The Joudas project

[trevordj]

Ok, I finished the practice box and, for simplicity, I took on axis measurements at 1m and nearfield. I elevated the box 4ft off the ground on a ladder and it at least 4ft from all walls.


Far field on axis woofer sealed box:



nearfield woofer sealed box



I am seeing an unexpected blip at around 500hz which is also showing up in the impedance measurement. Initially I had some box resonance (no filling)



Some of it cleaned up with stuffing but that spike at 475hz remains. Is this something I would need to worry about if I was building a speaker with this? What is likely to be causing this?



And here is the combined frequency response after running through response blender



Tweeter response

[rpb]

Suggestion.... Crop that measurement from about 80hz to 20k, and it will be easier to interpret.

Nearfield will do odd things up around 1k. The problem is that some of the sound is from the dust cap, and some arrives late since it comes from the edge of the cone.

In short.... Just think of the response below 300hz a a perfectly straight line. You can splice them together if you want. I just use my imagination, and focus on 300hz and up.

[trevordj]

Suggestion.... Crop that measurement from about 80hz to 20k, and it will be easier to interpret.

Nearfield will do odd things up around 1k. The problem is that some of the sound is from the dust cap, and some arrives late since it comes from the edge of the cone.

In short.... Just think of the response below 300hz a a perfectly straight line.

Got ya, here you go:

woofer blended response:


woofer far:



Woofer near



Overlayed response

Tweeter:

[rpb]

I think that this is the only measurement that matters. You either sim, or connect parts by trial and error to shape the response into something like shown.



Then you do the same with the tweeter, and it may look something like this. This was measured indoors. The XO was done by trial and error. (No sims.)

[trevordj]

I think that this is the only measurement that matters. You either sim, or connect parts by trial and error to shape the response into something like shown.

Right on! I'm getting closer . I am going to see this side project through to a functional speaker and then I will get back to the main LCRs. I really appreciate all of your input, I'm learning a ton

[rpb]

Right on! I'm getting closer . I am going to see this side project through to a functional speaker and then I will get back to the main LCRs. I really appreciate all of your input, I'm learning a ton

I wish you had some parts. I've wanted to attempt a trial and error xo online. For this, I'd estimate a 2.5mH coil, and 12uf cap as a start, and see what it looks like.

Wait... Well DUH on me! You can make your frd, and sim the 2,5mH, and 12uf cap.

[trevordj]

I wish you had some parts. I've wanted to attempt a trial and error xo online. For this, I'd estimate a 2.5mH coil, and 12uf cap as a start, and see what it looks like.

Wait... Well DUH on me! You can make your frd, and sim the 2,5mH, and 12uf cap.

If I had the part I would definitely do it. I am going to simulate it for now and order some and we will see how it goes.

[rpb]

If I had the part I would definitely do it. I am going to simulate it for now and order some and we will see how it goes.

The biggest coil I've used on a 2-way is just under 3.6mH. I unwound a few turns. A 2.5mH often gets close. I depends on how you want to voice the speaker, as well as the drivers.

[rpb]

Speaker pron. Doesn't this look fun. Sweep, change a part, sweep, change a part.

[trevordj]

Speaker pron. Doesn't this look fun. Sweep, change a part, sweep, change a part.

lol, looks super fun.

Actually all of this fiddly stuff really is fun. I have been wanting to learn how to do this for years and it really is enjoyable. I think whatever speaker comes out of this "instructional offshoot" will have to be named "The Diversion" or something like that.

[trevordj]

Ok, I continue to make progress, and I find all of this super interesting so win-win. Here is where I am with the measurements:

tweeter on baffle in waveguide:



Woofer in sealed box:



Crossover sim:



Reverse null



The crossover is super simple and the response looks reasonable. I'm not too worried about that response above 10kHz as it attenuated off axis. The only thing I'm not sure about it the impedance in the upper registers dips to 2. Is this a concern? Unless anyone has any objections I may order these parts and see how it goes. I am still working on the ported boxes and, since this seems to be working out, I am going to actually put these speakers together and give them to my seven year old son. He has been helping me with these and is super into them. I'll have to put together a subwoofer for him as well .

[trevordj]

Here is the design for the ported boxes:





And here is the box sim:



With no HP filter these could exceed the xmax of the woofer at 50 watts




But with a 60Hz 12dB/octave HPF everything looks ok

[Wolf]

Your acoustic phase alignment with this xover is not so good.

Wolf

[rpb]

I can't see the part values. Can you post something much larger?

What I do see is a resistor in parallel with the tweeter, but not one in series. That is probably why the impedance is low. Try a resistor in series before or after the tweeter xo.

[rpb]

I think maybe the wave-guide is affecting the woofer response. You can compare factor data maybe and see, or place something over the wave-guide, and remeasure.