HSU VTF-15H CEA Test Results

The original Audioholics 2010 Subwoofer Shootout Measurements procedure differ slightly from the CEA 2010 standard, because as we have explained here, in order to get a more even contribution at the relatively close microphone distance of 1 meter from the ports (typically located on the bottom of the cabinet), and the woofer (usually above the ports), it is best to place the microphone at a distance equal from all acoustic radiators (woofers, ports and passives all).  To accomplish this,  we laid the cabinet on its side to acquire measurement data when the ports and cabinet are both mounted on the front baffle of the system.  We also place the mike, not with respect to the geometric center of the cabinet, but with respect to the acoustic center of the cabinet, which we estimate as halfway between vertical lines running through the ports and driver.  These procedures are to produce the most accurate results, with the smallest possible errors due to differences in distance from the microphone location to the radiating elements (port(s) and speaker(s).  This method was not only applied to the Hsu system measured, but ALL THE SYSTEMS with front mounted driver and ports.  CEA 2010 also includes agreed upon methods for measuring systems with the driver and vent (or passive) mounted on the side surfaces, and we have pointed out to CEA (unofficially) these agreed upon configurations puts those type of systems (for example side firing systems) at a disadvantage when measuring maximum SPL by placing the acoustic centers of the radiators (speakers and ports or passive radiators) FARTHER away from the specified location of the ground plane microphone relative to front mounted systems.  Placing the microphone farther away from the acoustic center would give those systems with front mounted elements an advantage in the test for scoring the highest SPL numbers.  Because most listener are considerably farther from the system than the 1 Meter distance used for measurement, this is a measurable difference not likely to be heard in a real listening environment.  Again, this point covered in graphic detail here.

After spending an hour listening to a salesman pitch his factory where I was conducting the subwoofer tests (always a joy to be sure) I went back outside and discovered the Hsu subwoofer was still in one piece not having been run over by a truck.  (This was truly my lucky day).

Seeing said trucks go rolling over where my $600 microphone was laying some 30 seconds after removing it, I have made a habit of taking it out of the sun, and way over to the edge of the building where the equipment rack is located so it does not heat up in the sun.

That said, the Hsu subwoofer was now in the beautiful southern California midday weather, and the ambient temperature had increased by perhaps 10 degrees F, while the humidity dropped substantially.  My first order of business was to take another curve without having made any changes at all to the equipment settings.  I returned the mike to its rightful place 39.37 inches directly in front of the cabinet (grille was off, so the measure was taken to plane of the baffle board, not the grille,  and laid the microphone down on the ground.  Below is the resultant curve...

Same Drive voltage into the amplifier – 1 hour later, weather changed amp was warmer
Orange curve at around 10 AM -  Blue Curve around 11 AM

Clearly the change in the temperature of the weather and the amplifier had caused a shift in the gain of the amplifier.  This does not appear to be speaker related at all, or the difference between the curves would not be frequency independent.

My next step was to repeat the CEA testing, but this time with the CEA standard orientation, vertical orientation (ports on the bottom).   The program allows one to gather the data one sample at a time, or once one has discovered the “maximum” drive you can employ before failing the test, to run the signal several times and gather an average.  Since it appeared that in some cases, a drive voltage resulting in a failure of the speaker with a single pulse, actually passed the test when an average of 8 pulses were used, both methods were employed to make every effort to find agreement with the higher SPL numbers Hsu Research was claiming.  For sake of reference, below is the table from the original data presented.

Max SPL Output Data of the Hsu Research VTF-15H Subwoofer
                      Data taken November 11^th, 2010
                            box orientation: horizontal

It is worth noting again, the data above was taken with the cabinet on its side, and both ports open.  As per Dr. Hsu's instructions, the Q was set to 0.7 and the EQ was set to position 2, which raises the frequency of the subsonic filter and boosts the gain and available output in the 30 Hz region at the expense of a very small amount of bass in the 20 Hz region.  Below are for completeness the distortion curves for the above data as originally shown in the first Hsu VTF-15H review.

Hsu Research VTF-15H CEA-2010 Distortion Test Results

The retest data below was taken with the cabinet upright, and one port blocked.  The Q control was still set to 0.7, and the EQ was set to position 1, recommended for 1 port blocked.  In this orientation, the microphone is closer to the ports, and further away from the woofer.  The amplifier is also now warm, and has been on (without playing) for about 1.5 hours prior to re-testing.

       First retest of Hsu VTF-15H February 9^th, 2011 (1) Pulse Maximum
                                     Box orientation: vertical

Below is the concurrent distortion data, collected as before, except for cabinet orientation.

 Distortion data for first retest of Hsu VTF-15H February 9^th, 2011
(1) Pulse Maximum - Box orientation: vertical

So far the theory that putting the mike closer to the ports, lowering the tuning, and taking the extension over the maximum output seems to be coming at a pretty high price.  How much of this is due to the amplifier loses power with an increase in temperature versus a change in the weather is not possible for me to say without further additional testing under a multitude of conditions.  Below are the results with the same test conditions save for the program being set to run (8) pulses (1) second apart so as to improve the signal to noise ratio of the measurement.

Second retest of Hsu VTF-15H February 9^th, 2011 (8) Pulses averaged
Box orientation: vertical

Distortion data for second retest of Hsu VTF-15H February 9^th, 2011
(8) Pulses averaged - Box orientation: vertical

The final test phase was done after mike re-calibration, and seems to improve the output by approximately 0.5 to 0.6 db regardless of frequency relative to the second test, done with the prior calibration.

Third retest of Hsu VTF-15H February 9^th, 2011 – mike re-calibrated & (8) Pulses averaged
                                             Box orientation: vertical

 Distortion data for third retest of Hsu VTF-15H February 9^th, 2011
mike re-calibrated & (8) Pulses averaged - Box orientation: vertical

Despite the small improvement seen at 20 Hz, as expected, the cost was in sensitivity at all the higher frequencies.  There is nothing new or noteworthy here.  This is basic loudspeaker physics 101.  That the data is different, and also less than it was in the original test comes as no surprise to this author.   That the test results are only repeatable within about 1 db is completely in keeping with my experience using the CEA test program.  The consistency of this data is very good, and all the numbers agree with what is to be expected from systems of this size and power.