2-way esl

[chinsettawong]

Originally Posted by vmk

Hey, that sounds more like it. I need to redo the leaking stator anyway and now that we have bigger living room than what we had when I started the project I might as well go little bit bigger .

I made dipole bass units once for my little Magneplanar's and they were nice match for them. Have to start to selling the idea for my wife

cheers,
Vesku

Hi Vesku,

I meant to say that my stator size was 300x1200.

I put three little spacers (25mmx25mm) in the middle of the stator instead of a long one like yours. I think that it helps the diaphragm to vibrate easier.

Wachara C.

[vmk]

OK, now they are redone. 1-way and 2mm air cap. New stator dimension is 1250*305. The Mylar tension is lower, about 800g with ER-tension gauge.



The bigger size of the stator seems to be too much for my toroidal transformers, highs go down before 20kHz, but as I'm a half deaf it's not a big problem.

Had a little testing session with the transformers, there's a very big differences in the performance above ~12 kHz even if you just alter the primary winding by one round. Is it because the resonance freq of the trannys is below 20kHz? With a higher transfer ratio I get highs cut off earlier. With a low ratio the high freq resonance gets really big. The freq plot attached below is with the "right ratio" and 1 Ohm series resistor. Ratio is some where around 80, 5.5V to 230V... Some of the highs may be off because my extra long RCA-cables and crappy sound card...

I'm using Winamp and EQ-software . It's a pretty nice and easy way to find the sound you are after. Although all of the High End purists here probably don't agree

Obviously I want to get rid of the bass resonance, now I'm using about 60uF series capasitor to cut of lows (from about 100Hz). What is the prober way to take care of the bass resonance?

With the EQ-software I have tweaked +6dB's at 80Hz and sloping down to 450Hz (0dB) and speakers sound quite nice now (I have to do more listening before making my final call, I'm chancing the setting all the time...). What's the best way to do a similar equalising with passive components? I have done a few other ups and downs with the software, but I guess there's no easy way to do them with passive stuff. I wish I had mine Behringer parametric equaliser (feedback destroyer) with me here in NZ...

Now I need a beefier amp as well...

[SM7UYJ]

Have you tried to measure without the series resistor?
If I put a series resistor in my speakers, the highs totaly go off!
(also running 5.5V/220V trannys - self wound primarys)

/R

[vmk]

Yep, tried with out series resistor as well, but I get quite nasty resonance peak if there's no series resistor. Now I have 0,5 Ohm and that seem to be the best value in my system.

Only difference in pictures below was that I did put one more round to the primary. It seems that in my set up the step up ratio makes the biggest difference how the highs behave.

But the speakers sound really nice, that's the main thing, I think. Need to start a woofer project next, dipoles... Well, I will first wait for couple of weeks and see are these ones still alive...

[SM7UYJ]

Is your resonance peak at abt. 2k?
If so - it's spooky - it's exactly how my system behaved!
I solved this by putting a 9 Ohm resistor parallell with the primary winding of my transformers.

/R

[vmk]

We'll see tomorrow, I'll try to kill the resonance with a same value resistor

The stators are about 1.3nF, measured with a cheap DMM.

I used these transformers: 9 + 9 50VA Toroidal Transformer : Discount Electronics and unwinded the 9V side.

One of them shorted, I don't know what exactly happened. It was working fine with the speaker, but when I hooked it up to 230AC to check what the transfer ratio was after unwinding the secondary, it shorted the "230AC side". Rest of the three worked fine with 230 AC test. I might have damaged the winding some time through the process.

[Calvin]

Hi,

as CR You could start with a peaking Highpass of 2nd order. Fpeak ~80Hz. this does basically the same as Your EQ does now. It ´pulls down´ the high peak at Fs and fills the dip above the peak.
To get values measure the impedance of the system and the amplitude repsonse and simulate for example with LSP-Cad or similar.
Read thread Ping Calvin Multidoc4 describes exactly what You could try.

jauu
Calvin

[Capaciti]

Hi Vmk,

looks nice !

At which distance you measured your system ?

If you measure at about 40" distande frequency response should show a linear and constant loss of sound pressure from low to high frequency. The difference should be about 10 dB since your room looks quite undamped.

The current plot indicates an anemic lower midrange and maybe Crying sound of voices.

Try a 220µF capacitor in serie to the stepup and bridge it with a resistor. Variation of the resitor value will enable constant lowering of pressure towards the high frequency. The value is difficult to predict, maybe start with 4,7 ohms.

Capaciti

[vmk]

Originally Posted by Calvin

Hi,

Read thread Ping Calvin Multidoc4 describes exactly what You could try.

jauu
Calvin

Thanks a lot for pointing out this good thread!

After installing the impedance correction resistor (8.2 Ohm) no change to output, but huge difference how the HP-capacitor acts. Now I need to use bigger series capacitor (200uF). It don't cut off the resonance anymore, but it acts like a prober HP-filter now.

Next I need to do the notch and the RL. That's going to be a bigger job, with the impedance measurements and learning how to use a new software. Hmm, as I'm lazy, is there any way to get away with a trial and error method with the notch and RL?

The 2kHz peak vanished when I moved the measurement point a bit higher... Not a spooky thing after all My measurements are from about 10cm away from the speaker.

Just had a quick look at the impedance measurement with ARTA and seems to be pretty straight forward, I'll try to get that done next. Just need to do the "voltage probe"...

Below a plot showing where I'm now with this never ending tweaking process. Will try with a bit larger capacitor later, but now I will listen some music

[JonasKarud]

Quad solved this problem with the bass peak using a special very fine mesh, placed on the inside of one stator, to create a acoustical resistance that dampened the peak.