The Paulinator said:
Try Behringer DCX 24/96 digital active crossover, it will give you a lots of possibilities. There are quite a few treads on this forum about it.
Peter, your attention to details and craftsmenship is always inspirating. I had a chance to hear some of the Triangle speakers and they do have an amazing imaging.
phase_accurate said:
As I said in the other thread, those are the best sounding speakers I have at the moment. Much of it is probably due to a very simple crossover, consisting of just one series cap (Siemens MKV) with a tweeter. I might be also upgrading to Skan Speak 9700 drivers. Although the bass is not comparable to what I get with 15 and 12" in parallel in my main speakers, it sounds pretty fast and real and I like a lot.
Peter,
Congratulations on completing a project that is very easy on the eyes. I have also viewed one of your amp projects and I applaud your craftsmanship. You set a very high standard for the rest of us.
I am replying today because I am rather puzzled by the woofer frequency response graph and the schematic you posted. The two don't really go together, particularly that notch filter. I have always thought that center frequency of notch filters was easy to calculate.
In Excel use the following formula to determine center frequency of passive notch filters.
=(1/(2*PI()))*SQRT((1/(X*Y)))
replace X and Y with the cells containing the value of the capacitor and inductor. Capacitor values are in scientific notation for microFarads and inductor values are in scientific notation for milliHenries.
I apologize for the inelegance of the formula. I used to use one that looked much nicer, but with the last Microsoft upgrade it stopped working. Using Excel’s built in formula builder produced the above. I believe it works, what more can I say. I gave up long ago trying to figure out Microsoft upgrades and have learned to just go with the flow.
The values in the schematic yield a center frequency just below 5 kHz (4822.877Hz). The 2.2 ohm resistor degrading the quality of the series notch filter is quite small in comparison to the source resistance of the inductor about 2.5 to 3 octaves above its knee point. Even with the resistor this notch filter will have a fairly deep notch.
Now the puzzling part. In the frequency response supplied, there is nothing shown in the 5 kHz range that you would want to notch out.
Three possible hypotheses:
One, despite several decades of calculating filters and then verifying them experimentally, I have miscalculated the center frequency.
Two, the frequency response graph is not accurately showing the cone breakup response.
Three, the schematic supplied is irrational.
If hypothesis one is correct, then I have just wasted server space and bandwidth. If two is correct, then I am really happy that you are pleased with the sound of your speakers. If three is correct, then the driver manufacturer is seriously confused.
Again, thanks for sharing your project,
Mark
Congratulations on completing a project that is very easy on the eyes. I have also viewed one of your amp projects and I applaud your craftsmanship. You set a very high standard for the rest of us.
I am replying today because I am rather puzzled by the woofer frequency response graph and the schematic you posted. The two don't really go together, particularly that notch filter. I have always thought that center frequency of notch filters was easy to calculate.
In Excel use the following formula to determine center frequency of passive notch filters.
=(1/(2*PI()))*SQRT((1/(X*Y)))
replace X and Y with the cells containing the value of the capacitor and inductor. Capacitor values are in scientific notation for microFarads and inductor values are in scientific notation for milliHenries.
I apologize for the inelegance of the formula. I used to use one that looked much nicer, but with the last Microsoft upgrade it stopped working. Using Excel’s built in formula builder produced the above. I believe it works, what more can I say. I gave up long ago trying to figure out Microsoft upgrades and have learned to just go with the flow.
The values in the schematic yield a center frequency just below 5 kHz (4822.877Hz). The 2.2 ohm resistor degrading the quality of the series notch filter is quite small in comparison to the source resistance of the inductor about 2.5 to 3 octaves above its knee point. Even with the resistor this notch filter will have a fairly deep notch.
Now the puzzling part. In the frequency response supplied, there is nothing shown in the 5 kHz range that you would want to notch out.
Three possible hypotheses:
One, despite several decades of calculating filters and then verifying them experimentally, I have miscalculated the center frequency.
