The Joudas project

[trevordj]

It helps me to see the raw woofer response before any xo filters have been added. That way, you see potential issues that the x-over may need to deal with.

Here is the frequency response for both the woofer and tweeter. I messed up something along the way with my original simulation so I am back to square one. I will see what I can come up with and report back.

I was able to get simulated polar responses with baffle diffraction simulation for both woofer and tweeter. I am unable to figure out how to combine the woofers simulated free air FR with the box simulation, I think that is where I made an error before. It looks like it should be doable in vituixcad but I cannot figure it out. I think it should be fine for simulating the crossover given the crossover frequency is several about 4 octaves above the box tuning. If anyone knows how to make this work I would love to know though. I also tried using Jeff Baby's Response modeler but I can't get that to work either, I keep getting errors.

Back to the drawing board with the simulated crossover.

[trevordj]

Ok, lets try again. The crossover was much, much easier to make work this time.

Circuit



FR



Impedance



Phase



Reverse Null

[trevordj]

Directivity (Not sure how accurate this is given the tweeter is horn loaded)



Transfer function



Power dissipation of resistors. Unlike my previous crossover attempt, with flat signal selected, the resistors would get several hundred watts of power dissipation at 20khz. With M-noise the resistors are only seeing 0.1 watts of power.

[Chris Roemer]

I've traced your FR plots, but I'd need your Z(imp) plots to get an accurate sim.
I don't get your filtered response for your woofer (using a simulated Z-plot from WinISD), but I DO match your filtered horn pretty well (assuming a "flat" Z-curve for the AMT device @ 8ohms).

[trevordj]

I've traced your FR plots, but I'd need your Z(imp) plots to get an accurate sim.
I don't get your filtered response for your woofer (using a simulated Z-plot from WinISD), but I DO match your filtered horn pretty well (assuming a "flat" Z-curve for the AMT device @ 8ohms).

Thanks Chris, I'll be curious to see what you come up with. I have attached the frd and zma files I am using. The zma is measured free air for both the woofer and the tweeter. The box sim I posted above wasn't accurate, the box is 93L and tuned to 41hz with two 3" x 5" long ports, double flaired. The baffle is 20" x 30" with a 3/4' chamfer. The frd files are traced from drivervault. I'll attach the zma and DATs files in the next post since I can only do 5 attachments per post.

[trevordj]

Ok, well I guess the forum doesn't like the DATS files. Here are the zma files though.

[trevordj]

I have reworked things a bit, I think I got better phase alignment with a lower crossover point. This sim was done with the polar responses at 0, 20, and 40 degrees from drivervault rather than the simulated polar responses from vituixcad.

[Chris Roemer]

Immediately after my last post (#19) I tried to mimic your filters w/out going 4th order (electrical).
More often than not (on here) guys will target 4th order slopes (but that's "acoustical"), which are often achievable using just 2nd order (occasionally 3rd) electrical. The "extra" rolloff comes from the drivers' inherent rolloff.

Try this - HP (horn) 2nd order w/L-pad: I just used a 6uF series cap (left/amp end), then a 0.80mH shunt coil (to gnd), and lastly (next to tweeter) an "L" w/ SR=4n(ohms) and PR(parallel resistor)=10n.
The woofer is just 2nd order (elec): using a 1.5mH series coil (prob. iron core - or "low" DCR) and a 20uF shunt cap.

I'll probably run one more using your latest F/Z downloads.

[Chris Roemer]

Not sure what's going on, but I can NOT get your woofer rolloff to look like you show it using your F/Z files (and I've done about a million sims).
(I've just noticed guys using Virtuix Cad (or whatever it's called) rather lately on here - like THIS year. Most have been using a handful of softwares (like for 15 yrs or more) that are either the late Jeff Bagby's "PCD" (Passive Crossover Designer - one of the oldest) or one of the several "workalikes" that others have developed. Makes me wonder if VC has some inherent problems/omissions in its programming?)

