'Upgrading’ my tweeters (can I do this?)

Jump to Latest
#562
First order crossovers are usually a nightmare. Big drivers overlap, a lot of gain and loss in the range, problems with the polar characteristic, tweeters overloaded with low frequencies and the like. The drivers must be aligned in a vertical plane. The advantage is the lowest energy loss, no need for elements with tight tolerances and the lowest price. If a lot of corrective elements are needed, then they can be very complicated, which means more expensive.
 
  • Like
Reactions: tone?
#563
Stanislav said:
You can try a Series XO for the fun, but IMHO only 1st order Series XO is worth the effort and pays handsomely soundwise at least in my book.
I personally do not like 2nd and 3rd order filters and considering the quality of your drivers they will handle 1st order series without too much blinking.
With the right drivers Series XO is unbeatable I think.
Click to expand...
I totally agree. Can you tell me more about your speaker system and why you prefer series xo over parallel 1st order?
thx!
 
#564
And that interaction of woofers and tweeter really, REALLY matters. As Proac might (and should) have optimized not only on axis response, but power response and lobing too, you only should mess with the crossover if you’re prepared to do the full set of measurements on and off axis. You can import that all in VituixCAD and optimize further. But don’t go changing component values randomly, judging by your ears. They will fool you.
 
  • Like
Reactions: NIXIE62, tone? and fatmarley
#565
Stanislav said:
Well, here you are comparing 3rd and 2nd order parallel filters, not what I was talking about.
This is what a 1st order series xo looks like, it's a different project with different drivers but the idea is the same.
Actually I think if you try your speakers with the same xo values as per that file the result will be quite good. Then you can optimise.
Of course if you want to try high order filters go ahead and do it. I've been there and done that, long, long time ago.
Click to expand...

stan,

wouldn’t that xo cause issues at higher volumes though?
as in distortion from the excessive bleed of drivers?

thanks !
 
#566
tone? said:
i know this kinda sounds weird but I’ll know it when I hear it again. I do love the way they sound now.
if I had a bit more immediacy it would be nice.
little more snap so to speak.
My speakers have a lush bit midrange. I’m sure the MTM helps with that.
Click to expand...
You wrote you have literally spent months on speaker placement to squeeze all you can from your speakers to sound their best.

As you change crossover filters, the three different driver's phase relationship changes, causing dispersion changes, in effect changing the room's response to the speaker placement. "Breaking in" your hearing to those different "comb filters" takes time.

The "lush" midrange, or "acoustic smear" resulting from three individual sources combining on and off axis reduces "snap" and "immediacy".
Using a lower crossover point on the lower driver would reduce some of the overlap, increasing your perception of detail.

B&W uses that 2.5 way crossover approach on some of their speakers with similar topology to yours.
If you have used separate wires to connect the woofers, you may want to experiment with that.
That said, realize that the actual acoustic crossover, rather than the order (1st, 2nd, 3rd, 4th etc.) is what determines frequency and phase response- your two different woofers using the same filter set have different rolloffs, and each change you make results in a different slope for both, not the single one as depicted in your simulations.

Cheers,
Art
 
  • Like
Reactions: tone?
#568
markbakk said:
And that interaction of woofers and tweeter really, REALLY matters. As Proac might (and should) have optimized not only on axis response, but power response and lobing too, you only should mess with the crossover if you’re prepared to do the full set of measurements on and off axis. You can import that all in VituixCAD and optimize further. But don’t go changing component values randomly, judging by your ears. They will fool you.
Click to expand...

thanks so much Mark

i am in reality making once super minor adjustment.
i don’t want to make major changes because as I mentioned I already like the speakers. Just small small adjustments.
Arent MTMs one big plus having a great dispersion which solves a lot of lobing no matter what order filter you use?
that is what I have read and how it works.

i understand and you are correct that to make any changes I would have to take many measurements. And possibly have to buy something like DATS v3 so I can take accurate impedance sweeps of my actual drivers.
 
Last edited:
#569
tone? said:
what do you mean about separate wires for both woofers and experiment with that?
yes they are separate.
you mean make slight variations on one woofer ?
Click to expand...
From what you posted previously, it appears your speakers are not summing flat on axis around the acoustic crossover point, which may be destructive interference. MTM lobing... you may be calling that a "BBC dip".
?Xover .png

2kHz appears to be roughly the acoustic crossover point, there should be no dip at that point.
Rather than slight variations, I'd suggest rolling off the bottom woofer an octave or two lower.
That could level off the acoustic crossover dip, and also reduce the peaks in the 1000 to 1600 range.

IIRC, you have purchased new coils and caps, using series or parallel combinations you have enough parts for (near) endless possible combinations.
Putting two coils in series side by side, or on top of each other opposite winding or the same direction gives three different inductance possibilities without any unwinding. Putting steel bolts in the coils adds inductance- for low volume testing (or apartment volume...) there is no worry about core saturation.

Measure, listen, compare, have fun!

Art
 
  • Like
Reactions: youknowyou and tone?
#571
youknowyou said:
series xo over parallel 1st order?
Click to expand...
The first order two-way series crossover is a cool thing with the system adding to more than the sum of its parts. IIRC you at least get greater than 1st order slopes up to 2nd, impedance compensation and compensation for the directivity hole between woofer and tweeter when good driver choices have been made. If these were to be measured and duplicated in a parallel configuration it would require more parts.
 
  • Like
Reactions: Stanislav and youknowyou
#572
You have all the gear to make the measurements, I think. Apart from constructing two voltage dividers and 22R plus a 33R 5W series resistor together in the impedance jig you don”t have to undertake anything. I think I posted the links for this already, look for the REW and ARTA/LIMP manuals.
 
  • Like
Reactions: tone?
#573
Yeah I’m not doing radical modifications or anything.

pretty much just copying the original xo.
once I get that right then maybe I throw in a component.

but I need to get the new xo to emulate the old one first

I actually was just unwinding the air cores for the woofers a bit , because they were 16.5mH and not 16mH.
so I unwound them until they were perfectly 16mH

the actual iron core inductors that Proac use call for 15mH with a DCR of 0R3. according to the serial number of the iron core. And the serial number of the tweeter iron core should be 0.2mH DCR 0R1 not 0.22mH DCR 0R2 that the originals measure.
so i might try that as well later on. First I need to emulate the original as I said

to be honest the tolerances in their xo aren’t the best. But ok for this range of product.

good thing the caps in the tweeter circuit were spot on.
 
Last edited:
#574
I am going to solder them soon , ’ loosely’ no twisting of wires, just in case I need to make a change.

once I get things ’ right’ then I will pretty up the board and make everything as pro as possible.

im in the R&D phase now. Lol
 
#577
tone? said:
Two bass traps

and five panels.
Click to expand...
That is very vague (like saying you use an amp and speakers) ... what kind of bass traps and panels?

Usually more treatment is required at low frequency ... what kind of construction is your room?

How do you know it is your speakers that are the problem not the room ... ?

With respect I just get the feeling your chasing the dragon without a plan or objective ... it all seems very random.

Actually I just realised that I don't even know what it is your trying to fix?
 
  • Like
Reactions: tone?
#578
This is exactly how I educated myself as to what does what to perceived sound, it really paid off for me in of the fact i can EQ with the XO at listening position, yes thats taboo…..but unless you or your system is going to move around alot imo its the icing on the cake.
 
  • Like
Reactions: tone?
You must log in or register to reply here.
Top Bottom