|
Starling-TBFC The Starling-TBFC is the bass reflex version of the Dual Chamber Reflex Starling-DCR. It is for listeners that prefer a smaller box. In this version, I decided to simplify the crossover slightly. There’s no compromise in the sound quality. The advantage is a reduction in component count.
The Black plot in Fig 1 is the RAW response of the Seas ER18RNX. It has a beautiful flat midrange that extends to 3kHz. If nothing is done to lower the midrange, the speaker will sound mid heavy. This is a common complaint with the ER18. After some adjustments, the Blue plot is the response of the ER18 with a new low pass network. This network is not the same as the one in the Starling-DCR. I re-worked it for a flatter midrange in the crossover passband.
The Red plot in Fig 2 is the Seas 27TBFC tweeter with a new network too. The two drivers are crossing at about 2kHz which is my new target.
The Black plot in Fig 3 is the passband of the crossover. It indicates good summing as no cancellations are observed. However, there is a slight dip at 5kHz caused by the cone breakup of the ER18.
Fig 4 is the final frequency response of the Starling-TBFC. This response is flatter than the Starling-DCR. Where it differs is the slight dip at 5kHz. I didn’t have that in the Starling-DCR because I tamed the ER18 peak at 5kHz. To reduce components in this version, I left out the network
When I flipped the tweeter wires around, it resulted in a deep null (Fig 5). The tip of the null is at 2kHz, which is exactly where the two drivers are crossing (Fig 2). This is as good as being time-aligned.
The Starling-TBFC has a beautiful Step response (Fig 6). The sharp tip at the bottom indicates the ER18 and the 27TBFC are crossing perfectly. The transient is linear and fast. It hit the top at 300 microsec. I have no idea why the tip is not sharp as in the Starling-DCR.
The Waterfall (Fig 7) and Toneburst (Fig 8) plots are quite normal. There are some excess energy at 1kHz~2kHz which last for 8 cycles.
The Specrogram (Fig 9) gives a better idea of the excess energy. There’s a hot spot at about 1.2kHz, followed by some delayed spots. They are harmless as they are dissipated by 5 msec.
The Excess Group Delay is interesting. The Starling-TBFC registers only -1.52 msec at 50Hz. Sound of Starling-TBFCI forgot how much I missed the Seas ER18RNX until I heard the Starling-TBFC. The ER18 is miles ahead of the Peerless 830874, 835025 and the SB Acoustics SB17NRX. Everything is right with this driver. She just sounds so natural. I would even dare say the ER18 is the entry into high-end. In all my previous work with the ER18, I was using a 15 liters bass reflex. I was never really satisfied with the bass. The Starling-TBFC is the first time I used a 27 liters box and finally, the bass fell into place.
In my box simulation (Fig 11), this 27 liters box is slightly larger than the recommended volume of 21 liters. It is not an issue for me. It’s just that I have this test box in hand. Since the ER18 sounds great, there is no reason why it would not work in a 21 liters box. The Starling-TBFC did justice to my ER18. The next step is to mate her with a more expensive tweeter. Not that the Seas 27TBFC is not up to standard. By all accounts, I’m very happy with her. It is this never ending quest for excellence that drives me. 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. |
February 1, 2021Projects
Fig 1 – Black plot=ER18RNX RAW Response • Baffle Width=9-1/2″
Fig 2 – Blue plot=ER18RNX Low Pass • Red plot=Seas 27TBFC High Pass
Fig 3 – Black plot=Starling-TBFC Passband
Fig 4 – Starling-TBFC Frequency Response
Fig 5 – Starling-TBFC Null
Fig 6 – Step Response
Fig 7 – Waterfall
Fig 8 – Toneburst Energy Storage
Fig 9 – Spectrogram
Fig 10 – Excess Group Delay
Fig 11 – Starling-TBFC box modelling