[trevordj]

Not sure what's going on, but I can NOT get your woofer rolloff to look like you show it using your F/Z files (and I've done about a million sims).
(I've just noticed guys using Virtuix Cad (or whatever it's called) rather lately on here - like THIS year. Most have been using a handful of softwares (like for 15 yrs or more) that are either the late Jeff Bagby's "PCD" (Passive Crossover Designer - one of the oldest) or one of the several "workalikes" that others have developed. Makes me wonder if VC has some inherent problems/omissions in its programming?)

Weird, I will try working it in x-sim, although I do not see a way to add polar measurements in that program. I have been having some issues with Jeff Bagby's excel files on my computer lately. The only one reliably working for me is the Diffraction and Boundry Simulator and Woofer Box and Circuit Designer. I dedicated myself to learning Vituixcad because it seemed to be the most comprehensive so of course its the one that there have been problems with. It is entirely possible I am making an error somewhere. The frd I am using in my crossover sim does have baffle diffraction simulation applied so maybe thats it?

[trevordj]

Not sure what's going on, but I can NOT get your woofer rolloff to look like you show it using your F/Z files (and I've done about a million sims).
(I've just noticed guys using Virtuix Cad (or whatever it's called) rather lately on here - like THIS year. Most have been using a handful of softwares (like for 15 yrs or more) that are either the late Jeff Bagby's "PCD" (Passive Crossover Designer - one of the oldest) or one of the several "workalikes" that others have developed. Makes me wonder if VC has some inherent problems/omissions in its programming?)

Strange, I did just the raw woofer response with baffle diffraction applied in vituixcad using the frd I posed above



And I imported the frd above into x-sim and applied baffle diffraction which I took from Jeff's Diffraction and Boundry Simulator and I got a very similar response



I may be doing something consistently wrong? I appreciate your input as this is why I am cranking through with simulated measurements before I complicate things with another variable.

[trevordj]

Immediately after my last post (#19) I tried to mimic your filters w/out going 4th order (electrical).
More often than not (on here) guys will target 4th order slopes (but that's "acoustical"), which are often achievable using just 2nd order (occasionally 3rd) electrical. The "extra" rolloff comes from the drivers' inherent rolloff.

Try this - HP (horn) 2nd order w/L-pad: I just used a 6uF series cap (left/amp end), then a 0.80mH shunt coil (to gnd), and lastly (next to tweeter) an "L" w/ SR=4n(ohms) and PR(parallel resistor)=10n.
The woofer is just 2nd order (elec): using a 1.5mH series coil (prob. iron core - or "low" DCR) and a 20uF shunt cap.

I'll probably run one more using your latest F/Z downloads.

I tried these values with the following outcome:

[djg]

I know you've waded into the crossovers, but I wanted to say this is some very fine paintwork.

[trevordj]

Not sure what's going on, but I can NOT get your woofer rolloff to look like you show it using your F/Z files (and I've done about a million sims).
(I've just noticed guys using Virtuix Cad (or whatever it's called) rather lately on here - like THIS year. Most have been using a handful of softwares (like for 15 yrs or more) that are either the late Jeff Bagby's "PCD" (Passive Crossover Designer - one of the oldest) or one of the several "workalikes" that others have developed. Makes me wonder if VC has some inherent problems/omissions in its programming?)

Sorry for the hammer quotes, I'm trying to get to the bottom of this. I redid the trace and got the same frd



Here is how I am simulating baffle diffraction.



Your proposed crossover looks great without baffle diffraction simulated




After baffle diffraction frd



And your proposed crossover after baffle diffraction simulated

[trevordj]

I know you've waded into the crossovers, but I wanted to say this is some very fine paintwork.

Thanks a million! I wish paint/bodywork didn't take so much time and effort but there is nothing better than getting a nice finish that will last a lifetime. Up until about 1.5 years ago spraying automotive paint with an hvlp gun was completely new to me. It just took a lot of experimentation on scrap wood and getting the right gear.