Home Theater Forum and Systems banner

1 - 10 of 10 Posts

·
Registered
Joined
·
18 Posts
Discussion Starter #1
Well, first and foremost I am into audio, both home and car, big time. My home setup consists of a pair of B&W DM603 S3 hooked to a Denon DRA-697CI receiver. What I want to do isd build a set of bookshelf speakers for use at school or just in general. I want to use the Vifa XT25TG30-04 1" Dual Concentric Dome Tweeter for the tweeter and Tang Band W6-1721 6-1/2" Underhung Midbass Driver for the woofer. What I need is a crossover design, but I have no experience in this area. If anyone can at the very least assist me I would appreciate it greatly.
 

·
Registered
Joined
·
63 Posts
Are you looking for active crossovers (ie amplifiers for each driver in the speaker enclosure) or passive crossovers (one amp for all the drivers)?
 

·
Registered
Joined
·
63 Posts
I had a look at the driver specs via the links you posted.
Will you be using a vented enclosure design for the LF driver?
BTW, I have a speaker design spreadsheet on my site that may help with your cabinet design.

For your crossover you can probably get away with a second order passive design. For a Linkwitz-Riley filter at 2500Hz (ie -24dB at 600Hz), you can use 0.5mH inductors and 8uF caps. The HF driver you have selected gives about 3dB more output for 1W at 1m so you will probably need to attenuate it as well.

For the inductors, wind them as air-cored using 0.8mm or better copper wire. The construction specs for the inductors will depend on the formers you choose. Otherwise you can probably just purchase some ready-made from an electronics supplier.

You can make up 8uF capcitor using 4 x 2uF in parallel. Make sure the voltage rating is sufficient for the output of your amplifier. I normally use 250V polyester caps.
 

·
Registered
Joined
·
1,585 Posts
I think the post above would work in general, but you may not be happy with the results. Until you play with the actual drivers, it'd be impossible to know how a crossover would turn out.

In general, a 2nd order LR probably makes sense, but I couldn't tell you if the values for the parts above are good or not unless I had first hand knowledge.

There are several spreadsheets out there that you can plug the appropriate information in to get the values for the parts, but I really look at that as a starting point rather than the end.

That's the reason xo design is as difficult as it is.

JCD
 

·
Registered
Joined
·
63 Posts
Yes, it's a starting point. It has not included the Zobel network since we'll need impedance plots to work out the values.
 

·
Registered
Joined
·
1,585 Posts
Zaph has tested both of these drivers and all his results are on his site.
It's their results with the actual capacitors/inductors/resistors that I was referring to.

E.g., the Zobel network that gswan references is a circuit to flatten out the impedence curve -- the resistance of the driver changes with the frequency of the signal. This can cause problems with the xo as it is dependent on the resistance of the drivers.

JCD
 

·
Registered
Joined
·
18 Posts
Discussion Starter #9
It's their results with the actual capacitors/inductors/resistors that I was referring to.

E.g., the Zobel network that gswan references is a circuit to flatten out the impedence curve -- the resistance of the driver changes with the frequency of the signal. This can cause problems with the xo as it is dependent on the resistance of the drivers.

JCD
I know that, lol. I am studying EE, so I know a fair bit, but I do not know crossover design.
 

·
Registered
Joined
·
63 Posts
I know that, lol. I am studying EE, so I know a fair bit, but I do not know crossover design.
That's great. So you'll be aware that if you can create accurate electrical models of your drivers you can build a simulation of the drivers and filter network. The crossover and zobel network itself is basic filter theory, however you need to consider the fact that the drivers represent complex varying load impedances to the filters over the operating frequency range. This is what needs to be accounted for in the design. Then you need to adjust for driver sensitivity over the frequency range. Using a good analogue circuit simulator and good analogue models of your drivers will help you understand and play with the design without risking your drivers and having to fiddle with winding inductors until you are ready to test your design. Remember that the LF driver impedance should be modelled to include cabinet effects. The HF driver impedance is not affected by the cabinet.
 
1 - 10 of 10 Posts
Top