KR65 Kevlar Cone Woofer with Vifa PL27TG-35-06


the KR65 – 6.5″ Kevlar Cone Midwoofer
Active Bi-amped

In this test, I shall start off with a proprietary 6.5″ woofer, the KR65. I had a few hundred made to my specifications in the late 1990s for a project that I was involved in. It has a kevlar cone, cast frame, rubber surround and a bucking magnet at the rear. After 20 years in storage, it is still perfect. The surround did not deteriorate, no oxidation on the frame and the kevlar cone is still like new.

During that time, I commissioned another manufacturer for the enclosure. The material is 3/4″ MDF with black oak vinyl wrap. All measurements are made with the KR65 in this 18 liters bass reflex enclosure. Baffle width is 9.5″.

KR65 RAW Frequency Response

KR65 RAW FREQUENCY RESPONSEFig 1 – KR5 RAW Frequency Response. Measurements below 500Hz in Nearfield.

The Black plot in Fig 1 is the RAW response of the KR65. No crossovers are used. From this sweep, we can see the kevlar cone breakup starting at 3.5kHz. This is not really an issue for me because I intend to cross at about 2.5kHz.

KR65 @ 2.6kHz 24dB/octKR65 RAW 2K6 24dB

Fig 2 – Black plot = RAW. Blue plot = 2.6kHz (24dB/oct) LPF

The Blue plot in Fig 2 is the KR65 crossed at 2.6kHz (24dB/oct). As I had hoped, the cone breakup is well suppressed. 


Fig 3 – RAW Response of the KR65

Fig 3 is the Toneburst Energy Storage of the KR65 without any crossover. The bunch of light blue slices from 3kHz~9kHz are the stored energy. This is a 3-dimensional view of the cone breakup. 

KR65 TES 2K6 24dB/OCT

Fig 4 – KR65 at 2.6kHz (24dB/oct) LPF

Fig 4 is when I applied a crossover at 2.6kHz (24dB/oct). The cone breakup disappears completely. I am now left with a very clean woofer response.

REW Cumulative Spectral DecayWATERFALL KR65 2K6 24dB/oct

Fig 5 – Waterfall plot of KR65 at 2.6kHz (24dB/oct) LPF. Rise Time = 0.10 msec. Window = 1.3 msec ( 769Hz ). Time Range = 1.0 msec

Fig 5 is the Waterfall plot of the K65 with a LPF at 2.6kHz (24dB/oct). There are virtually no artifacts from 2.6kHz onwards. Now that the KR65 is measured, let’s take a look at one of my favorite tweeters, a vintage 1″ Silk Dome Vifa Premium Line PL27TG-35. In it’s day, this is the top tweeter from Vifa, Denmark.

Vifa PL27TG-35-06 Frequency ResponseVIFA PL27TG-35-06 RAW 2K6 FR

Fig 6 – Black plot = RAW. Red plot = 2.6kHz (24dB/oct) HPF

The Black plot is the RAW response of the PL27. The Red plot is with it crossed electronically at 2.6kHz (24dB/oct). I did not install a conjugate network yet as the tweeter’s impedance peak is well damped at it’s resonant frequency.

Toneburst Energy StorageVIFA PL27TG-35-06 TES 2K6 HPF

Fig 7 – PL27 with 2.6kHz (24dB/oct) HPF

Fig 7 is the Toneburst Energy Storage of the PL27 high passed at 2.6kHz (24dB/oct). It recorded some minor stored energy from 2.5kHz~6kHz. In the upper treble, the PL27 is extremely clean.

REW Cumulative Spectral Decay

VIFA PL27TG-35-06 WATERFALL 2K6 HPFFig 8 – Waterfall plot of PL27 at 2.6kHz (24dB/oct) HPF. Rise Time = 0.10 msec. Window = 1.3 msec ( 769Hz ). Time Range = 1.0 msec

In the Waterfall plot in Fig 8, 5kHz registered the longest decay. From 6kHz onwards, the decay is much smaller and getting weaker as it moves towards 20kHz. This is in line with what the Toneburst Energy Storage plot showed. 

