Morel CAW538 crossover help

[BWnAZ]

I have been experimenting with the Morel CAW538 in a 0.25 cu. ft. cabinet (8" wide) tuned to 53Hz for two different projects. One with the Morel CAT308 tweeter and the other with the Fountek NeoX 3.0 tweeter. The results have been very good with a textbook 8 ohm 2.1 KHz Linkwitz/Riley 2nd order crossover. I have also been using an 8ohm variable L-pad (potentiometer) for tweeter attenuation. I have also been using a textbook Zobel network of an 11 microfarad capacitor with a 6.5 ohm resistor. I have limited the BSC to -3db as greater attenuation seems to make the CAW538 dull and lifeless.
The results have been excellent but I would like to try a custom crossovers for both projects before beginning the build process. I would also like to convert the variable L-pad to a fixed voltage divider, however, I do not want to over attenuate the tweeter(s) and the impact of the added resistance on the crossover is very difficult to calculate. My Mac is too old (for Boot Camp etc.) to run windows and it seems like most crossover design software is developed for Windows only. Would anyone be willing to model a crossover for either or both of these projects for me? Thank you in advance for your help!

[rpb]

Can you measure the responses?

Do you have a good assortment of xo parts?

What cap, and coil values are being used on the woofer, and tweeter currently?

[BWnAZ]

Sorry, I do not have the proper equipment to measure the responses. I was hoping some of the available crossover software programs might have that data.
I do have a good assortment of crossover parts. I would buy whatever I needed if someone provided a schematic for a custom crossover for either project. I believe these drivers are worth the investment.
I am using 1.2mH coils and 4.7 microfarad caps(Audyn cap plus) on the 2nd order L/R crossover. This calculates out to 2115Hz @ 8ohms. I am using a 0.6 mH coil in parallel with a 2 ohm Mills resistor for the BSC which was originally calculated for a 7.5" wide box. Thanks for your inquiry. I look forward to any advice you could share.

[Chris Roemer]

Your box looks good, and (oddly enough) your "textbook" filter seems to be working well (along w/your Zobel).
For a 2.1 Fc w/+3dBSC, I don't think I'd change a thing.

Which tweeter are you going with?
You should be able to measure the impedance (resistance) on the legs of your L-pad to see what fixed values you could replace it with.

[BWnAZ]

I'm obsessed and proud of it too. Thanks for your words of encouragement. Everyone says you absolutely must build a custom crossover, but without crossover modeling software, I can only go by ear.

I'm actually going to use both tweeters with this woofer. I have 4 Morel CAT308's that I will use with the CAW538's and will likely give away. I have become obsessed with the Fountek NeoX 3.0's. I am using those in the front and center channels for my home theater system. I intend to use them with the CAW538's for the rear surrounds. I will probably also build a pair of the the 538/NeoX's for my music room as I think they will work well with my trusty old 12" Titanic Mk2 sub.

You're right, tweak the tweeter pot. until you like the way it sounds and then take the measurements. I really need to get a higher quality ohmmeter... Any advice?

Have you ever used the Jantzen silver-Z caps? I'd like to know if they are worth the money. I don't want to ruin good drivers with cheap crossovers. Thanks.

[Chris Roemer]

You can view the PDF files (even on a Mac) to see how much flatter that dome is vs. the ribbon.

You can incorporate a 2nd order rolloff on the woofer (incorporating about +3dBSC and the effects of your Zobel) by just using a 2.0mH series coil (low DCR, so "iron core") and an 8uF shunt cap (to ground). Fc near 2k. ALL are SOLID 8ohm designs w/Z-min near 240Hz @ 7.4 ohms or so.

Then for the dome: 7uF series cap & 0.85mH shunt coil (#18 or #20 is fine here) and a fixed L-pad: SeriesResistor = 5n(ohms), Parallel = 6n. Normal polarity is indicated (but try it both ways and go with whatever sounds best). (Runs near 82dB +/-1dB.)

For the ribbon (have you looked at its FR curve? - it's got a peaky top end that I'm rollling off a bit): small 0.10mH series coil (for that top end) and a 4uF series cap, followed by an 0.80mH shunt coil and the "L" : SR=12n / PR=4n. Reverse (negative) polarity is indicated. (Runs near 83dB +/-3dB.)

[BWnAZ]

Thank you so much for your crossover solution. I have already ordered the parts that I don't currently have. I will post my progress as soon as I put them together.

I truly appreciate your explanations of each of the filters. I am particularly impressed with the way you incorporated the BSC and Zobel into a simple 2.0 mH/8uF 2nd order low pass filter. What a brilliantly simple solution. Way better than my Rube Goldberg method of just adding correction circuits to the filter. I also want to thank you for the tip on the "iron core" coil. I can see where the relatively high DCR (0.7ohms) of a 2.0mH air core coil could be problematic. I always assume that air core is better but this may be a case of the enemy of good is better. Lesson learned.

Thank you for the heads up on the dome polarity. I would have assumed that it should be reversed. I will try it both ways.

You are absolutely right about the ribbon's response curve and thank you for the correction. Ironically, the ribbon's response curve is peaky with a dead flat impedance curve and the dome is exactly the opposite. I like them both for different reasons. The dome is very smooth and seems to blend beautifully with this woofer. The ribbon has an alluring transparency with a very crisp high end but I realize it may be a poor choice with this inefficient little woofer and I may just stick with the dome. I do, however, enjoy the experimentation part of this hobby.

[BWnAZ]

Many thanks to Chris Roemer for your help! The simple 2.0mH coil and 8uF cap worked perfectly for the woofer. You were absolutely right about the iron core coil versus the air core. I tried both and the difference with the iron core was amazing.

The high pass for the dome worked well, however, I felt it sounded a little too flat and reverted back to the variable L-pad to brighten it up a bit. Normal polarity was correct for the dome.

The real magic happened with the ribbon tweeter. Again, the attenuation seemed a bit too much for my taste and I reverted back to the variable L-pad. I also wound up removing the 0.10 mH coil. I repeatedly tried it with and without it and decided I prefer the peaky top end. I'm sure my ears are not perfect transducers so perfectly flat may not work for me.

I have decided to stick with the ribbon tweeter for this design. The dome is very nice but the ribbon is amazing with your crossover design. This is a huge difference from the crossover I was using. When the tweeter is attenuated just right this little speaker sounds fantastic! The bass is so strong and punchy they sound like much bigger speakers. The mids are clean, clear and uncolored. The highs are crisp and clear with amazing transparency. This crossover makes it all blend perfectly. Thanks again for your help Chris. I am really thrilled with the results. I bought some quilted maple and can't wait to begin the build. I will post when complete.

Final Design:
Morel CAW538, 0.25cu.ft. cabinet tuned to 53Hz (1.5"ABS pipe x 7.35"), crossover: 2.0mH coil, 8uF shunt cap (to ground)
Fountek NeoX 3.0, crossover: 4uF cap, 0.80mH shunt coil (to ground), 8 ohm variable L-pad (potentiometer), tweeter wired in reverse polarity

[Wolf]

So measure the Lpad and use those values.
Wolf

[BWnAZ]

So I finally broke down and bought a DMM of adequate quality to measure the L-pad values. The series value came out to 5 ohms and the parallel resistance worked out to 15 ohms. I verified this by installing the resistors and it sounds great.
I also tried some different capacitors and decided that the Jantzen Silver Z-Caps are worth the money. The Dayton1% Caps are great for the money but the Jantzen's really open up the high end.