Hi all,
I'm reviving an old project and need opinions and suggestions as to how to do the crossover from satellite to the subwoofer. The satellite speaker is HDS exclusive 5,5" and HDS 810921 tweeter in either vented or closed box of approximately 8 litres, one of the things I'd like to hear other peoples opinions on. Subwoofer is XLS 12" in ~40l closed box. I'm thinking of crossing over at about 80-85 Hz LR4 electrical, but I'm open to suggestions.
Few graphs I whipped up on WinISD:
Unfiltered response graph, no baffle-step simulated or corrected for HDS and WinISD seems to be over-zealous about factoring in voice-coil inductance (at least with this driver), XLS 12" should go straight to about 200 Hz:
Yellow is XLS 12", grey HDS vented and red HDS in closed box
Filtered rensponse:
Group delay (with LR4):
Max SPL @ xmax:
In vented box HDS will go about 4dB louder at xmax, wont need a linkwitz transform circuit and the group delay doesn't seem to be concern? Is there any downside for going vented? Also, the actual crossover frequencies are 80 Hz for the HDS's and 85 Hz for XLS, and there is not much overlap between the two, but neither doesn't seem to be a concern to me, judging from the group delay and frequency response graph. Am I overlooking something?
Looking at the transfer function graph it looks that at 85 Hz response starts to slope down gently F3 being @ 38 Hz and F6 @ 23 Hz. Does that seem ok for medium/small room, with the speakers about 0.7m/2 feet off the back wall, or will it sound thin/muddy?
Well.. that's it for now. I will be back with more question later.
I'm reviving an old project and need opinions and suggestions as to how to do the crossover from satellite to the subwoofer. The satellite speaker is HDS exclusive 5,5" and HDS 810921 tweeter in either vented or closed box of approximately 8 litres, one of the things I'd like to hear other peoples opinions on. Subwoofer is XLS 12" in ~40l closed box. I'm thinking of crossing over at about 80-85 Hz LR4 electrical, but I'm open to suggestions.
Few graphs I whipped up on WinISD:
Unfiltered response graph, no baffle-step simulated or corrected for HDS and WinISD seems to be over-zealous about factoring in voice-coil inductance (at least with this driver), XLS 12" should go straight to about 200 Hz:
An externally hosted image should be here but it was not working when we last tested it.
Yellow is XLS 12", grey HDS vented and red HDS in closed box
Filtered rensponse:
An externally hosted image should be here but it was not working when we last tested it.
Group delay (with LR4):
An externally hosted image should be here but it was not working when we last tested it.
Max SPL @ xmax:
An externally hosted image should be here but it was not working when we last tested it.
In vented box HDS will go about 4dB louder at xmax, wont need a linkwitz transform circuit and the group delay doesn't seem to be concern? Is there any downside for going vented? Also, the actual crossover frequencies are 80 Hz for the HDS's and 85 Hz for XLS, and there is not much overlap between the two, but neither doesn't seem to be a concern to me, judging from the group delay and frequency response graph. Am I overlooking something?
Looking at the transfer function graph it looks that at 85 Hz response starts to slope down gently F3 being @ 38 Hz and F6 @ 23 Hz. Does that seem ok for medium/small room, with the speakers about 0.7m/2 feet off the back wall, or will it sound thin/muddy?
Well.. that's it for now. I will be back with more question later.
Sorry, my graphs must've been misleading in that I forgot to mention that the sealed HDS has linkwitz transform circuit in place to flatten the response to match vented and allow 80-85 Hz XO-point. I'll edit that into the first post.
Here is the port gain graphed:
As can be easily seen the port offers gain at the operating band, even though it is tuned at 55 Hz and crossover is at 80 Hz. (Why the gain looks like it's centered around 85 Hz is because of the LR4 highpass in place) HDS in sealed box will exceed xmax at around 85 Hz llimiting the max. output to 101 dB, vented will go to 105 dB before exceeding xmax at 100 Hz. From both baffle-step correction of about 4 dB (?) will need to be taken off from max. spl.
EDIT: Can't edit the first post anymore, so It'll have to do
Here is the port gain graphed:
An externally hosted image should be here but it was not working when we last tested it.
As can be easily seen the port offers gain at the operating band, even though it is tuned at 55 Hz and crossover is at 80 Hz. (Why the gain looks like it's centered around 85 Hz is because of the LR4 highpass in place) HDS in sealed box will exceed xmax at around 85 Hz llimiting the max. output to 101 dB, vented will go to 105 dB before exceeding xmax at 100 Hz. From both baffle-step correction of about 4 dB (?) will need to be taken off from max. spl.
