Wondering if the plastic horn could be resonating badly, but not showing up in the freq-response?
It appears to reasonably solid, but I feel some vibration in the plastic at high levels.
Kind regards TroelsM
It appears to reasonably solid, but I feel some vibration in the plastic at high levels.
Kind regards TroelsM
You can't do these measurements indoors unless gated or close mice'd as the objects vibrating in the room add distortion.
Also your hearing protection may not be sufficient as attenuation is typically 20 - 30 dB.
The best way to do the measurement would be at distance outdoors but as that's not possible the next best thing is to close mic the drivers (so that direct sound dominates the room) and use a very high SPL rated mic capsule (>145dB are available) however these mics are very expensive, perhaps rental would be an option?
You would expect resonances to show up as narrow band increases in THD, you could also check waterfall plots etc.
Also your hearing protection may not be sufficient as attenuation is typically 20 - 30 dB.
The best way to do the measurement would be at distance outdoors but as that's not possible the next best thing is to close mic the drivers (so that direct sound dominates the room) and use a very high SPL rated mic capsule (>145dB are available) however these mics are very expensive, perhaps rental would be an option?
You would expect resonances to show up as narrow band increases in THD, you could also check waterfall plots etc.
The way you've blocked the outer half of each cone will theoretically reduce the center to center spacing of the drivers and reduce comb filtering at upper end of the range as well. Wonder how effective that is
Hi Troels,
the driver seems to be an A&D K6D210. If you use the drivers in parallel you have a sound pressure of about 122 dB at 40 Vpp and 0.6 m. For inexpensive microphones this is already in the range of the maximum SPL (e.g. Behringer ECM8000 = 3% THD @ 120dB, tBone MM1 = 3% THD @ 118dB; see also http://www.artalabs.hr/AppNotes/AN5-MicroMeasChamber-Rev4Eng.pdf, page 16).
Regards
Heinrich
the driver seems to be an A&D K6D210. If you use the drivers in parallel you have a sound pressure of about 122 dB at 40 Vpp and 0.6 m. For inexpensive microphones this is already in the range of the maximum SPL (e.g. Behringer ECM8000 = 3% THD @ 120dB, tBone MM1 = 3% THD @ 118dB; see also http://www.artalabs.hr/AppNotes/AN5-MicroMeasChamber-Rev4Eng.pdf, page 16).
Regards
Heinrich
The cabinet edges will cause a very high level of diffraction. Diffraction is known to increase in audibility at high SPLs. But there could also be a myriad of other issues as discussed; mic, ears, mic preamp, unknown amp clipping, maybe crossover shift with high thermal load, etc. It can be very hard to sort this kind of thing out without some effective resources, like good mics, measurements and measurement conditions.
FWIW, I also find that room reflections also become more audible at high levels, leading to a very "messy" sound which is easily recreated by listening to an omni mic via headphones.
A further FWIW, you can do distortion testing using a dynamic microphone. The frequency response won't be correct, but any decent dynamic mic will take >150dB cleanly.
Chris
A further FWIW, you can do distortion testing using a dynamic microphone. The frequency response won't be correct, but any decent dynamic mic will take >150dB cleanly.
Chris
ah that's a cool tip Chris.
You could try recording the speaker and listening on headphones at lower volume and see if the distortion is still there.
You could try recording the speaker and listening on headphones at lower volume and see if the distortion is still there.
I really appreciate all the input from you guys.
Diffraction: I had not thought it would be a big problem, but maybe Im wrong. Shouldn't massive diffraction show up in the freq-response? - how would you measure it?
Overloading the ears: I'm open to the possibility that this maybe be loud enough, to be too loud 🙂 - at least for critical nearfield listening, but I do believe some of the harshness/distortion i also present at lower levels or greater distance.
Dynamic mic: One more thing, I did not consider 🙂. It appears that dynamic mics start at relative low prices. I might get one.
This weekend I had the chance to compare the current project with 2 older (smaller) active speakers, and the 2 older also sounded a bit harsh in the same setting and with the same track, so Im inclined to think that the problem is not just with the new speaker, - although the new one is much higher and thus the problems become more apparent.
For some reason I ended using this track as test: Kiesza - Hideaway (Official Music Video) - YouTube
Her voice sounds fine in headphones but at high levels on my cheap active diy-speakers the voice becomes very agressive/resonant/harsh ( I dont know how to describe sound)
Kind regards TroelsM
Diffraction: I had not thought it would be a big problem, but maybe Im wrong. Shouldn't massive diffraction show up in the freq-response? - how would you measure it?
