Jims 3 way speaker build

[jimbones]

Here is a current version of the XO. I didnt have all the correct values so I have to order some parts. For some reason snip wont give me the option to save images anymore..just vanished lol

[DeZZar]

You can actually just 'paste' the snip directly into the text input area here when making a post.

[jimbones]

@DerZzar understood, I meant I could not save a snip as a jpg to my "Pictures" folder. I did some research and they said that a Windows security patch disabled the "save" feature. A lot of good that is ha ha

[DeZZar]

Hmm...thats annoying. Mine still has File->Save As available.....I'll have to watch out for this patch!!

I guess you can still just paste it into Paint and save it from there?

[rpb]

I'm not sure if I used the right files, but here's what I came up with.

[jimbones]

​[ATTACH=JSON]n1493055[/ATTACH]

test

[jimbones]

I'm not sure if I used the right files, but here's what I came up with.

Wow thats nice do you have the schematic to compare with mine? I like your idea of "sneaking up" so my idea is to get something I can listen to and make improvements or shall I say refinements along the way, Eventually I would like to understand about more than flat frequency response but phase response. Thats the advanced class. Thank you to all for your help.

[rpb]

Here's the mid filter. There's likely a less expensive way to do this, but this looked better in the sim.

[rpb]

Padding the mid like I did, helped get the impedance up. The phase may not be as good as the sim shows, but that's not all that important. Ideally, you want to measure the actual response, but at low frequencies, it's difficult to get good measurements.

Here's a sim without the big coil. It just requires the mid to work a little harder below 200hz. Likely not a problem.

[rpb]

You should add some resistance values for the coils in your sim. The ones I used were just made up. They can change the response slightly.

[rpb]

I lifted the top end by bypassing some of the tweeter resistor padding with a 5uf cap. Or was it 12uf? I forget. Somewhere in between maybe.

[rpb]

As you can see, how you pad a driver can change the xo parts needed. I was trying to get the final impedance up above 10 ohms for most of the frequency range. Not that it's a good idea, or anything, I just was trying to do it. You could try padding values using smaller resistors than I did, and see if it makes things easier, or harder,

[rpb]

I added too many comments in the post above. You may have to click on the top right corner to see some of them.

[jimbones]

rpb, I got your points above. I think the setback values for the mid were 2.18 in and for the woofer 10.5 in. Thats the numbers I am using in my sim. I dont know how well XSim and WinPCD track. I;ll try to mimic yours so we are at least in the same place.

[rpb]

rpb, I got your points above. I think the setback values for the mid were 2.18 in and for the woofer 10.5 in. Thats the numbers I am using in my sim. I dont know how well XSim and WinPCD track. I;ll try to mimic yours so we are at least in the same place.

I used about .75", and 3". I think the woofer might be closer to 4". I don't think the AC is ever behind the woofer, but I could be wrong. I make my x-overs by trial and error mostly, and don't give much thought to AC. I measure the response instead.

[jimbones]

rpb this is a hybrid. I used your mid bandpass and used my LP and HP for woofer and tweeter. If you think this is at least a good starting point I am going to mock it up.

[rpb]

The impedance gets dangerously low at high frequencies. There's a straight path to ground through C3, and C1.

Add a resistor in series with C3, and see how that looks. I can't sim that added resistor in PCD. I'd add about 3 to 6 ohms, and see how it looks.

I think we are using different frd files, but basically, your filter looks to be OK, but with a downward slope.

Try bypassing your tweeter 5.1 ohm resistor with a 6 to 10uf cap.

I have to go out for a while.

Try this on the woofer. 8mH, 100uf and 1.5 ohms to ground, followed by 2mH. This is a 3rd order, but with a little resistance added.

[jimbones]

made recommended changes. The impedance is better now. I dont mind a bit of a tilted down high end but I can tailor that tilt with cap values. 4.7 seemed to be pretty good,

[rpb]

Let me show you my tweeter, and woofer filters.

[rpb]

The thing about a tank filter is that frequencies above the tank notch come back strong. When I simmed your woofer filter, there was woofer output around 3k that should have been way down. Try a second or third order filter. Or a damped third order like mine. You don't have to cross at 300hz like I did. If you have the 7mH coil, maybe change the mid filter if needed, and cross a little higher.

My tweeter filter is weird. I was trying to raise impedance. I normally use smaller resistors on 2-ways.

[DeZZar]

Here's what I get with the last setup you posted (assuming we're working with the same files)

• You don't need that extra 4.7 on the tweeter
• The midrange lowpass is better second order
• I don't think you need the tank on the woofer. The extra resistor after 82 cap causes issues
• Impedance is all over the place

​




Here is what I would suggest instead:


​​​
​