Harmonic DistortionDISTORTION VIFA PL27TG-35-06 2K6 HPF

Fig 9 – Distortion of PL27. Red plot=2nd Harmonic. Violet plot=3rd Harmonic

Fig 9 is the Harmonic Distortion of the PL27 with a high pass filter of 2.6kHz (24dB/oct). I am relieved to find there are no peaks at the Toneburst Stored Energy region. This indicates the artifacts are benign.

Crossing the KR65 with PL27

KR65 LPF PL27 HPF 2K6Fig 10 – Blue plot = KR65 with LPF 2.6kHz (24dB/oct). Red plot = PL27 with HPF 2.6kHz (24dB/oct)

Fig 10 shows the acoustic crossover of the KR65 and the PL27. Eventhough my electronic crossover is set at 2.6kHz, the two drivers are actually crossing at 2kHz. This is not because of the PL27 but rather the KR65 response rolling off earlier.

Summed Response of KR65 and PL27 in Reversed Phase

SUM KR65 LPF PL27 HPF 2K6Fig 11 – Black plot = Summed Response of KR65 and PL27

The Black plot in Fig 11 is the summed response of the KR65 and the PL27. We can see the summation is very good. There are no cancellations on either side of 2kHz. For correct summing, the PL27 tweeter must be wired in Reversed Phase.

Frequency Response of KR65 and PL27 in Reversed Phase

FR KR65 LPF PL27 HPF 2K6Fig 12 – Final Frequency Response

Fig 12 is the final frequency response of the KR65 and the PL27. The summation at the 2kHz crossover is perfectly flat. However, farther up at 5kHz, there is a slight peak. This is seen in the Toneburst Energy Storage of the PL27 in Fig 7. My guess is this is not from the PL27 tweeter but is diffraction at the edge of the box.

Crossover Null

KR65 LPF PL27 HPF 2K6 CROSSOVER NULLFig 13 – Crossover Null when PL27 wired back to Normal Phase

To check on how close the acoustic centers of the KR65 and the PL27 are, I wired the PL27 back to normal phase. That resulted in a beautiful null centered exactly at 2kHz (Fig 13). This is very close to being Time Aligned.

Toneburst Energy Storage

KR65 LPF PL27 HPF 2K6 TESFig 14 – Toneburst Energy Storage of KR65 and PL27

Fig 14 is the Toneburst Energy Storage of the KR65 and the PL27 at 2.6kHz (24dB/oct). There are virtually no extra energy except for those from 4kHz~6kHz which is the 5kHz peak seen in the Frequency response plot in Fig 12. To eliminate this burst of unwanted energy caused by baffle edge diffraction, I will have to re-design the enclosure. Perhaps chamfering the sides around the PL27 at 45° will do the trick. 

the KR65 + PL27 in playback

All the while, I have been listening to this speaker at low volume background music while I’m writing this article. Now that I’m auditioning it, I decided to increase the volume to a level that I normally used. And mine, what a difference. The music came alive.

What struck me most is the improved dynamics and character of the bass. The bass notes are more “meaty”. This is not “bloom” but solid body. Out of curiosity, I rigged up the microphone to check the spl. With the mic at 1 meter tweeter axis, it recorded peaks at 95dB. The album I’m listening to is by Jamie Valle – ‘Round Midnight.

I then switched over to Diana Krall. Playback volume remains the same. I want to see whether her voice sounds over sibilant. In the Girl in the Other Room, Diana Krall sounds perfectly fine. What ever sibilance are there, they sound natural and not exaggerated. Cymbals sound smooth too, not glass like.

Lastly, I put on Santana – Supernatural. I want to see how the speaker sounds like with a fast track. With (Da Le) Yaleo, the KR65 kept up with the pace of the music.

Having tested and listened to the KR65 with the PL27, I’m more than satisfied with their performance. I am tempted to have a speaker of this quality made commercially. It will give a lot of speakers a run for the money.

Note: All measurements were made with the mic at 1m, tweeter axis. Impulse Window=5ms. No smoothing applied.