Starling-MDT (the Final Cut)

ER18RNX with Morel MDT29

If there’s one thing I learned from working on the Myna-DXT, I can forget about getting the same kind of bass as in the U18RNX. It’s just not in the ER18RNX. What I can do is to make the bass louder and add some oomph. That is the aim of the Starling-MDT.

Fig 1 – Seas ER18RNX RAW response • Baffle Width=8-1/2″

Fig 1 is the RAW response of the ER18RNX in a 18 liters bass reflex box. The frequencies below 500Hz is my in-room response, meaning it includes reflections. Disregard the suck-out at 150Hz. It’s caused by a floor bounce in my setup. 

From this basic plot, we can see the midrange and treble is about 8dB above the bass (125Hz to 50Hz). If a crossover is made from here, the music will have lots of midrange and treble. Bass will be quite soft. Such speakers are very tiring to listen to.

What I’m going to do is to lower the midrange and treble as close to the bass as possible without losing the clarity in the vocals. 

Fig 2 – Seas ER18RNX with Low Pass Network

After much trial and error, I found the right level. The Blue plot in Fig 2 is the ER18 with a Low Pass Network. This plot is actually found by ear first. The measurement is after the fact.

Fig 3 – ER18RNX Low Pass • Morel MDT29 High Pass

The Red plot in Fig 3 is the Morel MDT29 with a High Pass Network. The two drivers are crossing at 2.2kHz.

Fig 4 – Starling-MDT Passband

The Black plot in Fig 4 is the summation of the crossover. There are no cancellations in the passband, indicating the two drivers are crossing properly.

Fig 5 – Starling-MDT Frequency Response

The final frequency response of the Starling-MDT is in Fig 5. It is flat within +/- 2.5dB from 500Hz upwards. There are two dips, one at 5.5kHz and the other at 7.5kHz. I can’t do much about them because they are caused by the cone breakup on the ER18. Anyway, it is quite impossible to pick them up when you listen to music. What is important is the midrange is now only about 2dB to 3dB above the bass.

Fig 6 – Starling-MDT Null

When I flipped the tweeter wires around, it resulted in a deep null centered at 2.2kHz (Violet plot in Fig 6). I got this by chance. I was more focused on the overall tonality. 

Fig 7 – Starling-MDT Step Response

The Step response shows the ER18 is actually very fast. At the 90% mark, it’s at 200 micrsec. In the last 10%. it slowed down, hitting the apex at 250 micrsec. 

Fig 8 – Starling-MDT Waterfall

The Starling-MDT is remarkably “clean”. There are very few artifacts from 1kHz upwards. I don’t think I’ve ever come across something like this before.

Fig 9 – Starling-MDT Spectrogram

The Spectrogram confirms the waterfall plot. There’s scarcely any excess energy. This is quite an amazing plot. Normally, you’ll see some hot spots between 1kHz to 2kHz. 

Sound of Starling-MDT

If you plan on building the Starling, this is the one. Forget about all the previous versions that I did. It’s that good.

What sets this Starling-MDT apart is the bass. It’s not just any ordinary bass. The Starling-MDT bass is tight, dynamic, fast and believe it or not, it punches. I’ve never been able to get this before with the ER18 but I succeeded in the Starling-MDT.

If you want to know how the Starling-MDT punches, listen to Billy Ocean’s “Get Outta of My Dreams, Get into my Car”. The bass PUMPS. No sloppy bass here. It’ll make you want to dance.

Another track is Pointer Sisters “Slow Hand”. In the opening, you can clearly make out the bass note. This is immediately followed by a bass extension before dying off instantly. It is this bass rhythm that ties the song throughout.

This concludes my adventure with the Seas ER18RNX. It was a tough ride, 10 years to be precise, but it was worth the efforts. in the end, I managed to tame her and let her flourish.

Crossover is available on request. Free for DIY. Not for Commercial use.
Unless otherwise stated, all measurements were made with the mic at 36 ins, tweeter axis. Impulse Window=5ms. No smoothing applied.