I do not posses tools to measure each driver impedance so Im only relaying on what manufactures provided
In room responses of different types.
Additionally here is decay and waterfall graph of my listening position.
I cannot translate ‘weird’ and ‘unresponsive’ to speaker characteristics so don’t know how to help. Your measurement skills should improve. Read up on that and learn to understand what the contribution is of the speaker and what of the room, is my advice.
Wow, there is a lot wrong starting from post #1. Also, I did not see you say how you obtained the measurements that you used to design the crossover in the first place. Crossover modeling software is only as good as the data you give it, so if your driver and impedance measurements are not done properly then you are shooting in the dark. Can you please give some more detail about how you made those measurements?
So let's go back to this plot from post #1:
This is not a good system response. The SPL for the system should be relatively flat. We see one level (~90dB) around 500Hz, another level between 900 and 2.3kHz (~87dB) and then above 2.3kHz yet another level (84dB). This speaker probably sounds muffled or dead but with some annoying upper vocal presence. Also the woofer response trends up from 100Hz to 600Hz. You want this flat or slightly tipped down. Also, some designers will make the woofer level below 600Hz a little LESS than above it. This depends a bit on how you obtained the frequency response data. Measurements should be done under quasi-anechoic conditions, which means you collect an impulse response, set a window (sometimes called a gate) within which there are no boundary (room walls, floor, etc.) and then use that data to obtain your frequency response.
You should be able to cross over around 2kHz or slightly lower. The woofer has a breakup peak at 3kHz that you want to suppress, and the lower you can cross over the better. This depends on how low the tweeter can play, but I am not familiar with it. So anywhere between 1.5k and 2k Hz is a good target.
When you make the crossover and want to re-measure, you will need to redo the quasi-anechoic measurements. Put the speaker on top of a ladder outdoors and remeasure at a distance that is about 3 times the baffle width or larger. In-room measurements are rather meaningless at this stage.
So let's go back to this plot from post #1:
This is not a good system response. The SPL for the system should be relatively flat. We see one level (~90dB) around 500Hz, another level between 900 and 2.3kHz (~87dB) and then above 2.3kHz yet another level (84dB). This speaker probably sounds muffled or dead but with some annoying upper vocal presence. Also the woofer response trends up from 100Hz to 600Hz. You want this flat or slightly tipped down. Also, some designers will make the woofer level below 600Hz a little LESS than above it. This depends a bit on how you obtained the frequency response data. Measurements should be done under quasi-anechoic conditions, which means you collect an impulse response, set a window (sometimes called a gate) within which there are no boundary (room walls, floor, etc.) and then use that data to obtain your frequency response.
You should be able to cross over around 2kHz or slightly lower. The woofer has a breakup peak at 3kHz that you want to suppress, and the lower you can cross over the better. This depends on how low the tweeter can play, but I am not familiar with it. So anywhere between 1.5k and 2k Hz is a good target.
When you make the crossover and want to re-measure, you will need to redo the quasi-anechoic measurements. Put the speaker on top of a ladder outdoors and remeasure at a distance that is about 3 times the baffle width or larger. In-room measurements are rather meaningless at this stage.
The in-room in post #42 doesn't look to bad actually. Slightly tilting downwards - good. Usual floor-reflex succ-out at 300. Some peaks in bass - the 135 is unfortunate. Open baffle gives a spacious and open sound. Boxes tend to sond .... boxy. I simply dont think you like boxed speakers. Go back to OB.
//
//
I believe this is a lower resistance driver, often to get the best result with this sort of driver requires a lot more attention to the enclosure beyond what is suggested through modeling (..baffle-step & lower freq. response).
Typically: a deep enclosure bass-reflex design with a larger volume than modeled along with a low freq. port tuning having a larger port diameter than normal. Interior cabinet walls should be free from fibrous damping (at least 2" from any surface) while the middle-interior of the cabinet should have some lite fibrous damping suspended (away from walls, driver, and port entrance) to absorb higher freq. internal reflections.
