"Mechano", a small 2-way design with frequency power response target (with Satori and Seas drivers)

A humble 2way system designed with help of  XMachina. 

It's a small bass reflex design. The system was aiming to reach f6 around 40Hz with  8..10 liters BR enclosure. Other goals: generally balanced on-axis response and close to linear frequency power response.

General view


The drivers: Satori MW13P-4 [link] and Seas H1396-4 [link].

The enclosure is straight, no roundovers, chamfers or waveguides. External dimensions: 16x29x29cm (8l tuned to c.a.45Hz) . It's made of 18mm birch plywood. 

Some reinforcements are applied inside. Crossbar connecting side walls and  stripes of 12mm plywood (5cm wide) are glued along both sides and the upper wall. They are intended to strengthen the "sandwich" structure of the enclosure.

The drivers are placed in the middle, with centers at 4cm and 15.5cm from the top edge.

The enclosure is filled with polyester fiber.

Off-axis measurements were made in 10-degree steps to determine the power resp. A simple rotary table was used for this [link].

Initially I looked for crossover frequency between 2.5-3kHz, but matching these two drivers characteristics in terms of the assumed system power response was  difficult. I relaxed the criteria and then XMachina provided the following solution (system on-axis: cyan,  1.6dB/oct: yellow, power response: red):


But it turned out that crossover frequency is around 5kHz. The distance between breakup (8kHz) and edge of mid-woofer bandwidth (5kHz) requires sharp filtering (system on-axis: cyan, mid-woofer frd: gray, filter resp.70dB shifted : magenta):

Driver Freq Resp (gray), shifterd filter response (pink), system response (cyan)

A side effect of high crossover frequency are significant dips in vertical patterns which appear as early as 20 degrees (but it's all taken into account in the power response calculation).

Crossover is non-minimalistic, but also not very complicated.The resistors in series with the coils L1 and L2 were eliminated by using coils wound up with thinner wire. 

The resistor marked in green is used for correction. I made a mistake exporting woofer impedance characteristic (zma) and, as a result, the impedance module was much lower than the real one. This erroneous characteristic was used in the design. Unfortunately, I realized that when the filters were already assembled. 

First, I checked what would change on the characteristics if I load the correct impedance characteristics, hoping that the changes would be negligible. However, this did not happen, the power response characteristic was especially damaged, as it clearly bent above the assumed target, i.e. a linear decrease of about 1.5dB per octave.

Then I checked if it would be possible to do something to make the load impedance seen from the filter output resemble the impedance wrongly used in the design.

The solution turned out to be quite simple, it was enough to shunt the woofer with a resistor (12R, 20W). The power response characteristics returned to the target linear shape (at the cost of a small decrease in efficiency)

VituixCad 6pack before

and after the corrections:


Assembled filters (the resistors used for correction (12R, 20W) are not here, they are soldered to the mid-woofers connectors):


The crossover is attached to the lower part of the enclosure.


Impedance module (measurement).


Measured on-axis system response after correction (both pieces):


vs. LF box simulation:

Measured driver parameters differs from the values taken from manufacturer datasheet (e.g. fs = 58Hz instead of 41Hz), so measured characteristic differs from the simulation in the LF range. 

According to the measurements, the "f6 around 40Hz" target was not achieved, but that does not mean that bass is missing. Amazingly, there is even a very low bass (from single 13cm driver!), although of course not in large amounts.

List of components (for one unit):

  • capacitor 2,70 µF 5% e.g. Audyn Cap MKT 
  • capacitor 4,70 µF 5% e.g. Audyn Cap MKT 
  • capacitor  0,68 µF 5% e.g. Jantzen Audio Cross-Cap
  • bipolar capacitor  100µF 5% e.g Jantzen EleCap 
  • bipolar capacitor  18µF 5% e.g Jantzen EleCap
  • air core inductor,  0.18 mH 0.550 Ohm, wire 0.50 mm, e.g Jantzen Audio
  • air core inductor,  0.43 mH 0.545 Ohm, wire 0.70 mm, e.g Jantzen Audio
  • ferrite core inductor 0.560 mH 0.144 Ohm, wire 1.00 mm, e.g Jantzen Audio
  • air core inductor,   0.20 mH 0.200 Ohm wire 1.00 mm, e.g Jantzen Audio
  • resistor 2,00 Ohm 5 Watt 
  • resistor 4,70 Ohm 5 Watt 
  • resistor 5,60 Ohm 5 Watt 
  • resistor 12 Ohm 20 Watt 
  • resistor  18.00 Ohm  5Watt
  • bass reflex port for tuning 45-49Hz@8l, e.g. Monacor MBR-35


Komentarze

Popularne posty z tego bloga

3-way monitor with AMT (Mechano323)

Mechano23: small 2-way speakers with linear on-axis and power response characteristics with Scan Speak H2606/920000 and SB Acoustics SB13PFCR25-4,

"Mechano22" - new version of a small 2-way speakers, with linear on-axis and power response characteristics.