Please check out my xo design

[duanebro]

Ok, just a very simple xo. First the speaker:

And the modeled FR:

Measured:

And the xo: well I don't have a pic

Woofer:
inductor: 3mh
cap: 4.7mf

Tweeter:
3 ohms before xo (1.5 models better but measures hot)
cap: 4.7mf (3mf give a better reverse null haven't had time to test it out)
inductor: .2 mh
10 ohms series

Any big issues?
Thanks,
Duane

[Whitneyville1]

I'm ignoring the sub 200Hz "hump" as a room effect. Other than the apparent low sensitivity, the Bravox buyout and the buyout tweeter seem to be "playing nice" with each other. I wasn't sure the blues would play to XO at 4KHz+ without ugly break-up. Your sim and your measurement are with-in reality, to me. I'd like to see a schematic of your XO. You didn't put 300 parts in it! Play with the tweeter XO for sound (to your taste) at this point. Is that the 12 ohm woofer? (3mH makes me think so.) Johnny Richards would be proud of you! Looks like you may have gone about 20 liters ported, F3~40-45 Hz. Nice cabinet work on those too. I think you've got it nailed. If they sound as good as they measure, you should have some really nice speakers.

[Mark65]

Do you mean 10 ohms across the tweeter? If so, that would be parallel, and if not, you've got a HELLuva lot of padding!


Looks pretty good to me, in either case, and I don't see any glaring issues, although you didn't post the impedance curve, or the electrical phase, or the power response, so my approval would have to be conditional....


Mark

[duanebro]

Do you mean 10 ohms across the tweeter? If so, that would be parallel, and if not, you've got a HELLuva lot of padding!


Looks pretty good to me, in either case, and I don't see any glaring issues, although you didn't post the impedance curve, or the electrical phase, or the power response, so my approval would have to be conditional....


Mark

I do mean in series. And yes it is a lot of padding. The Electrical phase and power response are in the pcd screen shot. I uploaded the wrong one and edited the post a few minutes later. You could have seen the wrong info. Here is the Impedance:


I know the phase angle is pretty bad at 2k-3k. But I don't think it should be a problem because of the high impedance at that point.

Thanks for your replies,
Duane

[arlis_1957@yahoo.com]

your acoustical phase dose not look good to me. have you tried 3rd order with reverse polarity on that little guy? are you getting a good reverse null now?

[duanebro]

your acoustical phase dose not look good to me. have you tried 3rd order with reverse polarity on that little guy? are you getting a good reverse null now?

I started out with 3rd order. I dropped the last cap because if didn't seem to have an effect on the fr. I know the phase is not good. So I went back to 3rd order (6.8 mf) and the phase looks much better below the xo point. Nice reverse null at 3.5k. I have to run up to the store for some bacon. I will try to post a pic after breakfast

[isaeagle4031]

MMMMMhhhh bacon!

[duanebro]

Third order on the tweeter:


Still doesn't line up pretty, but the r.n. is about 35db deep.

[Jeff B.]

Third order on the tweeter:


Still doesn't line up pretty, but the r.n. is about 35db deep.

Don't worry about lining it up - what you have is fine. Now, can I have some bacon? I love bacon....

[Mark65]

Hmmm, never tried a resistor before and after the xo like that, interesting. And I agree, it looks great. I'd call it done, based on that, but how does it sound?

I like it!



Mark

[johnnyrichards]

You will grow to enjoy that Bravox woofer. Helluva series of drivers.

[duanebro]

So I was fooling around with the the tweeter xo and found something I think you might like to see: First I removed the padding on the tweeter -

Then I adjusted the xo to match what it was before - but at a higher/non padded level-

Then brought the level back down with an L-pad. The impedance was to low so I added a resistor back in before the xo and adjusted the L-pad again-

The response is almost the same as before. But the caps are larger and I needed a another resistor. Plus the impedance is lower and the impedance phase is over -60 degrees around 2K. Notice the grey line around 4K. This is the FR before my changes. It is almost totally covered by the new result.

[ReissM]

Third order on the tweeter:


Still doesn't line up pretty, but the r.n. is about 35db deep.

Doesn't line up? Huh? What you talkin 'bout Willis? Look at the phase curves... orange and sky blue. At the crossover freq the phases line up perfectly. That's why your reverse null is significantly deeper in this XO design than the other one you posted. Looks to me like you got the phase right in the above graph.

Now as far as phase-tracking goes, well, we can't have everything. Don't get distracted by the fact that the orange and sky-blue curves WRAP at different places... that doesn't matter. What is important is how well the phase curves track each other near the XO point. (and that the freq response looks good too!!!)

In a perfect world, the woofer and tweeter phases would overlay on top of each other for about two octaves... one octave below XO and one octave above. But it's really tough to get the frequency response to look good at the same time as the phase.

So just worry about getting the phases as close as you can at the XO freq. while keeping the freq response reasonably flat. In the graph above, you did just that. Nice!

[arlis_1957@yahoo.com]

Doesn't line up? Huh? What you talkin 'bout Willis? Look at the phase curves... orange and sky blue. At the crossover freq the phases line up perfectly. That's why your reverse null is significantly deeper in this XO design than the other one you posted. Looks to me like you got the phase right in the above graph.

Now as far as phase-tracking goes, well, we can't have everything. Don't get distracted by the fact that the orange and sky-blue curves WRAP at different places... that doesn't matter. What is important is how well the phase curves track each other near the XO point. (and that the freq response looks good too!!!)

In a perfect world, the woofer and tweeter phases would overlay on top of each other for about two octaves... one octave below XO and one octave above. But it's really tough to get the frequency response to look good at the same time as the phase.

So just worry about getting the phases as close as you can at the XO freq. while keeping the freq response reasonably flat. In the graph above, you did just that. Nice!

yeah, he is surly correct. There is just no way to line up phase AND have a good fr.

[mdocod]

Hmmm, never tried a resistor before and after the xo like that, interesting. .....

Mark

Hi Mark,

I see what us DIY folks would consider unusual placement of resistors pretty often in commercial designs. For DIY, before the x-over is practical for tweaking, for mass production, carefully placed padding can reduce component values and costs.

Regards,
Eric

[Whitneyville1]

You can get pretty squiggly lines or beat that buy-out tweeter into sounding right. With more parts you might do both, but how it sounds now seems more important. Listen to some tunes and decide if you're happy with them. Other than that 3mH (I assume iron core inductor)i it doesn't appear you have much money in these (YEA!!!!!) Here I didn't think there was much use for that buy-out tweeter...

[duanebro]

You can get pretty squiggly lines or beat that buy-out tweeter into sounding right. With more parts you might do both, but how it sounds now seems more important. Listen to some tunes and decide if you're happy with them. Other than that 3mH (I assume iron core inductor)i it doesn't appear you have much money in these (YEA!!!!!) Here I didn't think there was much use for that buy-out tweeter...

Yes, the iron core inductor is the biggest expense I have in these. I picked up the drivers at the PE tent sale last year. I don't like the drivers all that much. But I got a lot of experience with these and a 2.5 way I built using more of the same drivers. I think this sounds better than the 2.5 way (except for bass).

I have changed the xo a bit since I posted it, mostly because PE is out of .2 and .25 mH inductors. I used a .15 mH and adjusted the caps to bring the speaker back flat. I measured it earlier today and found that it is +/- 2 db above ~250 Hz. (I used 1/6 octive smoothing.)