Introducing the Falcon-III
The Falcon-III is a new addition to my Baffleless Concept Speakers. This time round, the heart of the system is a DMN-A Dome Midrange by HiVi.
The major difference between the DMN-A and the Falcon-II’s Dayton RS52AN is the diaphragm. The DMN-A diaphragm is made of fabric whereas the RS52AN is of aluminum. Because of that, the DMN-A doesn’t exhibit the nasty cone breakup. Furthermore, the bandwidth is also much wider.
As for the bass, I opted for the HDS 830869 by Peerless. This is an 8″ woofer with the highly regarded Nomex Cone. It is housed in a 35 liters vented box tuned to 40Hz. This will result in a deeper bass compared to the Falcon-II’s Dayton RS180s.
The last driver to address in this 3-way is the tweeter. I was tempted to replace the Peerless XT25SC90 with perhaps one of my ribbons but I couldn’t find any good reason except from being glamorous. So, the tweeter will stay the same for now.
HiVi DMN-A Dome Midrange
Fig 1 – DMN-A RAW Frequency Response
Fig 1 is the response of the HiVi DMN-A without a baffle. According to the manufacturer’s specs, the effective frequency range is from 700Hz~8kHz. What I got is from 1kHz~8kHz. The frequency plot, however, is similar to HiVi.
Fig 2 – DMN-A RAW and Bandpass Responses
The Blue plot in Fig 2 is the DMN-A with a bandpass filter. My corner frequencies are 1kHz on the high pass and 3.5kHz on the low pass. This narrow bandwidth is typical of 2″ dome midranges. Personally, it would be great if it can extend down to about 300Hz. For that I’ll need to resort to the Peerless GBS85N25PR03-04, a 3-1/2″ Midrange. The problem with the GBS85 is it needs a 7 liters sealed box. A baffleless design won’t work.
Peerless XT25SC90-04 Ring Radiator Tweeter
Fig 3 – XT25SC90-04 RAW and High Pass Responses
The Black plot in Fig 3 is the RAW response of the Peerless XT25SC90-04 tweeter. Surprisingly, this tweeter has quite a wide bandwidth. The Red plot is with my High Pass filter.
Integrating the DMN-A with the XT25SC90
Fig 4 – HiVi DMN-A with Bandpass and Peerless XT25SC90-04 with High Pass
Fig 4 shows the responses of the DMN-A and the XT25SC90-04. The two drivers are crossing acoustically at about 3.5kHz. This is quite encouraging as my targeted crossover frequency is 3.5kHz.
Fig 5 – Summed Response of DMN-A with XT25SC90-04
The Black plot in fig 5 is the summed response of the DMN-A and the XT25SC90-04. Even though there are no cancellations observed, the summing is uneven in that there’s more on the right than the other.
Fig 6 – Null Response of the DMN-A with XT25SC90-04
The Null response in Fig 6 flags the less than optimal summing of the midrange dome and the ring radiator tweeter. It appears that there’s an overlap at 5kHz, resulting in a hump instead of a deep notch. I’m not too concerned about this because my ultimate aim is the step response.
DMN-A and XT25SC90-04 Step Response
Fig 7 – Step Response of the DMN-A with the XR25SC90-04
Fig 7 is what I’m after, a clean and fast transient. This shows the DMN-A and the XT25SC90-04 are close to being Time and Phase Coherent. Not 100% but close enough for me.
Easing in the Peerless 830869 woofer
Fig 8 – Peerless 830869 RAW and Low Pass
The Black plot in Fig 8 is the RAW response of the Peerless 830869. This is with it mounted onto a 35 liters vented box having a baffle width of 10-1/2″. The Brown plot is with a Low Pass filter with a corner frequency of 1kHz. Note that the cone breakup peak at 5kHz is eliminated in the Low Pass roll-off.
Fig 9 – Peerless 830869 Low Pass with DMN-A and XT25SC90-04
Fig 9 shows the acoustic crossover of the 830869 with summed response of the DMN-A and the XT25SC90-04. It appears they are crossing at 1.5kHz as opposed to the 1kHz that I intended.
Fig 10 – Summed Response Overlay
The Black plot in Fig 10 is the summed response of the 830869 with the mid/high combo.The summation in the passband looks even and symmetrical. That indicates optimal summing.
Fig 11 – Null Response of 830869 with DMN-A and XT25SC90 combo
The beautiful deep notch in Fig 11 indicates the 8″ Peerless is phase aligned with the mid/high combo. This is important as our hearing is most sensitive in the midrange.
Fig 12 – Falcon-III Frequency Response
The final frequency response of the Falcon-III is shown in Fig 12. It is exceptionally flat throughout except for a light depression at 3kHz. Measurements below 500Hz are in nearfield.
Falcon-III Step Response
Fig 13 – Falcon-III Step Response
The Falcon-III Step shows the 830869 woofer is slower than the mid/high combo by about 700 microsec. This is to be expected because the acoustic center of the woofer is farther back.
Fig 14 – Falcon-III Waterfall
Fig 15 – Falcon-III Toneburst Energy Storage
In the Waterfall plot in Fig 14, it appears that the treble has virtually no artifacts. This is reaffirmed in the Toneburst plot in Fig 15. It is rare to see such a clean plot.
Fig 16 – Falcon-III Spectrogram
The Spectrogram of the Falcon-III doesn’t flag any anomalies in the response. Generally, the Falcon-III is very clean judging from the Waterfall, Toneburst and Spectrogram.
Auditioning the Falcon-III
The Falcon-III sounds vastly different from it’s predecessor, the Falcon-II. The sound is bigger. I attribute this to the Peerless 830869. The bass from this woofer is clear and detailed. At no time did it interfere with the vocals.
The midrange clarity is superb. The 830869 does a fantastic job with the lower midrange while the DMN-A handled the upper mids with ease. The crossover from the 830869 to the DMN-A is seamless. It is impossible to tell the vocals are actually from two drivers.
Further up the spectrum, sibilance is well controlled and sounds natural. There are no screeching female voices. I made sure of that when I was working on the crossover.
As for the treble, I really have no complaints. The XT25SC90-04 may be a bit dated but it still does a fine job. No harshness in the highs.
In the final analysis, I’m really satisfied with how well the Falcon-III turned out. I found it soothing. No listener fatigue, even listening to it from morning to night. The Falcon-III is worthy of advancing my Baffleless Concept Speakers line.
Unless otherwise stated, all measurements were made with the mic at 36 ins, tweeter axis. Impulse Window=5ms. No smoothing applied.
May 5, 2020Projects