Hello,
I'm planing to build a floorstanded speaker using a 10" mid-woofer and 1.2" tweeter. Hoped Fx will be around 1.2kHz steep slopes.
For floor boundary reasons I want the woofer center to be 11.6" above floor. For listening axis reasons I want the tweeter center to be 28.9" above floor.
Simulations of diffraction/boundary/room-gain are really nice.
But...the 17.3" distance center woofer/center tweeter is actually more than the wave length at 1.2kHz_11.26".
I know this fact will create frequency response cancellations but wish to know how harmful they will be sonically at say 10 feet listening position...
Any thought/experience about?
Thank you very much.
I'm planing to build a floorstanded speaker using a 10" mid-woofer and 1.2" tweeter. Hoped Fx will be around 1.2kHz steep slopes.
For floor boundary reasons I want the woofer center to be 11.6" above floor. For listening axis reasons I want the tweeter center to be 28.9" above floor.
Simulations of diffraction/boundary/room-gain are really nice.
But...the 17.3" distance center woofer/center tweeter is actually more than the wave length at 1.2kHz_11.26".
I know this fact will create frequency response cancellations but wish to know how harmful they will be sonically at say 10 feet listening position...
Any thought/experience about?
Thank you very much.
Yes, I thought about but then it would be some 8" below the ears listening height; so wouldn't this be detrimental on imaging?
edit: regarding the dimension of the woofer frame I could place the tweeter 26.8" above floor, thus limiting the drivers center to center offset to 15.2" so less than 3" more than wave length at 1.2kHz...And probably the 1.2kHz emission in a 10" woofer isn't at its center, so the one wave length rule should be almost OK...
Thus the only question remains about having the tweeter 4" below listening height...Could make the trick, no?
Thus the only question remains about having the tweeter 4" below listening height...Could make the trick, no?
I don't think that 4" will make that much difference. I looked at this thread this morning (but was distracted). If I remember rightly your lobing will give you a null at around 14 degrees above your listening position, and this is if you match the phase to suit your 4 degree vertical listening angle.
AllenB, thanks for your input but I'm not sure to understand what you mean...
Why are you telling about "14 degrees above my listening position" ?
My ears will be at 2.9° (6.4") above tweeter H axis and 9.6° (21.3") above woofer H axis.
I will design the filter to match drivers phases at Fx by placing my mic at listening height (33") and 59" distance from speaker (this is where I can windowing before first reflexion).
Why are you telling about "14 degrees above my listening position" ?
My ears will be at 2.9° (6.4") above tweeter H axis and 9.6° (21.3") above woofer H axis.
I will design the filter to match drivers phases at Fx by placing my mic at listening height (33") and 59" distance from speaker (this is where I can windowing before first reflexion).
As you move above your listening position your woofer will become further from you until it is a half wavelength further than your tweeter causing a cancellation. The further apart your drivers are at the crossover, the shorter this distance will be.
Anyway, I think you should have no problems with this. I mean it would not be ideal but you should be able to get good sound this way.
Anyway, I think you should have no problems with this. I mean it would not be ideal but you should be able to get good sound this way.
Supposing an LR xover, vertical beamwidth is given by:
Arcsin ( λ / 2d )*2 where λ is wavelength at xover F and d is vertical offset ( see Linkwitz for more about this). Which gives in your case about 22º*2=44º.
So that whether your offset is ok or not depends on your vertical coverage needs. 1/2 wavelength is for 90º, 1 wavelength for 60º, but do you really need so much?
Imho 40º is fine enough.
Arcsin ( λ / 2d )*2 where λ is wavelength at xover F and d is vertical offset ( see Linkwitz for more about this). Which gives in your case about 22º*2=44º.
So that whether your offset is ok or not depends on your vertical coverage needs. 1/2 wavelength is for 90º, 1 wavelength for 60º, but do you really need so much?
Imho 40º is fine enough.
Member
Joined 2009
Member
Joined 2009
I don't want to go to a 2.5-way or 3-way because my initial challenge is to mate a 10" to a tweeter.
As my need is to listen no more than ears height and favouring to avoid midbass floor cancellations, I will build my speaker with an offset of 15.2" and steep slopes at Fx to minimise vertical beamwidth issues.
Anyway thanks to all for your comments.
Hope I will remember to give you some news about this matter after completing of this speakers.
As my need is to listen no more than ears height and favouring to avoid midbass floor cancellations, I will build my speaker with an offset of 15.2" and steep slopes at Fx to minimise vertical beamwidth issues.