Two, the frequency response graph is not accurately showing the cone breakup response.
Three, the schematic supplied is irrational.
If hypothesis one is correct, then I have just wasted server space and bandwidth. If two is correct, then I am really happy that you are pleased with the sound of your speakers. If three is correct, then the driver manufacturer is seriously confused.
Again, thanks for sharing your project,
Mark
Nice work again Peter.
I wonder that where did you get all those tools and stuff to build these DIY things???
Did you have real daywork at all or where did you get time to do this nice audio stuff?
Keep going Peter....
PS. I can only dream that someday I have enough time and tools to bring my ideas alive
I wonder that where did you get all those tools and stuff to build these DIY things???
Did you have real daywork at all or where did you get time to do this nice audio stuff?
Keep going Peter....
PS. I can only dream that someday I have enough time and tools to bring my ideas alive
MarkMcK said:
Hypothesis one is partly correct - you have miscalculated the centre frequency because the formula you are using doesn’t take into account the woofer impedance at the target frequency. All woofers have raising impedance at higher frequencies due to voice coil inductance. Applying a textbook formula in this situation leads to missing the target. I use simulation software in these cases.
Unfortunately I cannot point you to more sophisticated formula right now either but I am sure somebody here can help us out.
Argo
Peter:
Have you ever considered a series filter?? They work really good in my oppinion.
If you are testing filters try it out. not many active components and really good sounding.
http://www.acoustic-reality.com/ar-sxo.htm
Check this danish site out ...they have a different aproach to series filters. Gonna go for that next time i build new speakers.
Would you mind terribly of i steal your box design
Well actually i guess that JM-LAB was first. But I like the combination of color and wood finish.
Always inspiring
Have you ever considered a series filter?? They work really good in my oppinion.
If you are testing filters try it out. not many active components and really good sounding.
http://www.acoustic-reality.com/ar-sxo.htm
Check this danish site out ...they have a different aproach to series filters. Gonna go for that next time i build new speakers.
Would you mind terribly of i steal your box design
Well actually i guess that JM-LAB was first. But I like the combination of color and wood finish.
Always inspiring
Here's the Raven setup http://www.diyaudio.com/forums/showthread.php?s=&postid=225064#post225064
It is promising, but I need to bi amp that system as the crossovers used here (again, a single cap in a tweeter) don't work well yet. With active crossover I hope for 2 - 3kHz, maximum third order filter. Also it seems like in wall speakers don't image that well and acrylic baffles, in spite what I expected, introduce too much of its own signature. But for movies it is fine as it is
As to the notch filter, I don't have a clue about it, that's why I didn't use it
And indeed to do all that stuff one cannot have a regular job, so that's why I don't have one.
I also like the combination of colors and don't mind if someone copies it. Initially I didn't plan for black surround on the tweeter, but it looked strange (with tweeter that low) and addin neoprene padding improved not only the visual appeal but also acts as decoupling tweeter from the box (the strenght of tightening the screws has an influence here) and helps with reducing some of the reflections. I didn't want to make the whole box in maple as this would look boring (like some of Red Rose Music speakers)
It is promising, but I need to bi amp that system as the crossovers used here (again, a single cap in a tweeter) don't work well yet. With active crossover I hope for 2 - 3kHz, maximum third order filter. Also it seems like in wall speakers don't image that well and acrylic baffles, in spite what I expected, introduce too much of its own signature. But for movies it is fine as it is
As to the notch filter, I don't have a clue about it, that's why I didn't use it
And indeed to do all that stuff one cannot have a regular job, so that's why I don't have one.
I also like the combination of colors and don't mind if someone copies it. Initially I didn't plan for black surround on the tweeter, but it looked strange (with tweeter that low) and addin neoprene padding improved not only the visual appeal but also acts as decoupling tweeter from the box (the strenght of tightening the screws has an influence here) and helps with reducing some of the reflections. I didn't want to make the whole box in maple as this would look boring (like some of Red Rose Music speakers
)- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.