These speaker will replace KS-223 as TV speakers. Speakers will be located close to flat TV without stands. Seas W15LY001 woofers are 'stolen' from KS-223 which will be demolished. Another option was Peerless 5.25" Exclusive (831882) but that's not so good looking. Black is black. Many years ago Seas 27TBFC/G tweeters were installed in KS-585 with customized Monacor WG-300 wave guide. This time I decided to try local MCM 54-580 clones designed for Seas 27T tweeters. This wave guide should work okay also with 5.25" woofer though it's probably designed for 6.5"...8" woofers. DI spectrum looks suitable for LR4 @1400 Hz where DI has dropped 2 dB. I don't have any special sound, quality or SPL requirements. I'm happy if they sound tolerable to me.

Kontiainen is Finnish name for Talpa europaea, European mole, a small, black and annoying animal which turns lawn upside down.

Technical specifications

General 2-way, vented
Dimensions 180 x 330 x 280 mm (W x H x D)
Volume 10 liters
Tuning 49 Hz, vent Ø47 x 171 mm curved
Crossover 1800 Hz, acoustically 3rd order Butterworth with inverted tweeter
Sensitivity 84 dB/2.83V/1m
Drivers Seas W15LY001
Seas 27TBFC/G, Wave guide with MCM 54-580 (6.5") profile
Impedance Nominal 8 ohm, min. 5.5 ohm @ 300 Hz


Drawing Kontiainen.pdf

Enclosure material is 19 mm MDF. Crossover boards are hot glued on the side wall.

50mm E25 plastic foam for damping is located on the back wall and top. Curved vent is rotated ~25 degrees clockwise to increase air gap between plastic foam and inner end.

Vent is Uponor 50 length 70mm through the back wall extended with Uponor 50x88.5 rounded angle inside the box. Locking seal is removed and groove cut off from inner end.

Important! Straight vent does not work well in this location (in the middle of rear wall). Amplitude of lowest pipe resonance (~800 Hz) would be stronger than Helmholtz resonance. Curved vent towards top damping decreases amplitude of pipe resonances about 20 dB, down to 15 dB below Helmholtz resonance at the exit. Flared outer end would help but aiming of inner end away from woofer's cone and towards plastic foam on the top is radically more effective solution.


Filter schematic:

Coils by Alphaudio have air core. 1.36 mH and 0.27 mH with 1.4 mm wire, both 2.7 mH with 0.9 mm wire.

Caps are SCR MKP 250 VDC.
11 ohm resistor is two 22 ohm 5W in parallel. Other resistors 5W.

Board dimensions are 120 x 200 mm

It could be simpler for sure but I can afford few tweaking coils and caps.
LR4 crossover would probably give equal result. See LR4 simulation.




Source 2.83 V, 0.0 ohm


Woofer's near field and far field measurements are combined with Merger tool. Splicing frequency is 400 Hz with blending of two octaves. Near field sum is corrected with cabinet impact response by LspCAD 5.25 Pro.
Far field measurements at 631 mm, time window 6 ms. 0-180 deg in horizontal plane with 5 deg steps, totally 37 measurements.


Far field measurements of tweeter at 631 mm, time window 4 ms. 0-180 deg in horizontal plane with 5 deg steps, totally 37 measurements.
Reference distance is 2500 mm. Origin of drivers & microphone is center of woofer. Reference angle (design axis) is horizontally 10 degrees off-axis.
Both horizontal and vertical planes are included in Directivity Index and Power response calculation. Intensity for power response is calculated on spherical surface.

Project data in VituixCAD 0.1-1.0 format: Kontiainen_demo.ZIP


Frequency response

Axial response
- measuring distance 100 cm, on woofer's axis
- time window 3.5 ms, freqLO=285 Hz

Room response
- measuring distance 250 cm, on-axis
- smoothing 1/3 oct.

Near field responses
red=woofer cone, black=port

Step response

- measuring distance 100 cm, on woofer's axis

Harmonic distortion

- measuring distance 30 cm, on-axis
- SPL 88 dB/1m

Excess group delay

- measuring distance 100 cm, on woofer's axis
- time window 15 ms


- room measurement at 250 cm, on-axis

Acoustical parameters during previous STI measurement
red=T30, green=T20, blue=EDT




Full baffle step compensation and 1-2 dB steeper tilt of power response compared to typical enables listening also in large rooms. No problem to get emphasized mid..upper bass by positioning speakers and listener. In our small TV room at least vent of corner located speaker should be blocked with foam or sock while listening music close to rear wall. Lower midrange is slightly thick but not very booming. Probably near field measurement of small exponential cone was not very accurate and far field measurement had slight floor dip causing too narrow baffle diffraction equalization in crossover. Excessive thickness can be handled by positioning or turning speakers upside down.

Sound is slightly muffled off-axis especially in damped acoustics. This is quite common with 2-way if power response has shallow dip on lower treble. This feature may turn upside down to slightly hard/metallic when listening distance is decreased close to two meters. Good balance should be found by careful positioning.

Very small amount of sibilance and rasp exists but mainly off-axis. This is quite commonly associated with 2-ways with wave guide - at least in Finland where this concept has been almost de facto standard for two decades. I think it's tolerable this time.

Attack and strength of piano is good thanks to timing and full bodied lower midrange. Much better than just thin tinkle.

SPL capacity and compression spectrum are poor. Short displacement and small cone of woofer with large vent limit sound pressure at low frequencies very early. Voice coil seems to heat up equalizing tonal balance after few minutes listening with high volume. Simple unplugged and elevator jazz work okay quite loud but Megadeath etc. kill the speakers in a seconds with careless listener.

Sound stage is logical, accurate and also quite spacious (3D) if listening triangle is small enough. Stage stays logical to different listening distances without need to rotate/aim the speakers. High resolution much further than 250 cm requires very short early decay time. Our living room with RT60 <400 ms and EDT <200 ms enables good resolution up to 250 cm with this kind of design.

Page created 29.9.2015
Updated -