I have 3 way DIY speakers. Midrange is Accuton 173-6-090, Scan Speak Revelator tweeter, ring radiator. The crossover is 6 dB per octave. The midrange is 120uF and 0.56 mH. The tweeter is 3.9 uF. Distortions, sibilants are heard in the area of high tones. I tried other speakers in the setup, there were no distortions or sibilants. I read about how the SS ring radiator can’t play low. I have experience with the Vifa ring radiator, which also couldn’t play low without distortion. I would like to keep the crossover 6 dB per octave, because I have expensive elements, and I also like the connection and phase correctness of such a crossover. I wonder if you know which tweeter could play with my Accuton midrange and 6 dB per octave crossover? Best from own experience with similar drivers.
You are crossing over at 5khz assuming 8 ohm tweeters which is by no means low. If it is 4 ohm unit you are at 2500hz which is getting low for a ring radiator with a 6db roll off. Go 18db and invert the phase on the tweeter to get it back in phase with the 6db woofer crossover. If you are worried about phase shift over the frequency range use a Bessel type crossover. It is more phase accurate than Butterworth but less steep of a roll off which is ok.
if you are hearing sibilance at low volumes something else is wrong.
if you are hearing sibilance at low volumes something else is wrong.
An electrical first order filter (which you have) very rarely (almost never) gives acoustically 6dB/octave Butterworth roll-off (which is necessary for linear phase response) with real-world drivers. So you probably don't have linear phase either with your speakers.
Usually it takes a very complex filtering to get a proper, acoustically 6dB/octave crossover.
I think an acoustically steeper crossover (a well executed, not just with an online calculator) may would help. At least, may worth a try to reduce the sibilance.
If you don't want other crossover than electrically first order, then I would measure what's happening acoustically with the current setup.
Acoustical measurements is advisable anyway.
Usually it takes a very complex filtering to get a proper, acoustically 6dB/octave crossover.
I think an acoustically steeper crossover (a well executed, not just with an online calculator) may would help. At least, may worth a try to reduce the sibilance.
If you don't want other crossover than electrically first order, then I would measure what's happening acoustically with the current setup.
Acoustical measurements is advisable anyway.
The main breakup of the Accuton 173-6-090 is around 5-6kHz, right in the middle of the sibilance area. The breakup seems well-damped in the frequency domain but can be problematic in the time domain because it's still have a sharp roll-off and it's still a resonance and maybe the electrically first order filter is not enough to attenuate this range adequately.
Those drivers can't work together with a simple 1st order electrical crossover. The breakup and dispersion characteristics of the Accuton will require a 2KHz more or less LR4 crossover point and probably also a notch for the breakup. A LR4 2KHz crossover point should be doable also for the SS tweeter, even if it is on the low side. The Accuton really can't work with a simple coil as a LP filter, the distortion you are hearing is probably from this driver.
The point you are mainly missing here is that what is important is the acoustic slope, which is a sum of the electrical slope and the inherent FR of the driver. This means that in almost all cases a simple 1st order electrical filter yields an acoustical slope greater than 1st order. So you won't get the phase behaviour you expect. All real 1st order acoustic crossover require much more components than a simple cap or coil.
One last point: a couple of your mids and tweeters cost in excess of 2000 €, and we don't even have factored in the woofers. With a fraction of this cost you can get a good calibrated non-usb mic and an appropriate mic preamp, so you can measure the drivers and design a tailored crossover.
Ralf
Just to clarify. I bought ready-made diy speakers. But I will obviously have to become a diy constructor 😂
BTW. the basses are Accuton 173-11-191, 3 pieces in a box.
BTW. the basses are Accuton 173-11-191, 3 pieces in a box.
Okay, as you can see, expensive drivers slapped in a box with a questionable crossover (passive 1st order electrical in this case, which is one of the worst imo) is not necessarily produces satisfying sound. 😀
You can use USB measurement mic too (Umik-1 for example), but to design a passive crossover, you also need to measure the impedance curves (because they are not straight lines or fixed values) of the drivers.
You can use USB measurement mic too (Umik-1 for example), but to design a passive crossover, you also need to measure the impedance curves (because they are not straight lines or fixed values) of the drivers.
Last edited:
So: don’t pick the Umik or any other USB mic, but go for the dual channel USB audio interface and ’ordinary’ measurement mic. A Behringer UMC22 and ECM8000 plus stand and cables and a few resistors for an impedance measurement rig will set you back about €100 and give you a good set for developing crossovers, assuming you have a PC and and amp.
PS, it’s jawdropping time and time again how diy’ers manage to finish ’suboptimal’ designs and sell them too. 🤣
Thanks guys. Which crossover with these drivers would be the most optimal? 2nd order, or 3rd order? What type of crossover? Btw. I have Zero Ohm bass coils, 5.6 mH and 3.9 mH. Can I use at least 2?
Are you willing to go for the development of a new crossover? If you can deliver the right measurements, sure folks here will be able to help you out. But these measurements are obligatory.
