Am I on the right track? Bookshelf speaker Fountek FW146 / Dayton DC28F

[rpb]

Here is version 3 of my Peerless tweeter response curve. It took a little while to figure out how to get the phase aligned better. What ultimately seemed to get me closer was adding the parallel resistor over the woofer. Is the high impedance at 5kHz an issue? And is the impedance phase okay?
Thx

Try replacing the 30 ohm R16 in the tweeter filter with about 6 ohms. Most of the other values will change as well, but you may find it easier to get the tweeter to roll off as desired. R10 will probably drop to about 4 ohms or so. This will reduce the impedance of the speaker above the x-over frequency.

[crossbound]

After removing the large resistors in the highpass to start over, I tried again, but added resistors in series with the capacitors, which seemed to help to keep everything a little more even and more manageable. Currently the reversed null is not dipping all the way down. I could correct that by increasing the C9 and C10 capacitor 5.1uF, but it would elevate the SPL in the crossover regions just slightly.

[rpb]

There's a thousand ways you could design this x-over, so don't take my suggestions as an indication that you are doing something wrong. . Just as an experiment, try adding 1, 2 or 3 ohms for R12, and tweak the caps a little.

If you put 6 ohms in parallel with the tweeter, it will tend to level the tweeter impedance. That sometimes makes it easier to get the desired roll off. It will probably change the top end too.

[crossbound]

My question is then, which impedance range should I shoot for?

[rpb]

My question is then, which impedance range should I shoot for?

I don't know if it matters all that much. What I would probably shoot for is keeping the impedance peaks in the mids, and highs from exceeding 20 ohms, so yours looks fine to me. On my current speakers, I added a cap, coil, and resistor across the speaker input to reduce a peak down to about 8 ohms. I plan to compare the sound with, and without the added parts. I don't expect to hear much difference, but who knows. I'm keeping an open mind about it. The biggest potential advantage of adding the parallel resistor, is reducing the effect of the tweeters impedance on the filter. It's not always necessary, but sometimes it helps.

[Wolf]

Keep in mind that a tube amp won't like the wild swings in impedance. A Class AB will be fine with it. A Class D might have altered response in the top-octave depending on tweeter impedance in this range due to the typically standard output filter they use to suppress the switching frequency at output.

Later,
Wolf

[crossbound]

Once again thank you guys.
I'm not planning on using a tube amp so no problem there.
I guess my question now would be: is what I got acceptable, or are there any glaring issues I should take care of?
I will try some more tweaks, but sure would like to wrap this up. :-)

Wolf is there a certain impedance that you try to stay within to minimize the effects or are you designing for a certain type of amp in mind? I know next to nothing about amps.

[Wolf]

It was just a statement and info to be aware of. Being that most Class D filters (some have another method that negates the effects by not having an output filter) are generally 1st order and 0.01-0.022mH in use. You can use the classic mH equation; L = [R/(2*pi*F) ]* 1000

In general, lower impedance means lower Fc of the coil, and an earlier rolloff ensues. Most are meant for 4 - 8 ohms.

Later,
Wolf

[WernerM]

You are doing fine. At this point anything else you decide to try will be for sheer learning and practicing. I can't tell you the number of Sims I have done just for experiments and learning. Wolf is definitely more versed at the terminology, some of the things that I mean aren't necessarily how I convey them... lol. If you chose to continue more experiments trying to flatten the impedance phase swings, then by all means do so. You aren't bothering any of us I'm sure. I personally like seeing these types of threads as they are great learning resources for more people than just yourself. I did the same thing when I first started. I also messaged Wolf and rpb plenty of times about theoretical builds and Sims to get their insight. For me it has been a pleasure trying to give you help and advice. Wouldn't you know it, I picked up a couple of tricks myself.

[crossbound]

What I learned tonight: As I was trying to reduce the impedance a little bit, I discovered that L1, C2, C9 and C11 had the biggest impact on reducing the impedance bump at 1500Hz. I reduced L1 just slightly, and increased C9 and C10. Also realized that R7 controlled the impedance bumps at 20Hz and 70Hz. And finally some other slight adjustments to keep the phase in line.