The Swallow-AP is the passive version of my active bi-amped evaluation that I did in Aug, 2020. The Swallow’s main driver is an unusual 3-1/2″ dome midrange, a Peerless GBS-85N25PR03-04. This midrange has a very slim profile, an incredible 0.6 ins deep. She is ideal for tight places like in-wall mounting or in a car door. However, that is not what I bought the GBS85N for. It’s more for her short acoustic center. She may be easier to align with the tweeter.
In this Swallow-AP, the tweeter is an Audiopur PLA114-6 ribbon. I don’t really know whether this tweeter is still available as there’s not much information available. She’s not the best tweeter to mate with the GBS85N but since they are already mounted, I might as well work on the crossover. Chances are I will replace the PLA114 with something else later.
Fig 1 – Peerless GBS85N RAW Frequency Response • Baffle Width=8″
The plot in Fig 1 is the GBS85N in a 7 liter sealed box with a baffle width of 8″. She has a relatively flat response up to 4kHz, after that is the cone breakup. There is a small dip at 1.8kHz but it’s not too serious. What is fascinating is she doesn’t exhibit any effects of baffle step.
Fig 2 – Peerless GBS85N with Low Pass network
The Blue plot in Fig 2 is the GBS85N with a bandpass filter. The High Pass is a 170uF capacitor to roll-off the bass. On the other end, the low pass is at 2kHz with a 2nd order network.
Fig 3 -Peerless GBS85N with Audiopur PLA114-6 (Flush Mounted)
The Red plot in Fig 3 is the Audiopur PLA114-6 ribbon tweeter. This is not the best high pass but considering how horrible the raw response is, it’s not too bad. After all this is just an evaluation with a passive network. I don’t want to expend too much time on it, more so when I have no intention of using this ribbon tweeter in an actual build. My aim is to listen to these two drivers so that I have an idea of what their sonic signatures are like.
Fig 4 – Swallow-AP Passband
The Black plot in Fig 4 is the crossover passband. As expected, the summation has problems. There’s cancellations on the right side of 2kHz.
Fig 5 – Swallow-AP Frequency Response
The depression from 2kHz to 4kHz in the final frequency response (Fig 5) is the result of the crossover cancellation. I can correct that but that would mean spending more time, time that I can’t afford.
There’s also a 5dB emphasis from 5kHz onwards. Strangely, I did not pick up any brightness during playback. i think ribbons are quite forgiving when it comes to that.
Fig 6 – Swallow-AP Null
The Violet plot in Fig 6 is with the tweeter wires reversed. There is a null at about 1.7kHz which is off from the 2kHz where the two drivers are crossing. Note that from 2kHz to 5kHz, there is actually a summation. This is what crossovers look like when they are not correct. Imperfect crossover leads to imperfect summation.
Fig 7 – Swallow-AP Frequency Response
Fig 7 is the final frequency response of the Swallow-AP. It is far from perfect but she doesn’t sound bad during playback. In fact, as an end user with no knowledge of measurements, the Swallow-AP sounds perfectly fine.
Fig 8 – Swallow-AP Step Response
The Swallow-AP step response is in Fig 8. The GBS85N is quite fast, hitting the apex at just below 200 microsec. Her transient is a straight up. I don’t think I’ve seen anything like that before. On top of that, she exhibits a fast and graceful decay.
Fig 9 – Swallow-AP Waterfall
The Swallow-AP treble is very clean (Fig 9). Even in the upper midrange, there’s not much decay.
Fig 10 – Swallow-AP Toneburst Energy Storage
The Toneburst plot in Fig 10 confirms what is in the Waterfall plot. There’s almost no excess energy from 500Hz to 1.5kHz. There’s a bit of excess energy just before 2kHz but other than that, the Swallow-AP has very few artifacts.
Fig 11 – Swallow-AP Spectrogram
The Spectrogram in Fig 11 shows a delayed spot at 2kHz. That is exactly at the crossover frequency. I did not pick up any issues during playback, so nothing to worry about.
Fig 12 – Swallow-AP Harmonic Distortion
The Swallow-AP is dominated by 2nd harmonics (Red plot in Fig 12). The 2nd harmonics are a bit on the high side but again, I did not detect any distortion on playback.
Fig 13 – Swallow-AP Impedance
The Swallow-AP is not a difficult load for power amplifiers. She dips to 3Ω is at 300Hz but other than that, she is well above 4Ω.
Summary of Swallow-AP
Surprisingly, the Swallow-AP sounds decent. Even with the horrendous +5dB bump in the upper treble, there’s no harshness.
To take care of the bass, I supported the Swallow-AP with the Palila. The Palila is a triple chamber bandpass sub with two units of Dayton DA135-8 inside. I didn’t need to add any low pass filter because the pipe resonance is low enough. The end result is delightful.
Obviously the Swallow is still a work in progress. The Audiopur ribbon tweeter will have to go. Pity. She does sound nice but her response is anything but flat. Stay tuned.
Unless otherwise stated, all measurements were made in Full Space (4 pi) with the mic at 36 ins, tweeter axis. Impulse Window=5ms. No smoothing applied.
March 22, 2021Projects