Playing around with mesaurements on my latest 3 way diy before i will do the last finetuning, and the wiggle bothers me.
Do you experienced people have some ideas about getting i better?
Mesaurement artifacts?
the mic is 120 cm away center midrange.
A SS 18M4631T00 doing 150-2,4 K in 4 liter box.
Regards John
Do you experienced people have some ideas about getting i better?
Mesaurement artifacts?
the mic is 120 cm away center midrange.
A SS 18M4631T00 doing 150-2,4 K in 4 liter box.
Regards John
Attachments
Can you measure outside? I see a lot of hard edges that will prevent a clean measurement. If you can't go outside at least move the DUT out into the middle of the room and and place it onto some kind of a stand to minimize the floor bounce. Also you need to gate it down to single milliseconds.
Get the gating right. And get the speakers outside, or anywhere you can apply a gate of 8 to 10ms before first reflection. Mostly by lifting speakers up in the air for about 2m measurement axis. Or measure using ground plane technique.
Reflection, diffraction due to box dimensions, floor bounce cancellation etc. Is the first thing to look for.
The 300hz+ midrange area is typical for such effects showing theyre ugly head.
You need to make a gated measurement as pointed out, that is so short you will not include any reflections or boundary effects.
Outside is the best as was pointed out.
The 300hz+ midrange area is typical for such effects showing theyre ugly head.
You need to make a gated measurement as pointed out, that is so short you will not include any reflections or boundary effects.
Outside is the best as was pointed out.
Thank you all for helping me out 
My stele operated neck and also my back is just to bad to move them at the time, im living with nervpain 24/7 but are a little better a few hours now and then.
I will try to move them more to "the middle of the room" when i can, and try with a gate of 8 to 10ms before first reflection.
Lowe powerful big Mcintosh,Krell and Pass lab amps and big speakers, & used to be able to move them around with ease, 82 kilo Mcintosh mc-602 or Krell fpb-600 was nemas problemas.
These days i can hardly lift or move anything, really sad for a guy who loves loud big things.
My stele operated neck and also my back is just to bad to move them at the time, im living with nervpain 24/7 but are a little better a few hours now and then.
I will try to move them more to "the middle of the room" when i can, and try with a gate of 8 to 10ms before first reflection.
Lowe powerful big Mcintosh,Krell and Pass lab amps and big speakers, & used to be able to move them around with ease, 82 kilo Mcintosh mc-602 or Krell fpb-600 was nemas problemas.
These days i can hardly lift or move anything, really sad for a guy who loves loud big things.
Small baffle mids are hard to get linear - ish response from. The little baffle area makes it more sensitive to other influences compared to a large, wide baffle. The diffraction and boundary cancelations are prominent, especially if the crossover is lower than the baffle area can support without significant baffle step compensation.
If the location you're measuring at is the final chosen placement of the speaker, then you should optimize the crossover for that location. Including all of its influencing conditions, so you end up with a speaker that.sounds better in that specific location. This may be a bigger compromise than is acceptable, making the solution more complicated and the desired FR much more placement dependant.
Gating the measurement definitely is suggested, but that will skew actual FR and won't include the physical placement's influences on the final FR, not fixing the response in the end. I'd redesign the mid enclosure with a larger baffle area unless you absolutely want a minimal baffle size and width. You can also play around with the mid placement in relation the LF section. I'd focus on finding the acoustic center first, aligning the drivers better which makes filter design easier and more predictable. You'll end up with a better performing speaker in the end. Using DSP would of course be the easier way but that brings another set of requirements with it.
If the location you're measuring at is the final chosen placement of the speaker, then you should optimize the crossover for that location. Including all of its influencing conditions, so you end up with a speaker that.sounds better in that specific location. This may be a bigger compromise than is acceptable, making the solution more complicated and the desired FR much more placement dependant.
Gating the measurement definitely is suggested, but that will skew actual FR and won't include the physical placement's influences on the final FR, not fixing the response in the end. I'd redesign the mid enclosure with a larger baffle area unless you absolutely want a minimal baffle size and width. You can also play around with the mid placement in relation the LF section. I'd focus on finding the acoustic center first, aligning the drivers better which makes filter design easier and more predictable. You'll end up with a better performing speaker in the end. Using DSP would of course be the easier way but that brings another set of requirements with it.
