Hi,
ok so far ;-) Is the measurement taken with a series R in place?
As You can see from the impedance plot the Panel´s capacitance and the Trannie´s leakage inductance form a resonce at ~14kHz with a peak in the amplitude response. This peak is dampened by the series resistor. Above this peak amplitude response drops with -12dB/oct (2nd order lowpass; LC-filter).
Now You can start simming with LSP-CAD. But expect some rather unusual values for the CR-parts unless You linearize the impedance with a big parallel resistor or a RLC-impedance matching network.
jauu
Calvin
ok so far ;-) Is the measurement taken with a series R in place?
As You can see from the impedance plot the Panel´s capacitance and the Trannie´s leakage inductance form a resonce at ~14kHz with a peak in the amplitude response. This peak is dampened by the series resistor. Above this peak amplitude response drops with -12dB/oct (2nd order lowpass; LC-filter).
Now You can start simming with LSP-CAD. But expect some rather unusual values for the CR-parts unless You linearize the impedance with a big parallel resistor or a RLC-impedance matching network.
jauu
Calvin
It was the plain panel with out any resistors.
Yep, but it seems to work now. The CAD is plotting about same sort of impedance curves that I'm able to measure. OR IS IT working? What makes the difference in the frequency where it drops 400Hz(meas.) vs. 200Hz(calc.)?
The measurement and the simulation below were made with an 8.2 Ohm parallel and a 3.3 Ohm series resistor + a 25uF series capacitor. On left is the measurement and on right the cad simulation.
If I would like to use a bigger series capacitor to make the notch doable would that be a right guess?
Yep, but it seems to work now. The CAD is plotting about same sort of impedance curves that I'm able to measure. OR IS IT working? What makes the difference in the frequency where it drops 400Hz(meas.) vs. 200Hz(calc.)?
The measurement and the simulation below were made with an 8.2 Ohm parallel and a 3.3 Ohm series resistor + a 25uF series capacitor. On left is the measurement and on right the cad simulation.
If I would like to use a bigger series capacitor to make the notch doable would that be a right guess?
Attachments
those look fantasic! making me a little more eager to have my own. great work, nice catch. I will toast with a beer with you soon.
Here's the situation after a couple of days of the tweaking. I got the highs working OK when I lowered transformer ratio and used a bigger series resistor. I don't need to use any equalising on the PC for the highs any more and the over bright sound is gone. Attached a picture of the Xover at it's current stage, it's not ideal, made from the bits and pieces I found lying around.
I still need to use Winamps EQ between 1kHz to 125Hz. Got that sorted too with the xover, but only when I used higher xover freqs.
Unfortunately I have only a single subwoofer at the moment and I'm forced to use low xover frequency (with high xover >60Hz I can't listen the speakers at all
). Now I have a weak bass between ~150-70, but I can live with that (for a moment or a couple of years).
"ping Calvin"-thread has been a great help, much appreciated...
2 things to do now:
1st REBUILD, 2-sided tape that I used gave up in one of the horizontal supports (we had a warm day...and the stator has some internal tension thanks to guy who cut it and then dropped it). I will but more tension on the Mylar when I'm on it, aiming to a ~100Hz resonance.
2nd come up with a bass unit to cover area approx. 50Hz-200Hz. Any ideas?
I still need to use Winamps EQ between 1kHz to 125Hz. Got that sorted too with the xover, but only when I used higher xover freqs.
Unfortunately I have only a single subwoofer at the moment and I'm forced to use low xover frequency (with high xover >60Hz I can't listen the speakers at all
). Now I have a weak bass between ~150-70, but I can live with that (for a moment or a couple of years)."ping Calvin"-thread has been a great help, much appreciated...
2 things to do now:
1st REBUILD, 2-sided tape that I used gave up in one of the horizontal supports (we had a warm day...and the stator has some internal tension thanks to guy who cut it and then dropped it). I will but more tension on the Mylar when I'm on it, aiming to a ~100Hz resonance.
2nd come up with a bass unit to cover area approx. 50Hz-200Hz. Any ideas?
Attachments
Hahaa, now I'm listening with out an external EQ, just the speakers xovers. I made a one little change, LSP-CAD is just great for "testing", increased the value of parallel resistor from 8.2 Ohms to 22.4 Ohms. Much, much better result. Still room for improvements of course, just have to wait till an another sleepless night comes with new ideas 😉
Attachments
Hi,
with 200dB of diagram range everything looks flat and linear. Increase resolution (say to ~60dB) so that you can really see what happens! 🙄
jauu
Calvin
with 200dB of diagram range everything looks flat and linear. Increase resolution (say to ~60dB) so that you can really see what happens! 🙄
jauu
Calvin
Yep, but you all should know by now that it's all about the looks with this project (and that's failing too...), even the sound is judged here by the looks of the graphs 😉
Wife likes the new sound, she says that they sound more like "nothing" now. Looks, well she's not that exact with that (I'm a living proof).
Wife likes the new sound, she says that they sound more like "nothing" now. Looks, well she's not that exact with that (I'm a living proof).
Attachments
More bad parts added to the signal path. Random. Is the xover OK like that or did I break some rules of how things can be done? And it's really beautiful as well!
I'm not able to test the speakers now with a high sound levels, but at least with a moderate volume level all seems to be OK.
I'm not able to test the speakers now with a high sound levels, but at least with a moderate volume level all seems to be OK.
Attachments
OK, they are dead again. I did thrown them down from a window. Distortion levels kept going up with highs, like cymbals. Vocals were still crystal clear. Isolated the problem to the toroidal transformers. They are leaking like mad, sparks with a bit of power from 50VA amplifier. I think this it for me. I'll buy a commercial version or something.
Now listening a good pair of box speakers and they sound like $hit even compared to the not properly working ESL's. But the good music still sounds good. Now I don't have to considerate on the sound (since everything sounds ****) I can just enjoy the music.
Maybe tomorrow...
Now listening a good pair of box speakers and they sound like $hit even compared to the not properly working ESL's. But the good music still sounds good. Now I don't have to considerate on the sound (since everything sounds ****) I can just enjoy the music.
Maybe tomorrow...
Last edited:
I've bought Martin-Logan Purity's. My home made ESL's were better... These Purity's lack the wow factor and sharpness that my diy's had. I guess it's because of the smaller membrane size? Or the curved stator? On the bright side they're a lot easier to live with, you can even move your head a little with out strange effects. And these have bass, mine didn't.
So, I guess I've have to make another diy-pair someday. If only I had the stomach to start an another project...
Good luck to all DIY'ers, you need it or skills 😉
So, I guess I've have to make another diy-pair someday. If only I had the stomach to start an another project...
Good luck to all DIY'ers, you need it or skills 😉
Hi,
my experience is that the typical ESL magic is lost below a certain panel size or a too high crossover frequency. Then it will be just a decent mid-high transducer. Size does count after all 😀
jauu
Calvin
my experience is that the typical ESL magic is lost below a certain panel size or a too high crossover frequency. Then it will be just a decent mid-high transducer. Size does count after all 😀
jauu
Calvin
- Status
- Not open for further replies.