The Joudas project

[trevordj]

I'd drop the notch that uses the 10mH coil. You are loosing too much sensitivity. Whatever the woofer response is at 300hz would be my approximate target sensitivity.

Here we go. My pocketbook thanks you for this recommendation. That notch filter wasn't doing as much as I thought and it more than doubled the cost of the crossover. The 100uF cap alone was $25!

[rpb]

Try a 3rd order on the tweeter, or maybe back the woofer xo down to a 3rd order.

[LOUT]

I have zero experience building passive crossovers so when, for example, someone like you looks at a crossover network and can tell from the schematic that something will get hot I have zero clue why that is.

If you're using Vituix for these crossover sims, clicking "VIEW" near the top/left and selecting "PowerDissipation" will display a graph that lets you see how much wattage is hitting each part in your crossover (with checkboxes for things you want to show or hide). It should let you adjust the amount of power to match what you'll be sending the whole speaker-box and then see if that exceeds the limits of your resistors. Some designs might need you to pair resistors together so they can share/split the load, but often it'll probably be better to tweak the XO so it isn't sending that much power into a resistor in the first place.

[trevordj]

Thank you!

[trevordj]

Try a 3rd order on the tweeter, or maybe back the woofer xo down to a 3rd order.

Ok, I tried a bunch more. Here is the best I could come up with:







Vituixcad kind of allows for target curves. Through the optimizer program you can get it to show, it just won't show both filters at the same time. I tried running the optimizer a bunch of times and it really just makes a mess of things. I think this may be my favorite crossover so far. Although phase alignment could be better I get a nice deep null.

[rpb]

Well, now you should see a trend in part values that you may need. Don't expect the sim to be better than plus or minus 20%. Buy extra parts, since you have the other set of speakers too. PE will allow returns if the parts look unused.

[trevordj]

Well, now you should see a trend in part values that you may need. Don't expect the sim to be better than plus or minus 20%. Buy extra parts, since you have the other set of speakers too. PE will allow returns if the parts look unused.

Cool! I went ahead and purchased some components. I bought enough to make the top three designs x 2 so it ended up being a lot of parts, and like you said, many that are close in value. If anything I will have a good start on crossover component stores. -Trevor

[trevordj]

Well, I have nothing but time until the crossover components arrive so I am going to build an accompanying subwoofer for my son to go with these speakers. I also need to design some stands for these to go on.

I have four Dayton RSS210HO-8 subwoofers so I am going to build two boxes (my older son will get one and my younger son will get the other). Eventually I am going to have to build some speakers for my younger son too. Since this thread has somewhat become a stream of consciousness for these different things I'll just keep it going. I ordered a 250 watt plate amp from PE with DSP but it won't be in stock until November. I will probably wire these in series to 16 ohms as we don't want him rattling the walls.

Here is what I am thinking:

[rpb]

Tune it to 22hz, and add a 60hz low-pass

The 100w amp seems plenty for my use.

[trevordj]

Tune it to 22hz, and add a 60hz low-pass

The 100w amp seems plenty for my use.

That tune is just a bit on the low side for my taste, for music anyway. I would have to figure out a 34" long port to do that and I think tuning to the mid 30s with an F3 of 30hz (lower with some room gain) will sound nice for music.

[rpb]

Yes, the box would need to be bigger for that to work. Maybe 100l.

[rpb]

I found a file on my computer for a 4 ohm NE180. I altered the impedance, and came up with a woofer filter that might have a chance. Third order using 3.7mH, 10uf, and 1mH. I'd be curious to know if that looks close at all.

[trevordj]

I found a file on my computer for a 4 ohm NE180. I altered the impedance, and came up with a woofer filter that might have a chance. Third order using 3.7mH, 10uf, and 1mH. I'd be curious to know if that looks close at all.

Ya, interestingly that was really close to one of the crossovers I already had (3.7mH, 12uF, and 1.5mH).

[rpb]

Looks like the tweeter padding needs some adjusting, but the woofer response looks like an easy spot to start when you get parts.

[rpb]

So, when you start measuring, and assembling a xo, will you do it where you can play music on one, or both speakers quickly? When I work on a xo, I listen if I get close. Even if I know it will be changed, I listen to see how good or bad it might sound, and to see if what I hear matches the measurements. If I hear a peak in the response, I like to see it in the measurements. This also helps you gain experience in identifying what needs improving.

[trevordj]

So, when you start measuring, and assembling a xo, will you do it where you can play music on one, or both speakers quickly? When I work on a xo, I listen if I get close. Even if I know it will be changed, I listen to see how good or bad it might sound, and to see if what I hear matches the measurements. If I hear a peak in the response, I like to see it in the measurements. This also helps you gain experience in identifying what needs improving.

Definitely, I have them in my home theater and I can play both movies and music on them. Training my ear is definitely something I will have to work on. I like the idea of correlating subjective and objective measurements.