EDIT: Can't edit the first post anymore, so It'll have to do
I have considered doing that, but I'll be probably quite happy with the SPL/distortion I can get from the speakers as it is (and I want to keep them small). Actually I just gained a healthy 1 dB more by changing the tuning frequency to about 80 Hz, equalizing out the pump the port makes and used 2nd order filter instead of 4th (the port now supplies the other 12 dB of roll-off).
Tracking around the XO-point and group delay in general improved also and the driver is moving less at 70-100 where it's 3rd component distortion starts to rise, so I'm pretty happy about the chances, unless someone points out issues in the design as it is now.
BTW. max spl from the HDS driver is constrained by power handling now, rather than xmax.
Tracking around the XO-point and group delay in general improved also and the driver is moving less at 70-100 where it's 3rd component distortion starts to rise, so I'm pretty happy about the chances, unless someone points out issues in the design as it is now.
BTW. max spl from the HDS driver is constrained by power handling now, rather than xmax.
Am I the only one who doesn't see any graphs?
Not being able to see anything I'll hazard some observations anyway. Since you can tweak the enclosures for response and smooth integration with a sub consider this: If you can develop an enclosure for the HDS with a roughly .707 Qt roll off than you can implement a BW2 high pass and LR4 lowpass at the F3 of the HDS enclosure for perfect summation.
Not being able to see anything I'll hazard some observations anyway. Since you can tweak the enclosures for response and smooth integration with a sub consider this: If you can develop an enclosure for the HDS with a roughly .707 Qt roll off than you can implement a BW2 high pass and LR4 lowpass at the F3 of the HDS enclosure for perfect summation.
Excellent rule to remember, thank you! I was 2 Hz off (hehe), tuning mostly by group delay graph. The graphs are in png format, maybe your browser doesn't support those?
Does anyone know if there's some side-effects of forcing a way too highly tuned vented box (peaking response at tuning frequency of 4 dB or so) to flat by equalization?
Does anyone know if there's some side-effects of forcing a way too highly tuned vented box (peaking response at tuning frequency of 4 dB or so) to flat by equalization?
augerpro,
That is a very sound idea. If the natural rolloff from a seal boxed is used we don't need a high pass there which keeps the component count at a minimum (cleaner sound) and the group delay low. That is what I do with my current speakers. But do you mean a second order rolloff for the HDS and a LR4 high pass for the sub?
Samuli,
For a 3 way I would not normally cross the HDS 5" lower than 300Hz but if you don't demand a high SPL then your plan seems fine. I believe your 80Hz is a good choice because the XLS demands very high power when it is above 80Hz. Check Linkwitz site for his XLS sub and power chart. Below 80Hz, a 100W amp can do very well. Above that, the power requirement is so large that it is really impractical if possible. I have done quite a bit of modeling on this.
I would choose sealed rather than ported enclosure for the HDS to integrate with the XLS better. If ported, frequencies below 80Hz can cause large excursion that may damage your HDS.
Regards,
Bill
That is a very sound idea. If the natural rolloff from a seal boxed is used we don't need a high pass there which keeps the component count at a minimum (cleaner sound) and the group delay low. That is what I do with my current speakers. But do you mean a second order rolloff for the HDS and a LR4 high pass for the sub?
Samuli,
For a 3 way I would not normally cross the HDS 5" lower than 300Hz but if you don't demand a high SPL then your plan seems fine. I believe your 80Hz is a good choice because the XLS demands very high power when it is above 80Hz. Check Linkwitz site for his XLS sub and power chart. Below 80Hz, a 100W amp can do very well. Above that, the power requirement is so large that it is really impractical if possible. I have done quite a bit of modeling on this.
I would choose sealed rather than ported enclosure for the HDS to integrate with the XLS better. If ported, frequencies below 80Hz can cause large excursion that may damage your HDS.
Regards,
Bill
Thanks for confirming the 80 Hz XO-point is fine. As for the HDS in it's modelled enclosure now (8,35 litres and port tuning at 80 Hz), I couldn't possibly get more SPL out of it because the thermal power rating for the element reduces output before the driver exceeds linear xmax at ANY frequency. There's a 2nd order highpass in place @ 80 Hz that keeps it from bottoming out at about 50 Hz, where the highest excursion lies (+-3mm). At that frequency HDS is already down 24 dB.
As for vented vs. non-vented I can get virtually the same roll-off and group delay by forcing both with active electronics, vented will just go louder/less distortion, but am I overlooking something here?
As for vented vs. non-vented I can get virtually the same roll-off and group delay by forcing both with active electronics, vented will just go louder/less distortion, but am I overlooking something here?