Anyway thanks to all for your comments.
Hope I will remember to give you some news about this matter after completing of this speakers.
Member
Joined 2009
Some news from my almost finished project using a tweeter (SB29) placed far away from a (close to floor) woofer (SEAS CA26RE4X)...
Finally it turned out to be quite interesting in term of listening results, with a lot of dynamics mixed with a very good level of details and smoothness; obviously the challenge was not only to control V directivity issues but also to design the filter without measuring close (usually 1 or 1.5 meter) to speaker_yes phase cancellations are huge at this distance...So I decided to measure at 3M, window 300Hz and smoothing things at 1/3octave for sims; this process gave great results; however the filter design was a lot of work, 2 x 4th order aren't easy to design right, particularly in these conditions! Transfert functions are particularly clean.
Room bounces didn't give me the possibility to measure the reverse nul, but it can be view on sims; non-windowed measurement at listening position gives an almost linear range (+/- 1.5dB) slightly decreasing (no screen capture, sorry).
Almost no BSC and some 91dB/2.83V are partly due to the woofer placed close to floor; all in all I would say that this drivers combo and filter doesn't need any mid-driver; the tweeter doesn't undergo any form of stress in spite of the 1.1kHz Fx and listening stays quite acceptable even if I stand up at listening position...
the drawing, to remember drivers placement:
Finally it turned out to be quite interesting in term of listening results, with a lot of dynamics mixed with a very good level of details and smoothness; obviously the challenge was not only to control V directivity issues but also to design the filter without measuring close (usually 1 or 1.5 meter) to speaker_yes phase cancellations are huge at this distance...So I decided to measure at 3M, window 300Hz and smoothing things at 1/3octave for sims; this process gave great results; however the filter design was a lot of work, 2 x 4th order aren't easy to design right, particularly in these conditions! Transfert functions are particularly clean.
Room bounces didn't give me the possibility to measure the reverse nul, but it can be view on sims; non-windowed measurement at listening position gives an almost linear range (+/- 1.5dB) slightly decreasing (no screen capture, sorry).
Almost no BSC and some 91dB/2.83V are partly due to the woofer placed close to floor; all in all I would say that this drivers combo and filter doesn't need any mid-driver; the tweeter doesn't undergo any form of stress in spite of the 1.1kHz Fx and listening stays quite acceptable even if I stand up at listening position...
An externally hosted image should be here but it was not working when we last tested it.
the drawing, to remember drivers placement:
An externally hosted image should be here but it was not working when we last tested it.
Reverse the polarity of the tweeter, use a real time analyser and pink noise. Hold the mic in your hand and sweep it up and down in front of the speaker, maybe a few feet away. You should be looking to see that all frequencies around the crossover region are reaching their minimums close to on axis.
You'll also need to do something about C5, it really shouldn't be there.
Allen and jerome,
I thought such low impedance at so high fr wasn't important ?!...
A screen shot with R10ohm in series with C5 and without the R:
jerome, thanks for your congratulations!
I do never analyse the filter topology I design, so LR6 or 8 ?:
I thought such low impedance at so high fr wasn't important ?!...
A screen shot with R10ohm in series with C5 and without the R:
An externally hosted image should be here but it was not working when we last tested it.
jerome, thanks for your congratulations!
I do never analyse the filter topology I design, so LR6 or 8 ?:
An externally hosted image should be here but it was not working when we last tested it.
You see the difference 1.5ohms and 4.5ohms. Fortunately you don't push the test higher.
Yes it is important because some (bad in general) amplifiers can have high frequency oscillations and can overheat. It could reduce stability of the amp. with such low impedance at high frequency. The third reason i see, is you can amplify radio frequency and it's not good for the (very good) amplifier to drive 0.5ohms at 100kHz.
I see near 8th order (45dB/oct.) for the tweeter and near 6th order (36dB/oct.) for the tweeter. But in fact you have a LR2 near the crossover region
Yes it is important because some (bad in general) amplifiers can have high frequency oscillations and can overheat. It could reduce stability of the amp. with such low impedance at high frequency. The third reason i see, is you can amplify radio frequency and it's not good for the (very good) amplifier to drive 0.5ohms at 100kHz.
I see near 8th order (45dB/oct.) for the tweeter and near 6th order (36dB/oct.) for the tweeter. But in fact you have a LR2 near the crossover region
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.