Appreciate your detailed reply profiguy
No the speakers real placement will be 1 m to the left and pointing in another direction.
Trying to build a potent 3 way that doesn't cover half my windows in the living room, like my XXL JBL 250 clones do
But at the time my previous diy XXL JBL 250 clones stand in the way....and 155 kilo each.
Also the woofer was build for playing sidemounted (to get a slim visual impression 40 insteed of 60 cm wide)
Is both wide and high equally important for a midrange baffle?
Or is width more important then hight?
It kind of depends on their relative sizes. A 20cm wide 1.2m tall enclosure will be less sensitive to a slight height adjustment than width. Keep in mind that the top panel in that case has less angular representation (looking outward from the driver).
Something to consider: profiguy pointed out that room interference influences the perceived sound at the listening spot in a(ny) room. First suspect here is the ‘floor bounce’ or first reflection from the floor. That one often causes a well defined ripple in the in-room response at the mentioned frequencies. You can correct it by equalizing, but the power response will be affected.
Floor bounce can be a big issue. The way I prefer to reduce it is by using multiple drivers covering the lower midrange. That sort of diffuses its effects and also helps with diffraction as well as the benefit of a slight increase in sensitivity (multiple driver array gain). I usually build all my systems with more than one LF driver due to the multiple benefits it brings with it. Some other bonuses are better LF dynamic range, smoother in room response across a wider listen area and less IMD and THD in the lower mids.
The only drawback IMO is the uneven directivity in the lower mids it can produce, which can result in a thick sounding lower midrange. This is more of an issue with 2.5 way designs, as the LF drivers have to cover more range not having a separate midrange as with a 3 way. The issue can be mostly eliminated though careful crossover design (ie. staggered separate LP filters on each LF driver). It also reduces the need for BSC and allows the use of lower sensitivity LF drivers with more xmax, which opens up LF driver selection considerably.
The only drawback IMO is the uneven directivity in the lower mids it can produce, which can result in a thick sounding lower midrange. This is more of an issue with 2.5 way designs, as the LF drivers have to cover more range not having a separate midrange as with a 3 way. The issue can be mostly eliminated though careful crossover design (ie. staggered separate LP filters on each LF driver). It also reduces the need for BSC and allows the use of lower sensitivity LF drivers with more xmax, which opens up LF driver selection considerably.
Last edited:
@jawen While I can imagine the benefits of extending those biradial WGs with spheres on your existing soeakers, I also have to think how tolerant your better half must be living with the asthetic aspect. My wife puts up with my hobby quite a bit, especially how connected it is with my main line of work, but she would never agree to having speakers like yours in our living space. Just an observation... sorry for side tracking.
HxW sets polar response, so whatever works best overall in your app:
Horn mouth/baffle polar frequency: c = 10^6/(a*b)
a = wall angle in deg
b = mouth width (or height) in inches
c = -6 dB frequency
metric = (2.54*10^6)/(a*b) in cm
Trying to learn "IR window aka gate" settings in REW.
And haven´t done new mesaurements BUT have change the same mesaurements IR wIndows in REW. ( had bad days)
Settings now is Left window 312 ms, Ref time is 0 and Right window is 8 ms.
Impulse look like this after "new" IR window settings in REW
Does it show "things" right when i change "old" mesaurements IR windows afterwards?, or must it be changed before the mesaurements?
So much to learn in diy, and my "new" medecine Gabapentin helps me a little to endure the nerve pain, but also makes my brain sluggish, so new learning and problem solving or mathematics are now impossible for me to do some hours after taking medicine. (very frustrating)
best regards John
Attachments
Hahaha...understand what you mean
My 16 & 18 years old daughters think they look like boobs.
And great benefits of extending these JBL 2344a biradial!
Old 10 kilo Swedish Aga/Baltic cd is playing from 800 hz to 9 K (12 dB filter), and can be used from 450 hz and all the way up.
So quite large spectrum and because of that = more benefits
Sweetspot is so wide and still superb placement of singer and the rest of the orchestra/band
About "better half" ..Live alone sinse 12 year back now which largely is "self-selected"....even if is boring some times.
And last relationship that lasted a long time, I already had 1200 kilos of stereo equipment in the living room when we met ( think much it's all about habits and sincerity)
So she "begain" with that, and was sincerely real happy when i let one pair of speakers go hahaha (even if i had 3 pair left)