DS135-8 and XT25TG30-04 Crossover considerations

[jaste]

Hi!

I realized I already had most parts to build one of theese

http://projectgallery.parts-express....helf-speakers/

It uses the Dayton DS135-8 with a Vifa DX25 tweeter. I have all crossover parts on hand and the woofers but a pair of XT25 instead of DX25. I substitutet the DX25 with the XT25 in Xsim, loaded FRD and ZMA files for the drivers, and added a second capacitor for the tweeter (the 10uf one). This is the first time I´m fiddling with crossovers and first time using Xsim so I´m quite a rookie at constructing speakers, only used to building them. To my rookie eyes the sims look rather ok, but looking around the web the XT25 seemes to be mated to alot more complicated x-overs. The x-over point is somewhere around 2700 hz if i understand correctly.

Does anyone have any input? I understand a more elaborate crossover might do a better job but my budget is low and I was happy when i realized I already had most coils and expensive parts on hand.

[jaste]

And here we see both drivers

[LOUT]

I feel like the baffle-step of only 2-3db might make the low-end sound a little thin unless you're planning to have these on a desk/table or very near the wall behind them, but I think the tweeter with your less complicated crossover looks like it should work really nice as you've modeled it. I'm also pretty sure a larger BaffleStepCorrection would need you to replace the 0.8mH coil on the woofer with something more like 1.0-1.2mH while messing with the tweeter's Lpad to follow, and I'm guessing those are parts you don't have on hand. Plus it's not like you can't tweak it a little later if you decide you want to change it.

[Chris Roemer]

I agree (esp. about the BSC). They seem "okay" (if you're happy).
How do they sound to you? Any complaints?

[jaste]

Oooos, just realized I posted in the wrong part of the forum! Should have posted in the Tech talk audio section since I havent built the speakers yet. Sorry about that and moderators please move if it’s against forum rules. I guess I’m obligated to build them now and post pictures here...

Thbaks for your input!

Regarding the BSC, I will have the speakers placed fairly close to a wall on a sideboard so that will help i guess, but if I were to start fiddling with the crossover it’s the LC secrion that is affecting the BSC if im not missunderstanding it, as well as being a second order filter. Would adding a second coil after the capacitor help maybe, ”cascading” is something I’ve read somewhere...

[LOUT]

I believe you can add 2-3db more of a bafflestepcomp to that design like you mention by adding a coil after the capacitors in the woofer's low-pass, BUT I'm pretty sure it'll still have to be something like 0.8mH and you'll want to run something like an 8ohm resister parallel with the added coil to avoid it making a nasty peak right before the drop-off.

[jaste]

Sooo, I’ve finaly made some progress with theese. I should have added volume for some bracing, and I found some 4mm roofing felt that i could line the insides with but again i’m worried that i reduce the alredy small volume in there. I’ve also fiddled a bit in xsim with the bsc but I dont know if its an ok implemantation or not. Hopefully I will finish them the coming weeks and be able to listen to them and try some things out.

[Chris Roemer]

Your BSC LOOKs like it's occurring in the 1000-1500Hz range. This is more typically applied between maybe 200-1000Hz. By your sim I'd expect an elevated midrange presentation.

[jaste]

Yes it doesnt kick in at around 600 wich is what i got from an online calculator based on my baffle widht. I will try my first version of the x-over out and see how it sounds.

[Chris Roemer]

To get the BSC right, you normally need a larger series coil on the woofer.
Try - LP: 1.5mH low DCR (iron-core) series coil, then the 8uF shunt cap (to gnd), and a "Zobel" using 15n(ohms) + 10uF.
HP: 3.3uF series cap, 0.40mH shunt coil; L-pad: SR=4n / PR=8n.

This should at least get your BSC in the right range, and set the tweeter level appropriately. The Fc is near 3kHz.
To get the "summing" right, you have to use "minimum phase" .frds and also dial in your drivers' X,Y,Z coordinates. Have you done that?

[jaste]

Yeah I really should get a bigger coil (1,5mh instead of the 0,8mh), but like the maker of the original sade theese will probably be placed quite close to a wall behind them and hopefully get some support from that. Part from that you advice me to take the 10 uf and use it in the LP section instead is that right? Or would that only work if I were using a bigger series coil. My values for the L-pad seems to be a bit off compared to yours (and I truly trust your numbers more then mine considering your knowledge), I will try yours.

Regarding the summing and minimim phase, I used .frds from Dayton and for the Vifa I dont remember what site I found it on. Honestly I have little too know knowledge about summing and phase and no clue about dialing in coordinates. I have just copied the original construction and changed tweeter (I found a speaker made by Troells Gravelssen if i recall correctly where the dx25 and xt25 were interchangeble but with a change in L-pad).

[jaste]

So! Ive got them up and running and sounding quite nice, without having anything to benchmark against, but I would like to try the filter Chris wrote about previously. I do have a par of 1.4mh air core (1.4mm thick wires), Chris you say iron core, how of is it to use air instead? Thanks!

[Chris Roemer]

Sounds close to #15 gauge, which should have a DCR around 0.4ohms. An iron-core'd be near 0.2?
Should be fine !

[Wolf]

That resistor across the 0.9mH will likely get hot. Above the Fc of the shunt cap, that resistor is directly across the amp terminals.
Please revise this.
Wolf