That was in the original post.
The time of flight distance between the woofers and tweeters still varies with distance as described.
Yes, a larger distance will reduce the path length difference.
What is the vertical scale of your measurement?
What is the center to center distance between the woofers?
What was the measurement distance?
Moving this a bit forward.
I had to settle for a woofer-filter based on the components I had in the drawer. Design-goal is around 400-500W @ 4Ohm so current and voltage will be pretty high. The only high-voltage film caps I had was 30uF/600V... Im a bit worried about the potential loss in the RC across the woofer. Right now its a 20W resistor. Will maybe have to simulate (or measure) temperature.
The inductor in my horn-filter was not the value I remembered, and that probably messed with my previous tests. My first attempts had a 5Ohm resistor before the horn-filter (damping, - part of the commercial filter I based it on). I could not get acceptable performance with this resistor before the filter and tried to move it to after the filter. This will change the horn-impedance as seen from the filter, but I will try to compensate for that. I have a small cap aross the resistor to lift the very high end a little.
Ii had originally aimed for a crossover at 2500Hz to protect the horn as much as possible. Its quite a bit lower now, but on the other hand it dampens the woofer brakeup more.
So this is current versions is like this:
Overall response is green-ish and looks very OKk with me. The dip after 10Khz is odd, but it dosnt bother me much.
Response dosnt change drastically with distance varying around 100-150cm indoors.
Filter-schematic and more measurments to follow.
Questions:
I had to settle for a woofer-filter based on the components I had in the drawer. Design-goal is around 400-500W @ 4Ohm so current and voltage will be pretty high. The only high-voltage film caps I had was 30uF/600V... Im a bit worried about the potential loss in the RC across the woofer. Right now its a 20W resistor. Will maybe have to simulate (or measure) temperature.
The inductor in my horn-filter was not the value I remembered, and that probably messed with my previous tests. My first attempts had a 5Ohm resistor before the horn-filter (damping, - part of the commercial filter I based it on). I could not get acceptable performance with this resistor before the filter and tried to move it to after the filter. This will change the horn-impedance as seen from the filter, but I will try to compensate for that. I have a small cap aross the resistor to lift the very high end a little.
Ii had originally aimed for a crossover at 2500Hz to protect the horn as much as possible. Its quite a bit lower now, but on the other hand it dampens the woofer brakeup more.
So this is current versions is like this:
Overall response is green-ish and looks very OKk with me. The dip after 10Khz is odd, but it dosnt bother me much.
Response dosnt change drastically with distance varying around 100-150cm indoors.
Filter-schematic and more measurments to follow.
Questions:
- What do you think? Better than first try?
- Is the distance from resonance to actual cross-over big enough?
Filter looks like this. Dosnt simulate very good, but the horn in it selvf is not very linear, so maybe thats to be expected.
I dont know if the correct approach, but I tried to simulate with 2 different sources to be able to delay the horn-signal according to the physical distance.
I dont know if the correct approach, but I tried to simulate with 2 different sources to be able to delay the horn-signal according to the physical distance.
Hmm. potential problem. Overall impedance looks strange.
I dont suppose that the impedance-dip at 2500Hz is good?
The current crossover measures good for response, but it appears that my trial-and-error approach has resulted in a silly impedance.
Any Ideas? - how dangerous would that impedance be to a PA amp?
I dont suppose that the impedance-dip at 2500Hz is good?
The current crossover measures good for response, but it appears that my trial-and-error approach has resulted in a silly impedance.
Any Ideas? - how dangerous would that impedance be to a PA amp?
Attachments
My hypotenuse length math was off by an order of magnitude, 1m to 1.5m is not going to present a phase shift problem, the time difference only 10-15mm (not 100 150mm..)
Looks better. The acoustic crossover is ~1800Hz, but over 24/dB/oct.
The rise in the HF horn response below the 900Hz bump may be a concern.
Is it real, or something in the noise floor?
It would be helpful to see the horn and woofer in cabinet raw frequency and impedance response.
No, it's not.
Yeah.
Probably more dangerous to the crossover components are dissipating all that heat.
What is the cause and will you fix it or are you asking assistance? We'd need to see more details. Perhaps individual filter responses and impedances?
Unless they are just reactive.
AllenB: havnt had much time to look into it, but I think it's caused by the series LC of C6 and L5. A series LC will have an impedance-dip and for some reason this dip becomes deeper when the first capacitor in the 3rd order filter is higher than the second. (Or at least that's my findings from playing around in the Weekend). I thinks it's purely reactive, but maybe a hard load for the amplifier?
The current version of the filter is based on trial and error and apparently I ended up with something that gives a ok frequency response, but with troublesome impedance. I have googled a little and its not hard to find other speakers designs with similar impedance-dips where the highpass-filter looks like a series LC.
I can do more graphs of the individual components.
Goal: get to good-enough and learn a little along the way. This is my first passive filter for high power.
The current version of the filter is based on trial and error and apparently I ended up with something that gives a ok frequency response, but with troublesome impedance. I have googled a little and its not hard to find other speakers designs with similar impedance-dips where the highpass-filter looks like a series LC.
I can do more graphs of the individual components.
Goal: get to good-enough and learn a little along the way. This is my first passive filter for high power.
Individual drivers. Horn protected with a 10uF in series. Hope thats OK?
Indoors. 1meter to centerdome on woofer. Mic at horn level.
Combined response (drivers same phase, still 10uF on horn
Indoors. 1meter to centerdome on woofer. Mic at horn level.
Combined response (drivers same phase, still 10uF on horn
Alternative version of filter. Without going into details: is it safe to assume that phasing of drivers is pretty good when phase is reversed and response shows a deep dip?
Graps shows new filter and same filter with horn inverted. Response is not good, but I have the option of "cheating", calling it good enough and adjusting response with EQ...
Individual drivers, same setup:
Graps shows new filter and same filter with horn inverted. Response is not good, but I have the option of "cheating", calling it good enough and adjusting response with EQ...
Individual drivers, same setup:
Small modification to filter. Less dip when phase is reversed, but slightly better response?
EDIT: thanks for al the feedback. Really appriciate it.
EDIT: thanks for al the feedback. Really appriciate it.
Total impedance for last version. Still low in high freq, but smaller Q (added series resistance to coil in horn-filter). Maybe easier load for amp?
Have you looked at notch filters? Say centered a 6K of so. Help notch the peak out.
Rob 🙂
https://www.diyaudioandvideo.com/Calculator/ParallelNotchFilter/Help/
Rob 🙂
https://www.diyaudioandvideo.com/Calculator/ParallelNotchFilter/Help/
Last edited:
No. The slopes of phase are clearly not at the same rate. Can you just use phase plots?