Try this out without any crossover. (..if you've got access to digital eq. (along with measurement) - eq'ing it "flat" at listening position, this can be beneficial to make sure that linearity isn't also affecting your judgment, though usually the lack of "high-end" sound in these instances is noticeable regardless of linearity.)
I really don't know what you, wadimek11, find difficult about this!
The waveguide tweeter:
https://sbacoustics.com/product/sb26stwgc-4-fabric/
To reiterate on the woofer:
https://www.daytonaudio.com/product/1415/es180tia-8-7-esoteric-series-woofer-8-ohm
Those graphs tell you almost everything you need to know about the FRD and ZMA files. A medium inductance woofer whose impedance correction should be around 9R plus 15uF.
Qts = 0.4, Vas = 20.7 litre indicates a reflex cabinet around 18l should work OK. Port around 5-6 cm diameter and about 15 cm length. And if you don't like reflex, you can stuff a rolled-up sock in the port.
It's the crossover you have made a complete hash of. Sorry, but it's the facts.
I have already given you two competent projects with this woofer which are driver specific:
https://projectgallery.parts-express.com/speaker-projects/premios/
https://projectgallery.parts-express.com/speaker-projects/clarino/
Why, even this SEAS Idunn project (an 8 ohm tweeter) would be a decent start point:
http://www.seas.no/index.php?option=com_content&view=article&id=380:idunn&catid=
Just adjust the tweeter attenuator to maintain the designed load resistance and get level right, and it will probably work well enough, though I would try a 5kHz tank notch on the woofer, which would be about 15R and 0.47uF across the bass coil.
Excellent project by Mr. Troels Gravesen here on what to do with typical 6" woofers. Very educational:
http://www.troelsgravesen.dk/DiscoveryW18.htm
The waveguide tweeter:
https://sbacoustics.com/product/sb26stwgc-4-fabric/
To reiterate on the woofer:
https://www.daytonaudio.com/product/1415/es180tia-8-7-esoteric-series-woofer-8-ohm
Those graphs tell you almost everything you need to know about the FRD and ZMA files. A medium inductance woofer whose impedance correction should be around 9R plus 15uF.
Qts = 0.4, Vas = 20.7 litre indicates a reflex cabinet around 18l should work OK. Port around 5-6 cm diameter and about 15 cm length. And if you don't like reflex, you can stuff a rolled-up sock in the port.
It's the crossover you have made a complete hash of. Sorry, but it's the facts.
I have already given you two competent projects with this woofer which are driver specific:
https://projectgallery.parts-express.com/speaker-projects/premios/
https://projectgallery.parts-express.com/speaker-projects/clarino/
Why, even this SEAS Idunn project (an 8 ohm tweeter) would be a decent start point:
http://www.seas.no/index.php?option=com_content&view=article&id=380:idunn&catid=
Just adjust the tweeter attenuator to maintain the designed load resistance and get level right, and it will probably work well enough, though I would try a 5kHz tank notch on the woofer, which would be about 15R and 0.47uF across the bass coil.
Excellent project by Mr. Troels Gravesen here on what to do with typical 6" woofers. Very educational:
http://www.troelsgravesen.dk/DiscoveryW18.htm
Last edited:
I simulated your project in VituixCad: the tweeter curves are from Dibirama, those of the woofer from the DA website. Both have been manipulated in Diffraction tool to simulate the response on the baffle.
Vb 18 L tuned at 35 Hz
Port > 50x220 mm
Baffle > 220x420 mm
W Height > 210 mm
W-T distance > 150 mm
Vb 18 L tuned at 35 Hz
Port > 50x220 mm
Baffle > 220x420 mm
W Height > 210 mm
W-T distance > 150 mm
I'm thinking the same thing with regard to you and others. 😉
When I read his first comment and other subsequent comments, sure - the design itself needs work (a lot), but it's not the principal reason he has posted this thread.
