Koapong monitor

The Dynaudio 15W75-4 ohm (=MW150) midbass married to a Vifa DX25TG tweeter makes a great mini-monitor. I have had this speaker playing in my workshop now for some months, and the more I listen, the less I feel inclined to make any changes. This is probably as good as it gets from these two drivers and the series filter.

During the summertime when the wind and noise is very low, it's possible to do out-door measurements. Due to the (almost) absence of reflecting surfaces it is possible to measure really low frequences. In-door we usually have reflecting surfaces within 1-1.5 metres not allowing frequency response measurements below 300 Hz. By adding even more absorbents to the lawn between the speaker and the microphone than seen here, it is sometimes possible to measure below 100 Hz. Getting the birds to be quiet is sometimes the biggest problem!

Well, back to the reason behind the -
Koapong

Dynaudio 15W75-04 (alias MW150), some 3" voice coil!

Vifa DX25TG.

The reason for this construction came from Koapong, residing in Thailand. Koapung told me he had bought a pair of Dynaudio MW150 car speakers and the Vifa DX25TG tweeters and now he was short of a crossover design.
Well, this is obviously a thing you should never do and under normal circumstances the purchase of these drivers would have been a waste of money, as you cannot put together a crossover by theory that will have any chance of revealing the qualities of these excellent drivers.

Under normal circumstances my response would have been “best of luck” and that would have been it. However, I have a pair of Dynaudio 15W75 drivers – in 4 ohms version – that are similar to the MW150. I bought these never-used drivers second-hand for a very reasonable price and have had plans to mate them with the 21W54 bass drivers and the Esotar tweeters, fig. 5.
The 15W75/MW150 is a particular “troublesome” driver as it has two intrinsic resonances that have to be taken care of in the crossover design.

I have been testing the drivers in a two-way design with the Esotar tweeter and got a very nice response profile – and a quite complex parallel crossover. Check this one out at: http://home1.stofanet.dk/troels.gravesen/index_b/15W_ESO.htm

Based on the recent W11 and SP95 construction I couldn’t help trying something completely different from the Esotar set-up and a couple of series crossovers later I ended up with this:

Koapong series crossover, version 3.

A test cabinet was available and it has the following dimensions:

Koapong cabinet dimensions.

Test cabinet and preliminary crossover.

My test cabinets are made from 20 mm Baltic birch as can be seen from the pictures. These test cabs are somewhat deeper as they are meant for housing a wider range of drivers giving 6-10 litre volume, here 7.6 litres.

Final Koapong cabs under construction.

Front panels routed and ready for assembly.

Left: Impedance plot of Koapong speaker. Impedance stays above 4 ohm and phase shift below 100 Hz is moderate making this an easy speaker to drive. Actually my Copland CTA505 valve power amp loves this speaker – best from the 8 ohms tabs. As can be seen the vent tuning is around 60 Hz.

Right: Response of individual drivers (purple and green) and response with inverted tweeter (red). Yellow = response of 15W75 without 10-12 kHz notch-filter. Further refining of this crossover may include getting rid of the 3.5 kHz bump/dip at 2 kHz by trying an 18 dB/octave approach. This will probably eliminate the need for a 10 kHz notch filter. However, there does not appear to be much interference from the 15W75 in the 3-8 kHz region on tweeter performance.
I cannot for sure say I can hear the impact of the 10 kHz notch filter. Tr

Left: Frequency response from left and right speaker. The difference between speakers is less than 1 dB.

Right: Step response. Note smooth decay from 15W75 driver and positive polarity of both drivers. The first very sharp peak is the tweeter response. Then comes the midbass. As can be seen the tweeter peak is very short and sharp, a fraction of a millisecond, where the midbass will last for more than 2 ms before it dies.
What is important here is that the decay curve of the midbass doesn’t display dips or bumps as a result of stored energy. Controlled break-up that is. One of the good things about polypropylene.

Step response explanation: When the impulse peak is up the driver has positive polarity. The black part of the graph is the reflection free time frame chosen for the SPL response graph.