Jims 3 way speaker build

[jimbones]

maybe this is better??

[johnny5jz]

You compare the system curve to the get file curve.

[jimbones]

thats weird Im using the system curve (all 3 drivers no XO) and the imported frd of the 3 drivers I measured and they look nothing like yours.

[jimbones]

My attempt

[Billet]

That looks pretty good to me.

[rpb]

I don't have my computer available at the moment, so I can't try this myself. Try adding a notch tuned to 800hz on the woofer, and then add a coil for the low-pass filter. Eliminate, or greatly reduce C1.

[jimbones]

Im not following, why 800hz? by adding a coil going to 18db/oct? Not sure what the advantages are adding 3 parts for one fairly large C1?

[rpb]

To eliminate the bump in the woofer response at 800hz. If you eliminate, or reduce C1, it might raise the impedance slightly, and might blend better with the mid. The roll off will not be steeper, in fact it might be shallower. I know it's more parts. The real question is will it sound better? The only way to decide is to sim it, or to actually try it. I'm not saying that it will be better. I was curious how it looks in a sim.

[jimbones]

I may have found part of my problem, apparently I was way behind in software versions of Omnimic. I dont believe I have to remeasure everything do I?? GaHH!!

[jimbones]

johnny5jz I redid the curves and I believe it came out better and our numbers agree. Let me know what ytou think and if I did this correctly.

So if I understand what we are basically doing is comparing our individually measured responses (summed in the tool) with the measured system response of the 3 drivers in parallel. the tweeter reference is always zero and the mid and woofer are set back so we are "adjusting" those to match the responses. Once we have this data we design the crossover taking into account this setback. So the million dollar question is Does this mean I will have effectively designed a XO that will have acoustic centers aligned?

[rpb]

Do you really think that the woofer AC is 10" behind the tweeter? I don't.

[johnny5jz]

The woofer is the lowest in the cabinet and has a delay from both its physical distance behind the tweeter, and its height below the tweeter. That number is not out of line for a 3-way.

[rpb]

Perhaps xsim is different than PCD. In PCD the AC is relative to the tweeter AC, not the mic.

[Chris Roemer]

Uh, most likely not - depending on several things.
"Passive" filters (caps & coils - post amp) "rotate" the drivers' phase, forwards/backwards depending on whether they're "inductive" or "capacitive" - the amount of rotation generally increasing with the "order" of the filter (1st, 2nd, 3rd... ).

In XO design, you're basically just aiming to get adjacent crossing drivers to both move in and out together ("in phase"). All "in phase" drivers are not necessarily "time-aligned", while "time-aligned" drivers WILL BE in phase.

[jimbones]

Uh, most likely not - depending on several things.
"Passive" filters (caps & coils - post amp) "rotate" the drivers' phase, forwards/backwards depending on whether they're "inductive" or "capacitive" - the amount of rotation generally increasing with the "order" of the filter (1st, 2nd, 3rd... ).

In XO design, you're basically just aiming to get adjacent crossing drivers to both move in and out together ("in phase"). All "in phase" drivers are not necessarily "time-aligned", while "time-aligned" drivers WILL BE in phase.

Yes I understand regarding the reactive components and order of filters. So what is the best way to go about this. Do I have the info I need to proceed. Am I proceeding correctly? Should I be attempting to do a design that is "time aligned"? Inquiring minds would like to know ha ha.

[rpb]

Yes I understand regarding the reactive components and order of filters. So what is the best way to go about this. Do I have the info I need to proceed. Am I proceeding correctly? Should I be attempting to do a design that is "time aligned"? Inquiring minds would like to know ha ha.

You have parts, so I'd start trying actual filters. Your measurements are at least suspect to me. Your AC calculations as well. That doesn't really matter. You can still make a x-over starting with your sim, and tweak / redesign from there. I hope that you have lots of test leads for connecting parts.

Since you are unable to move the speaker, I'd start with the mid filter, and look closely at the response. Next the tweeter. Woofer last.

You want responses to look like your target if you have that feature. (PCD does.) If you miss the target, you might can change it to match the response. In other words, if you were shooting for a 3k cross, but it came out perfect at 3.5k, then maybe use it. You can always change it later. The initial goal is to get it in the ballpark.

[jimbones]

Ill start with the mid

[rpb]

Try to get nice smooth roll offs. They will blend better.

[jimbones]

Todays update. I did frd and imp of each driver individually through the XO. your advice is appreciated.

[rpb]

The mid xo looks nice! Did you measure the summation of the mid, and tweeter? Have you gotten far enough along to listen to it?

[jimbones]

Lol I didnt sum them because I was waiting for someone to tell how far off I am and doing it all wrong lol!! ! That is my next step to do all 3. I only have the XO for one speaker. I didnt want to build 2 useless crossovers.

[rpb]

You have to kinda sneak up on a nice response!

If two drivers are in "good" phase, they add 6dB higher than alone. See if the mid, and tweeter are each down 6dB at the xo point. Your post is small on my screen, so I can't tell for sure. Play them together and measure. If you get a big dip,that's OK. Flip the polarity of one driver, and the dip should go away. What you want is flat one way, and a big dip the other way.

The woofer is probably close, but your measurements with it near the floor will make it harder to know.

Look at your sim to verify the impedance is safe for your amp.

[jimbones]

ummm woofer is really strong

[rpb]

Flip the midrange polarity That dip at 2.5k will probably flatten out, maybe the dip at 400hz too.

[jimbones]

you beat me too it because my XO schematic shows the nmid should be inverted but isnt wired that way

[jimbones]

still real strong. the woofer is not far from the floor so it is getting support from the boundary. damn..