When someone say's it "sounds good open baffle and not in box - and not sounding hi-end", the principal problem is almost never about *linearity, rather it's about sound-staging effects. As in a subjective result that is "closed" "shut-in" "lacks depth" "lacks width beyond the loudspeakers" etc.. You can have an exceptionally linear result and still lack these effects (Revel has a number of less expensive speakers that epitomize this). In fact you can have this result without any crossover at all, just using the single driver (and with dsp equalized to linear on-listening-axis).
*imagine how linear it was/was-not open-baffle.
At this point you either fix this underlying problem FIRST (and only thereafter proceeding to the crossover design) or you do as motokok suggests and sell the drivers that are providing a poor result.
-of course I could be wrong
but that's the way I read his comments.
My thoughts are to remove the LR across the woofer. While it may help with the Fs spike, you are reducing sensitivity by quite a lot.
I think you are right. While it puzzles me how to get a dipole sounding right with the back 10cm from the front wall, but maybe that wasn't the setup of his open system. OP, any comments?
Sensitivity is quite low, 84.8 dB before baffle step according to Dayton Audio specs. I tried to remove the LR but you have a 5 dB peak at 1 KHz and sensitivity at 200 Hz was the same.
In any case, that was just a simulation to be used as a starting point as we do not have any measurements and do not know the baffle size.
The actual driver was 40-50cm from wall in open baffle maybe even more. The speakers are in the same spot in enclosure tho but enclosure adds lenghts. Literally this configuration sounded better by quite a big margine from what I have now. In both cases I use same crossover+dsp+sub. Theoretically they sound alright now and I asked even few people to listen to it but they could be beaten by market 1000usd speakers but they in open baffle that was temporarily made to test them was way way better sounding than any of my previous speakers.
Attachments
Get them out at least another 50cm and Open Baffle them, I'd say. Forget the CB or BR. Cross to sub at about 200Hz (use at least two subs spatially distributed). You could use Basta! for a reference to correct the response with dsp and measure for the finetuning. But apply a decent crossover from bass mid to high!
Your woofer is 84dB sensitive (more or less), so your system response should be around the ~80dB mark for your average speaker in your average listening room. i.e. allowing ~ 4dB bafflestep compensation.
The lead series inductor which most simple 2 way designs use to apply both bafflestep and LP filter is usually around 1.5mH to 3mHfor 8 ohm nominal woofers. Most designs are around the 2 - 2.5mH range.
given your woofer rises to 90dB ~ 700Hz - I suspect either you have an odd baffle shape or suspect your woofer farfield (gated) measurement is invalid.
Shadowplay's crossover above looks far more reasonable, with a larger value woofer inductor applying sufficient bafflestep. Phase alignment also looks better, as well as overall system (minimum) impedance.
The lead series inductor which most simple 2 way designs use to apply both bafflestep and LP filter is usually around 1.5mH to 3mHfor 8 ohm nominal woofers. Most designs are around the 2 - 2.5mH range.
given your woofer rises to 90dB ~ 700Hz - I suspect either you have an odd baffle shape or suspect your woofer farfield (gated) measurement is invalid.
Shadowplay's crossover above looks far more reasonable, with a larger value woofer inductor applying sufficient bafflestep. Phase alignment also looks better, as well as overall system (minimum) impedance.
I commented earlier on the quality of the measurements. If and only if this above says anything, the mean level between 200 and 600Hz is too low. Combined with a rather high level at 700-1200Hz, the tonal balance is way off. And that complies somehow a bit with OP's complaints about the sound. Furthermore I'm clueless how you would design a crossover without having the proper impedance measurements (be it, old school trial and error certainly is an option). So I repeat: get your measurements done right.
Impedance measurements are far less important than frequency response measurements. If the drivers have build consistency from the manufacturer then a sealed dome tweeter will have no difference and enclosure loading on a woofer will likewise not affect the impedance range near your usual 2 way crossover point ~2KHz
True, but measuring impedance is really simple compared to measuring the acoustic output. Why wouldn't one?
It depends on hardware available. If the OP is using a laptop with only single 3.5mm dual purpose headphone / line in jack (not possible) and no DATS or other alternative.