Development of open baffle 3-way plus subs

[Lewinskih01]

I decided to try an open baffle and built 1 unit: twin 10" midbasses on a 45cm wide panel, an 8" midrange bafleless hanging on a swing, and a waveguided AMT for the time being. Supported by a coupe of sealed subs. I built just one because I used parts I had available, especially the 10" woofers and amps (would be short 1 stereo amp for a full 4-way stereo system).

Of course OB requires EQ because of the raising 6 dB/octave up until the dipole peak. I'm running the twin 10" from 70 to 500Hz which is still below the dipole peak, and the 8" from 500 to 1500, also below dipole peak, so to insure constant directivity behavior. At 1500Hz the waveguided AMT takes over.

20200920_100000.jpg (1.12 MiB) Viewed 4064 times

20200913_131647.jpg (973.05 KiB) Viewed 4064 times

I used Linkwitz 2nd order xo just because I didn't know what tends to be better for OB. Anything better I should use?

Then onto linearizing the drivers...a challenge in and of itself! Linearizing the tweeter was as usual, done nearfield. But the midrange and midbasses didn't seem a good idea to measure nearfield because I would be missing the bounce from the back wall. So I measured at the listening position. For this 1-speaker setup the listening position 1.5 meters away from the speaker, on axis horizontally. I took a Logsweep with the linearized xo, applied Room Macro 1, and the result is displayed on the red curve.

OB response linearized xo Macro 1.png (91.79 KiB) Viewed 4067 times

You can see the large drop around 100Hz. Subs are xo at 70Hz, which is the last peak you see before the drop. When establishing the target curve to avoid the suckout at 100Hz, I ended with about 20dB attenuation. The amp driving the twin 10" has plenty of power (Hypex UcD400) to drive the two 10", and it looks to me like I could take advantage of the linearization process and apply a normalization to a play at a higher level, so the response would be significantly higher at 100Hz, which in turn would push up the 200-300Hz region too.

I wasn't able to to extended listening yet. Last night I played a couple songs and was a bit underwhelmed.

Some questions:

• What are good crossover types to use in OB?

• How should I raise the 70-500Hz midbass region to avoid the huge attenuation at inversion? I'm thinking like +10dB. Or al alternative approach?

• Where should I measure to linearize OB drivers?

• Any additional input from very welcomed!

Thank you!

[UliBru]

Horacio,

the big dip at 100 Hz may be caused by an Allison effect. The speaker is placed on a hard floor. What happens if you put the speaker on e.g. a table?

For OB my experience is that it does not make sense to measure close-up. So I always do this at the listening position, even for a possible driver linearization. Maybe it would make sense to measure OB outside in free field without influence of the back wave.

Crossover: you can try all XOs offered by Acourate. UB/NT XOs are very steep. LBessel or RBessel have the least ringing. Buttterworth/Linkwitz are classical XOs. Usually I try to measure the harmonic distortions and then I decide about XO frequencies and steepness.

[Lewinskih01]

Horacio,

the big dip at 100 Hz may be caused by an Allison effect. The speaker is placed on a hard floor. What happens if you put the speaker on e.g. a table?

For OB my experience is that it does not make sense to measure close-up. So I always do this at the listening position, even for a possible driver linearization. Maybe it would make sense to measure OB outside in free field without influence of the back wave.

Crossover: you can try all XOs offered by Acourate. UB/NT XOs are very steep. LBessel or RBessel have the least ringing. Buttterworth/Linkwitz are classical XOs. Usually I try to measure the harmonic distortions and then I decide about XO frequencies and steepness.

Thank you Uli.

Will try with different heights.
Good to know you also measure at listening position. Measuring outside is pretty difficult for me. Maybe can set the OB in the middle of a 10mx5m room and measure 1.5m away on axis.

How do you measure harmonic distortion? I have to admit the speaker didn't sound good. Maybe this is a good place to start.

[UliBru]

Horacio,

for some reasons the English version has disappeared.
But please try https://www.audiovero.de/acourate-wiki/ ... tsprechern in combination with Google translator.

I need to rebuild the wiki page.

Usually people do not listen at 1.5 m distance. So I would try to measure at > 2m

[Lewinskih01]

Horacio,

for some reasons the English version has disappeared.
But please try https://www.audiovero.de/acourate-wiki/ ... tsprechern in combination with Google translator.

I need to rebuild the wiki page.

Usually people do not listen at 1.5 m distance. So I would try to measure at > 2m

Hello Uli.

Trying ta measure distortion per Wiki. On step 2 it refers to an Inverse48.dbl. How do I create that file?

Regards,
Horacio

[UliBru]

The Inverse48.dbl is located in \Documents\Acourate\LogSweep

[Lewinskih01]

The Inverse48.dbl is located in \Documents\Acourate\LogSweep

I measure the 2x10 inch section only, still on the floor, but 4m away from front wall, mic 2m away from speaker and 4m awy from back wall. Looks pretty ugly!!

2x10G40 distortion.png (82.86 KiB) Viewed 4018 times

I wonder if I'm doing something wrong. Will measure the mids to understand better.

Interestingly, when I measured yesterday thee drivers Acourate noted positive polarity. I have not changed the setup, and now it's telling me negative polarity.

Ideas?

[Lewinskih01]

Here's the midrange (B&C 8PE21), more like I was expecting. About 45dB between k1 and k2 or k3.

8PE21 ditortion.jpg (102.3 KiB) Viewed 4017 times

The good news> apparently I was doing it right. The not so godd> I seem to have a problem with the 10 inch woofers. I'm thinking the measurement level was low with the woofers and maybe the noise floor is afecting the measurement. The low gain in the woofer's amp is actually an issue I'm also solving for.

[Lewinskih01]

I re-ran the 10" woofers, now at higher SPL, and looks more like what I was expecting

10G40 distortion 2.jpg (189.22 KiB) Viewed 4015 times

The Hilo output I was using didn't have enough gain for the amp driving the woofers. Now I changed that, which turns routing into a mess, but oh well.

Uli, now that I was able to run the distortion measurements, what are your thoughts about xo types and slopes? My starting points for xo points are 70, 500, 1500Hz. A couple 12 inch sealed subs take care below the first xo, and a TPL-150H takes over at 1500Hz.

[UliBru]

Horacio,

the 10G40 seems to be quite inefficient below 200 Hz. Can the 12-inch subs run up to 200 Hz?