Overloading the ears: I'm open to the possibility that this maybe be loud enough, to be too loud 🙂 - at least for critical nearfield listening, but I do believe some of the harshness/distortion i also present at lower levels or greater distance.
Dynamic mic: One more thing, I did not consider 🙂. It appears that dynamic mics start at relative low prices. I might get one.
This weekend I had the chance to compare the current project with 2 older (smaller) active speakers, and the 2 older also sounded a bit harsh in the same setting and with the same track, so Im inclined to think that the problem is not just with the new speaker, - although the new one is much higher and thus the problems become more apparent.
For some reason I ended using this track as test: Kiesza - Hideaway (Official Music Video) - YouTube
Her voice sounds fine in headphones but at high levels on my cheap active diy-speakers the voice becomes very agressive/resonant/harsh ( I dont know how to describe sound)
Kind regards TroelsM
To check room / speaker resonance, the most practical way would be playing back steady sine wave, not sweep. Increase the frequency very slowly (or step), and when you hear the resonance, stop increasing, then you can easily detect which object is resonating.
Diffractions effect on the frequency response has been shown to be small when compared to its audibility especially at higher SPLs. This means that it might be hard to find the diffraction in a frequency response measurement alone even though it may have a strong effect on perception at high SPL. Detecting this would involved a close look at either the non-minimum phase and/or the impulse response.
And please keep in mind that diffraction can and does occur from objects in the room that are seperate from the speakers themselves. In my system I have taken great pains to remove all these early diffractions and my system never gets "muddy" even at excessive SPLs - even in a very small room.
What about vibrations in cab, horn-plastic and other parts. I know, I can build everything form lead and sand to avoid any resonance, but thats not very portable. Cab is 14mm plywood ( birch/spruce). there is solid vertical brace in middle and overall it dosnt ring badly.
I have considered making a "pickup/microphone" from and old speaker to sniff out which parts of the construction emit sound, - apart from the cones. - any idea if that would work?
Kind regards TroelsM
I have considered making a "pickup/microphone" from and old speaker to sniff out which parts of the construction emit sound, - apart from the cones. - any idea if that would work?
Kind regards TroelsM
Sure you can do that, you'll get some numbers, but like everyone else you will find that interpreting them is really problematic. I.e. they don't really mean a thing.
The idea would be that if a panel/part/horn had specific (high) resonant peaks, then those could be addressed with damping?
Maybe its a waste of time.
I'll try to damp some of the diffraction-edges and see(hear) if that makes any difference.
I did a quick freq response indoors and overall its appears "flat" from 200Hz (+/-4dB).
Kind regards TroelsM
Maybe its a waste of time.
I'll try to damp some of the diffraction-edges and see(hear) if that makes any difference.
I did a quick freq response indoors and overall its appears "flat" from 200Hz (+/-4dB).
Kind regards TroelsM
Did a quick listening test yesterday with some thin felt added to some of the edges on the baffle. it maybe helped a little on the harshness, but no major difference.
I'm considering making a bigger test-baffle (50*80cm?) and flush-mount all 3 drivers to see(hear) if that makes any difference to the sound. If the same harshness is heard on a much larger baffle, then I would assume its not diffraction from my small baffle?
Kind regards TroelsM
I'm considering making a bigger test-baffle (50*80cm?) and flush-mount all 3 drivers to see(hear) if that makes any difference to the sound. If the same harshness is heard on a much larger baffle, then I would assume its not diffraction from my small baffle?
Kind regards TroelsM
This observation would be consistent with the theory.
This observation is making a big leap of faith that the difference are due to diffraction alone, which is most likely not the case.
Hi Earl.
Thank you for participating in the discussion.
I'm not sure that I understand your input.
TroelsM
Thank you for participating in the discussion.
I'm not sure that I understand your input.
Consistent in that a small amount of felt would not make any difference, or that diffraction-distortion cannot be damped this way?
I realize that a bigger baffle ( and changed cabinet) will change more than one parameter, but if the harshness is STILL present on a bigger baffle ( with fewer diffraction issues), then it is not ONLY diffraction thats causing my problems?
TroelsM
I meant that your statement
is a consistent observation with what I have encountered regarding diffraction.
Larger baffles do not necessarily reduce diffraction and then there are all the other issues to consider. Subjective tests like these are always highly uncertain.