My first OB loudspeaker, seeking advice.

[Kim Schultz]

I have decided to make an OB loudspeaker after having listened to this experimental setup.

An externally hosted image should be here but it was not working when we last tested it.

This is inspired by Simon's OB with Eminence and Skytronic drivers.

My project will use an Eminence PSD2002 driver with a 90 degrees OS waveguide as tweeter. I had the driver already, and it was easy to remove the backchamber to make it dipole.

The mid will be an Eminence Beta 8a and the woofer a Fostex FW405.

The Beta 8a will cover as much as possible and hopefully run wide open without any crossover.

The tweeter will use a single cap and a lpad to bring the level down, to match the mid.

The Fostex will be in a H-baffle with a separate woofer amp and electronic crossover.

I have played around today and can see I need to have the front of the Beta 8a, about 7cm back compared to the mouth of the waveguide. This is need to time align the drivers.

An externally hosted image should be here but it was not working when we last tested it.

An externally hosted image should be here but it was not working when we last tested it.

I can do this by mounting the horn flush with the baffle, and then mount the mid on an extra spacer behind the baffle, with a 90 degrees flare out to minimise reflections.

After reading some info about OB's I don't know how wide I should make the baffle.

The H-baffle will be about 45cm wide, and I thought about a 40cm wide baffle for the mid to help it go down lower, but I don't know if it is to wide.

I've also considered to make the baffle ready for a second mid like the Bastanis loudspeaker.

This would mean the tweeter will be lower than the mid, is this ok?

Should the drivers be offset on the baffle, or placed in line ?

Thx in advance, all help appreciated.

 

[tinitus]

and the woofer a Fostex FW405.

thats a very expencive woofer

any special reason to use Eminence PSD2002 driver ?

 

[Kim Schultz]

I have it from an old 3way studio type loudspeaker, and the previous owner had it refoamed with a smaller roll surround.
This has caused the fs and qts to go up a bit and actually making it better for OB, it sounds quite good in the simple baffle.

If the mid and tweeter section turns out good, I might get a more suitable woofer later.

Only reason for using PSD 2002 is that I had it already and it was easy to remove the backchamber.

 

[tinitus]

I need to have the front of the Beta 8a, about 7cm back compared to the mouth of the waveguide. This is need to time align the drivers.

why ?

btw, Im still waiting for someone to try diaphragm coating on Beta-8

 

[Kim Schultz]

I needed to move the the tweeter forward to have the impulse measurement align.

Simon did ENabl his Beta8 and some coating too I think.

 

[tinitus]

I needed to move the the tweeter forward to have the impulse measurement align.

caused by crossover ?

 

[tinitus]

I've also considered to make the baffle ready for a second mid like the Bastanis loudspeaker.

This would mean the tweeter will be lower than the mid, is this ok?

Should the drivers be offset on the baffle, or placed in line ?

with double Beta-8 I would suggest to use them as 2.5way (MMT)
and upper 8" would be your listening height

With a compression driver especially Im not sure its wise with just a small series cap
and you may best try to match its rolloff with that of the midrange
not too difficult, but still a bit more tricky than just a small series cap

btw, your impulse measurement will change with a crossover

 

[Kim Schultz]

I'll experiment with a test baffle and some different crossovers tomorrow, to see how it affects time alignment.

 

[Godzilla]

Good luck with your project. I see you have some neat waveguides turned over there. I am curious how the Beta 8 sounds to you without a circuit and on open baffle.

Zilla

 

[Kim Schultz]

I had a board 38cm x 68cm already cut that fit nearly perfect, so I used it to make a test baffle.

The mid driver is mounted behind the baffle with the edge rounded over, the waveguide is mounted flush and offset to the left, in case I want to use another mid below the other.

The measurements are taken 1m out and 30cm to the right.

The mid is run wide open and the tweeter with a 1uF series cap.

I think it looks very promising.

Sorry for the crap photos, they were taken in poor light with my phone.

 

[Kim Schultz]

here is a measurement from 30cm to the left.

Would It be better to place the drivers in a vertical line ?

 

[Kim Schultz]

Here is another test with a 2nd order (3,3uF + 0,28mH) + lpad on the tweeter.

I alos moved the baffle forward to clear the other test baffle, and to make it more seamless to the wooferbaffle.

I think it looks and sound better.

 

[Kim Schultz]

I did some more experimenting today.

First attempt:
2nd order HP on tweeter, 4,2uF + 0,24mh and 21ohm + 3ohm lpad.
Mid run fullrange.
Measurement done at 1m/15° to the left (near tweeter) 1/3octave smoothing.

Second attempt:
3rd order, 2,4uF + 0.24mH + 2,8uF and 17,5ohm + 4ohm lpad because I thought there was a bit to much overlap in the 2nd order.
It also has a gradual rolloff which means it sounds great even when played very loud.

 

[tinitus]

....... because I thought there was a bit to much overlap in the 2nd order.

looks great

what you heard could be xo anomali caused by drivers impedance ressonances below 2khz

 

[Radugazon]

Exciting project ! It's nice to see the evolution of your filter curves.

As using open back drivers too (see the avatar) with kind of waveguides (ugly in comparison ) I can tell you:

*** Take care of the deadly love of the magnet for any iron part ! Dead dome issue.

*** With titanium it's ordinary to found little peaks in the curves, they will probably sound irritative, especially that the treble must have OTOH great qualities.

*** For me, damping of the surround resolved a big part of the problem. For the sake of easy removal I used high temperature grease, dirty but efficient. I've seen that too much kills the high, that's why it's better to go with this product.

*** Le Cléac'h demonstrated something else about drivers : when using a serie R (kind of 20 ohms) instead of a L pad, then the driver's curve follows more it's imp curve and in some cases gets more linear. Impulse is not affected. Can do a test.


As I see you are not lazy for posting measurements, can you show us a 0° front, vs 180° back vs 90° side ???...I never had a dipole fig of 8 pattern but almost omni

BTW, good luck

 

[Kim Schultz]

I did some more tweaking today and realised I had too much lower midrange, so I put a 10uF cap with a 6,5ohm bypass resistor i parallel with the mid driver.

I also decided that using the subwoofer amp with the Fostex for OB bass, will not work.
The crossover can only be set as high as 150hz, then it sounds boomy and there is still a hole between the woofer and the mid.

I´ll try to use a BR with the fostex until I get a better OB woofer.

Radugazon, here are the measurement you requested, but they are taken from 15°, 90° and 165° as the baffle is designed to work best at 15°.

 

[Radugazon]

Thx Kim,

I was asking front/side/back curves, to learn more about the dipole behavior of opened 1" drivers. Mines have a very fat fig of 8 pattern, but, if I dn't miss something, where is your back radiation? Closed again, or with an other tweeter ?

OTOH, the suck out you have at 0° (and disappearing so fast at 15°) could come from offsetting too much from the center line. Maybe, it's better to ge back to the safe way: all in the middle, at least for starting.

 

[Kim Schultz]

I made a mistake yesterday and measured the unmodified tweeter, that´s why no backway could be seen.

Today I made some new waveguides and centered them under the mid driver.
I still have an issue with time alignment, even though the WG mouth has been moved forward, and I also need to adjust the crossover again.

Here is the 0-90-180° measurement.