Frequency Response of PL18WO-09 in 18L Vented Box

When a woofer is mounted onto a box, it is subjected to internal forces which are different from the manufacturers'. In this chapter, we shall compare the response of the PL18 when loaded with a 18L enclosure to the manufacturer's Frequency Response Plot (Figure 13).

Note that this plot was made with the PL18WO09 in a 320 Liters Box mounted in Infinite Baffle.

This is likely a pit measurement.

It is probable that the PL18 was mounted onto a baffle board which was then bolted over a 320L enclosure that was previously sunk into the ground.


(Figure 13) Vifa Frequency Response Plot of PL18WO09

A microphone suspended 1 meter overhead would be positioned for measurements. With no obstructions within a 30 feet radius of level ground, this would constitute an Infinite Baffle. Measurements down to 20Hz can be made successfully with this setup.

For manufacturers, this is possibly the only way to provide meaningful data. It is impossible for them to anticipate the various shapes and sizes of enclosures their products are going to be mounted on. Unfortunately for us, infinite baffle is hardly practical. The majority of speakers is a box of some kind. We will use LMS to help us detect any deficiencies in the frequency response when the PL18 is mounted onto a 18L box.


Near Field of PL18

Let's refer back to the scaled 1 meter plot of the Near Field measurement of the PL18 previously taken (Figure 14).

This plot, while effective below 100Hz, is not accurate for mids and highs.

Note the deep notch at 1,000Hz.

Rolloff is at about 2,000Hz whereas Vifa specifies it up to 4,000Hz.


(Figure 14) Scaled to 1 meter

Gated SPL Sweep

To determine the accuracy of the higher frequencies, a Precision 300 points, Gated SPL sweep is done on the PL18 (Figure 15).

A Gated Sweep is a very accurate measurement. It is quasi-anechoic, in other words, like done in an anechoic chamber but not really in one. It accomplishes this by opening the meter's "gate" for the primary signal to pass through, then closing the "gate" before the first boundary reflection reaches the microphone.

With the microphone placed 1 Meter away, Input Signal at 2.83V and Gate "Meter On" set for 2.3 mSec, the lowest frequency it can read to is about 500Hz. Note the breakup below 500Hz.


(Figure 15) Gated SPL Sweep of PL18WO09


Splice

With the Gated measurement done, the two graphs are "Spliced" together at 500Hz to form one continuous plot (Figure 16).

It is interesting to note that the curve below 500Hz was actually made Near Field, then reference to 1 Meter mathematically.

When spliced with the Gated curve made at 1 Meter, the Scaled Near Field measurements are at the correct SPL. We can conclude that the Scaling Factor is accurate.


(Figure 16) Spliced Scaled Near Field with Gated SPL


Smoothing

For a more uniform appearance, the new curve is Smoothed by 0.1dB (Figure 17). This irons out the "jaggedness" of the Gated measurements.

Though it is a composite, this graph is a good representation of the actual frequency response of the PL18 when loaded with a 18 Liters Box. We are now in a position to compare with Vifa's Frequency Response Plot (Figure 18) and determine the effects of box loading.


Effects of Box

In Figure 17, we can discount the frequencies below 1,000Hz as they are relatively flat and bear close resemblance with Vifa's.

Slightly higher up, there is an obvious notch at 1,500Hz. This is reflected in Vifa's plot in Figure 18. I can only deduce that it's inherent of the PL18.

From 2,000Hz upwards, we are met by some bumps followed by a horrendous chunk bitten off at 3,500Hz. This is obviously not from the PL18 as can be verified from Vifa's plot.

I will hazard a guess that the culprits responsible are internal standing waves and cone breakup causing havoc at the upper frequencies.

With this discovery, I would be greatly concerned if I plan to cross anywhere between 2,000 - 4,000Hz.


(Figure 17) Smoothed 0.1dB


(Figure 18) Vifa PL18WO09

But at 1,500Hz, I am quite confident that with a 4th order filter, these "distortions" will be sufficiently suppressed so as not to intrude with the tweeter's performance.

We shall now proceed to examine the PL27TG35 tweeter in greater detail.

 


60 Downes Street | Calais | ME 04619 | USA