Playing with idealised drivers (FR and impedance response linearly smooth) may be nice for learning purposes. But you should not consider putting the results of such simulations into practice. Just in case this is not already clear.
I would also use simulations with data traced from data sheets at most for feasibility considerations, even before I start with the actual development.
In my opinion, there is no way around measuring.
I would also use simulations with data traced from data sheets at most for feasibility considerations, even before I start with the actual development.
In my opinion, there is no way around measuring.
Doing so creates that spike at 2k.
While you're playing around with sims, add a 1uf cap in parallel with the 2.825mH.
Or...
Add a 2 ohm resistor in series with C1.
Or...
Add a 1uf cap in parallel with the 1.413mH.
Incorrect phase matching. This can be seen in the crossover frequency: although the drivers do not add up, they do not cancel each other out perfectly either. Otherwise, reversing the polarity of the tweeter would make the whole thing look very different.
Phase diagrams of the individual drivers would also be useful, of course only if there was a timing reference during the measurements.
Phase diagrams of the individual drivers would also be useful, of course only if there was a timing reference during the measurements.
I didn't expect to even be close. I was just showing you some tricks.
Try reducing L2 to about 1mH.
Try some larger values for R5.
The frequency of the deep notch can be moved around by changing that 1uf cap value. A smaller cap moves it higher in frequency. I have some .1uf caps in my parts box, and some .22uf also.
Try reducing L2 to about 1mH.
Try some larger values for R5.
The frequency of the deep notch can be moved around by changing that 1uf cap value. A smaller cap moves it higher in frequency. I have some .1uf caps in my parts box, and some .22uf also.
Last edited:
Btw.: the vertical bandwidth of the diagram is about 120 dB. If you reduce this to 50 to 60 dB, you will be able to recognise more. Such a bandwidth is also more or less standard, at least in forums like this one (but probably beyond).
here we are, now that I have checked XSim and the various installation files, where I also found some examples and also the .zma and .frd files - all this, together with your suggestions, makes me think that the crossover in my other topic is far from finished/concluded. I'm thinking of a way to tell @Galu that the matter is not concluded.
😵
😵
Hey, FYI, dayton provides FR and ZMA files for all their drivers. You can always download a tweeter and woofer and try this. Also, push the woofer back some random amount like .25 to 0.75 inches so you have a realistic view of how phase matching will affect your end result.
You said the speakers sounded good. What do you want: blood?
P.S. I think the new contributors to your saga deserve this image:
Sims are fun!
What I like to do is start with the raw woofers response, and speculate on what I want to do with it. Then, keeping it on the same graph, I try filters. I use PCD (Passive Crossover Designer ). Since there's so many options to make a x-over, I try many, unless I have reason to think one approach is better than another.
What I like to do is start with the raw woofers response, and speculate on what I want to do with it. Then, keeping it on the same graph, I try filters. I use PCD (Passive Crossover Designer ). Since there's so many options to make a x-over, I try many, unless I have reason to think one approach is better than another.
Last edited:
Visually, it looks like your tweeter is closer to the mic, and listening position, than the woofer. This is common. The x-over has to account for that to make things add correctly. There should be a place in x-sim to estimate the offset, ie, you tell the software the tweeter is .6" closer. My estimate, and an estimate is close enough most times.
do you mean closer to the woofer on the vertical axis?
can I measure...from the center of the Tweeter to the center of the Woofer?
can I measure...from the center of the Tweeter to the center of the Woofer?
No. The cone of the tweeter is closer to the mic than the cone of the woofer. This is the case with almost all speakers.
If you had a slanted baffle, the distance might be equal, or very close to equal. Yours is about 5/8". Just an estimate is all you need.
No need to measure, but the estimate needs to be entered somewhere in the Xsim software, a long with the height difference of about 6". That you can measure if you want.
If you had a slanted baffle, the distance might be equal, or very close to equal. Yours is about 5/8". Just an estimate is all you need.
No need to measure, but the estimate needs to be entered somewhere in the Xsim software, a long with the height difference of about 6". That you can measure if you want.
ok, now I understand. yes, in these girls the tweeter is a little different from the Orig. LS 3/5a. in this case I used a 4ohm slightly larger dome and it is also positioned slightly more towards the woofer by about 4mm - or the thickness I used for the fixing plate. let's say that compared to the woofer subtracting the thicknesses of the front, 5.5/5.7mm. in the LS 3/5a with the classic T27 flush with the panel/front, about 9.5/10mm.
I use buffle specs. originally: 9.5mm
I use buffle specs. originally: 9.5mm