|
Hummingbird-TA In Aug last year, I worked on the Hummingbird which is based on the Dayton DSA135-8. This is one midwoofer that I really like. It has a fabulous midrange coupled with a fast, dynamic upper bass. In the Hummingbird-TA, I integrated the DSA135 with a Dayton ND25FW tweeter. I opted for this tweeter because with the waveguide, the acoustic center is farther back. It will be easier for me to Time Align the two drivers without having to resort to a step. The box is the same as the earlier Hummingbird, a compact 7 liters bass reflex with a baffle width of 8″. Box tuning is at 60Hz, slightly higher than the DSA135’s Fs of 55Hz
The plots in Fig 1 show the two drivers crossing at 3kHz. I left the DSA135 peak at 8kHz unattended (Blue plot). At -20dB below the fundamental, I doubt it’s audible. The difficulty is not with the DSA135 but rather the ND25FW tweeter. It is not as friendly as I would like it to be. After a few tries, I eventually settled on a 3rd order network with a shelving network to flatten the response (Red plot).
The Black plot in Fig 2 shows the summing in the passband. No cancellation is observed, indicating proper summing of the two drivers.
To check on the summing, I reversed the phase of the tweeter. It resulted in a beautiful deep null centered at 3kHz (Violet plot). This time alignment was achieved without adding a delay network, using a step or tilting the front baffle.
Fig 4 is the final response of the Hummingbird-TA. Measurements below 500Hz are in Nearfield. No smoothing is applied. Microphone is at tweeter axis, 1 meter away.
To have an idea of how the bass responds in my lab, I added the Black plot below 500Hz. This measurement includes my room reflections. Disregard the null at 150Hz. That’s an anomaly caused by a floor bounce. The area of interest is from 100Hz ~ 55Hz. Looking strictly at the bass, the Hummingbird has a fairly healthy extension down to 55Hz. Bass sensitivity is about -3dB to -5dB less. Note that this measurement is in Free Space (4 pi Steradians). The bass will be louder in a smaller room or against a wall (Half Space).
Fig 6 is the Step response. The DSA135 transient is almost vertical, indicating the cone reacts to the electrical input very quickly. Even though the Hummingbird-TA is time aligned, the time of flight shows the DSA135 takes off about 200 microsec after the tweeter. This is attributed to the delay in the passive crossover network.
The Waterfall (Fig 7) and Toneburst (Fig 8) plots show some excess energy (light blue slices) but it’s nothing serious. What concerns me is between 1kHz to 2kHz.
The Spectrogram in Fig 9 gives a better view of what is happening from 1kHz to 2kHz. There is some excess energy which could be cone ringing but they dissipate by 5 msec. From 2kHz onwards, whatever artifacts in the tweeter are insignificant. the sound of the Hummingbird-TAFor those who treasure vocals, I strongly recommend the Hummingbird-TA. Voices are crystal clear. Female voices in particular, are reproduced exceedingly well. There’s no hint of shrillness in all the tracks I played. Amanda McBroom, Diana Krall, Jewel all came out flawlessly. The DSA135 may be a 5″ midwoofer but don’t let the small size fool you. You’ll be surprised at the bass. It’s tight, dynamic and articulate, the very qualities I look for in bass. As for the ND25FW, I can’t heap enough praise for this budget tweeter. It’s a bit of a pain to tame her but once done, she’s a winner. No harshness. In fact, the tweeter is largely invisible. She’s that natural. Unless otherwise stated, all measurements were made with the mic at 36 ins, tweeter axis. Impulse Window=5ms. No smoothing applied. |
Fig 1 – Blue plot=DSA135-8 LPF • Red plot=ND25FW-4 HPF
Fig 2 – Black plot=Summed Response
Fig 3 – Null Response
Fig 4 – Frequency Response of the Hummingbird-TA
Fig 5 – Nearfield and Room
Fig 6 – Hummingbird-TA Step Response
Fig 7 – Hummingbird-TA Waterfall
Fig 8 – Toneburst Energy Storage
Fig 9 – Spectrogram