencies. When distortion does start to rise it is dominated by the third harmonic for the most part which is not unusual. If the distortion threshold is removed completely the SVS sub can produce a bit more output below 20Hz. The maximum output is ultimately limited by the amplifier and compressor circuit so no matter what signal happens to show up at the inputs to the PB13-Ultra, it will not damage itself. Overall this is an impressive scorecard that indicates that this subwoofer has plenty of clean headroom to tackle most situations it will encounter. SVS has leveraged the extra power of the new Sledge amplifier in conjunction with masterful execution of the compressor and limiter circuits to extract as much output as possible from the PB13-Ultra.

Editorial Note

While working with Gene DellaSala to develop and establish how we would approach the reviewing and testing subwoofers for Audioholics going forward, we sought opinions and feedback from many respected people in the industry, one of which was Ed Mullen of SVS, who himself was once a professional reviewer and tester of subwoofers. There had also lately been some controversy over testing subwoofers per CEA2010 and whether or not the results were repeatable between different test equipment and reviewers. In light of this we thought it best to get a “control” check on my test equipment and calibration with the first subwoofer review. In an effort to make sure things are where they needed to be, we chose an SVS subwoofer as our first product to test under our new measurement protocol. SVS has long been a supporter of third party reviews and testing of subwoofers. The PB13-Ultra, since being released in its initial configuration with the BASH 750w amplifier years ago, has been reviewed many times, tested thoroughly by multiple people and rightly established a reputation as a very good performer, with the measurements to back it up. Since the PB13-Ultra has lately been upgraded to the new Sledge DSP amplifier platform, which is claimed to produce improved performance over the original configuration and since SVS also conducts internal testing, including CEA2010, that we could compare our results against, we thought it seemed like the logical choice to approach them for a review sample. SVS promptly agreed to supply a unit for the review. The good news is that when my initial CEA2010 results for the review PB13-Ultra were sent to SVS for them to compare with their internal data, it turned out that the maximum SPL variance between the two data sets at any center frequency was a mere 0.7dB! A couple of the data points were exactly the same! The results for all three modes were confirmed to be very close to the SVS internal data. In light of the possible differences and tolerance variations in: equipment, calibration, atmospheric conditions, test site, unit to unit variation, etc., this is a very tight match. I do not think we could ask for a better confirmation that the CEA2010 results presented in this and future Audioholics reviews will be as close to accurate and consistent as is feasible.

 

Confused about what AV Gear to buy or how to set it up? Join our Exclusive Audioholics E-Book Membership Program!

Recent Forum Posts:

dmusoke posts on February 15, 2014 02:36
fuzz092888, post: 1016877
Based on the CEA results, some of the other graphs, and my own personal preferences/experiences I would go with the 20hz tune, at least based on the raw numbers. The 20hz tune is going to have lower distortion overall, in room gain will boost the extension you get slightly and you'll get more output/have more headroom over the 15hz tune.

As to the distortion question, at moderate volumes with distortion far less than 5%, I doubt you'd notice a difference between either mode. At higher volumes, depending on the content you may notice that extra bit of extension, you may not. Distortion may begin to come into play, but I doubt it. Both tunes will, by my estimation still sound similar. At full on reference playback with a demanding soundtrack is where you'd probably notice the biggest difference. Since 20hz and below is more felt than heard, I'd still probably go for 20hz tune since I'm not really seeing enough meaningful output over the 20hz tune to risk rising distortion while also sacrificing output and increasing thermal compression. IMHO of course and I'll throw in the disclaimer that I've never actually heard this particular sub so my opinion might be different if I did get to audition it. Then again maybe not as well.

I agree… Thank you so much Fuzz for the help you've provided explaining to me the details i was confused about .
fuzz092888 posts on February 13, 2014 01:26
dmusoke, post: 1016873
I see what you are saying now with the graphs. You're observations are correct in this regard. Is distortion of 10% or greater at 20Hz or lower audible? If so, then I'm tempted to think that the 15Hz mode is best overall based on the graphs iwth THD less than 5% up to the tuning frequency?

Based on the CEA results, some of the other graphs, and my own personal preferences/experiences I would go with the 20hz tune, at least based on the raw numbers. The 20hz tune is going to have lower distortion overall, in room gain will boost the extension you get slightly and you'll get more output/have more headroom over the 15hz tune.

As to the distortion question, at moderate volumes with distortion far less than 5%, I doubt you'd notice a difference between either mode. At higher volumes, depending on the content you may notice that extra bit of extension, you may not. Distortion may begin to come into play, but I doubt it. Both tunes will, by my estimation still sound similar. At full on reference playback with a demanding soundtrack is where you'd probably notice the biggest difference. Since 20hz and below is more felt than heard, I'd still probably go for 20hz tune since I'm not really seeing enough meaningful output over the 20hz tune to risk rising distortion while also sacrificing output and increasing thermal compression. IMHO of course and I'll throw in the disclaimer that I've never actually heard this particular sub so my opinion might be different if I did get to audition it. Then again maybe not as well.
dmusoke posts on February 13, 2014 00:25
I hope the new SB13 gets reviewed soon…
dmusoke posts on February 13, 2014 00:24
fuzz092888, post: 1016655
Looking more closely at the graph, you'll see that the 15hz mode doesn't offer lower distortion than the 20hz mode, except below the tuning point. The swept tone graph isn't mislabeled.

This graph shows the same thing except with a 110db tone. The distortion of the 15hz mode starts to raise higher than the 20hz mode at around 55hz and continues to rise higher than the 20hz mode until the 20hz mode gets close to the tuning frequency. Below the tuning frequency, the 15hz mode has lower distortion, as it should, since it is tuned lower. The tradeoff for that extra extension is rising distortion starting at 50-60hz. The CEA results show the same thing.

Data-Bass

I see what you are saying now with the graphs. You're observations are correct in this regard. Is distortion of 10% or greater at 20Hz or lower audible? If so, then I'm tempted to think that the 15Hz mode is best overall based on the graphs iwth THD less than 5% up to the tuning frequency?
fuzz092888 posts on February 12, 2014 08:37
Looking more closely at the graph, you'll see that the 15hz mode doesn't offer lower distortion than the 20hz mode, except below the tuning point. The swept tone graph isn't mislabeled.

This graph shows the same thing except with a 110db tone. The distortion of the 15hz mode starts to raise higher than the 20hz mode at around 55hz and continues to rise higher than the 20hz mode until the 20hz mode gets close to the tuning frequency. Below the tuning frequency, the 15hz mode has lower distortion, as it should, since it is tuned lower. The tradeoff for that extra extension is rising distortion starting at 50-60hz. The CEA results show the same thing.

Data-Bass
Post Reply