HiFiNutNut said:
Qt of .707 is a BW2 already. Cascading an additional electric BW2 results in an LR4 rolloff. So yes, elctrical BW2 high pass for the HDS and electrical LR4 for the sub results in the typical LR4 summation. You can use the typical HT receiver to do this or an external active XO of some sort.
augerpro,
That makes sense. At 80Hz I would consider 2nd or 3rd order though. 2nd order may not be optimal for the XLS but for the HDS it means lower group delay and less components (ditching the electronic LR2). From what I read with regards to group delay, which may be controversal, I would hesitate using 4th order below 100Hz.
Samuli,
You can always stuff the port to make it behave like a sealed box. Sound waves produced from a port have very different phase characteristics and you may have problem producing a flat response when crossing over to the XLS.
80Hz is good for the XLS. Whether it is good for the HDS depends on how much SPL you want.
Regards,
Bill
That makes sense. At 80Hz I would consider 2nd or 3rd order though. 2nd order may not be optimal for the XLS but for the HDS it means lower group delay and less components (ditching the electronic LR2). From what I read with regards to group delay, which may be controversal, I would hesitate using 4th order below 100Hz.
Samuli,
You can always stuff the port to make it behave like a sealed box. Sound waves produced from a port have very different phase characteristics and you may have problem producing a flat response when crossing over to the XLS.
80Hz is good for the XLS. Whether it is good for the HDS depends on how much SPL you want.
Regards,
Bill
Hi Twahach,
Is 106-107 dB (-4 dB from BSC, +6 dB from the other speaker) not enough? Because that is what HDS's can do in the enclosure/XO without exceeding linear xmax at any frequency. The set will not be used for HT, btw.
If choose to go sealed, it'll go 101 dB, but I can't see any benefit from sealed enclosure. The response and group delay are identical for both, although what HifiNutNut is saying about the phase response makes me wonder... isn't already taken account in the simulation?
After all, it is the frequency response that dictates the phase and GD of a system. Those seem to be in order because HDS tracks the sub very well (within 1ms, or under 36 deg phase disprecancy, and that's the worst case, everywhere else it's more like 0,1ms difference). Soo.. right now I'm going by the group delay figures WinISD gives, but is there more to vented boxes the program might be overlooking?
Using LR4 @ 80 Hz won't be problem for me concerning group delay. It actually improves very little by raising XO to, say, 100 Hz. LR2 is out of the question, I'm lot more willing to risk having some group delay in the system, of which audibility is at debate, than letting sub go strong into the 150 Hz-200 Hz region and probably having to use the inferior (?) sealed cabinet with less output/more distortion.
Is 106-107 dB (-4 dB from BSC, +6 dB from the other speaker) not enough? Because that is what HDS's can do in the enclosure/XO without exceeding linear xmax at any frequency. The set will not be used for HT, btw.
If choose to go sealed, it'll go 101 dB, but I can't see any benefit from sealed enclosure. The response and group delay are identical for both, although what HifiNutNut is saying about the phase response makes me wonder... isn't already taken account in the simulation?
After all, it is the frequency response that dictates the phase and GD of a system. Those seem to be in order because HDS tracks the sub very well (within 1ms, or under 36 deg phase disprecancy, and that's the worst case, everywhere else it's more like 0,1ms difference). Soo.. right now I'm going by the group delay figures WinISD gives, but is there more to vented boxes the program might be overlooking?
Using LR4 @ 80 Hz won't be problem for me concerning group delay. It actually improves very little by raising XO to, say, 100 Hz. LR2 is out of the question, I'm lot more willing to risk having some group delay in the system, of which audibility is at debate, than letting sub go strong into the 150 Hz-200 Hz region and probably having to use the inferior (?) sealed cabinet with less output/more distortion.
Hi Twahach and all.
I'm pretty much decided on going with the vented version of the box. I can easily seal it if it doesn't work out.
Here's a few graphs how it's looking like now.
Filtered response:
Red is HDS sealed with linkwitz transform and LR4 highpass in place
Grey is HDS vented with equalization and LR4
Yellow is XLS 12" in 40l sealed box with LT, subsonic 6dB @ 15 Hz and lowpass @ ~80 Hz.
Group delay:
Driver excursion: Vented is driven 80W/106-107dB and closed 25W/101dB (the program says, actually it's more because WinISD doesn't seem to factor LT-equalization in, dB figure is right though)
Unfiltered/unequalized HDS vented:
I'm still a bit worried if there's problems ahead with the vented I'm not taking into account... is the parametric equalization I'm using really panacea for all?
What about the roll-off towards the subbass, too shallow, too deep, wrong curve? The crossover curves from XLS to HDS clearly aren't perfectly aligned either, but I'm not too worried about that, at this point anyways.
I'm pretty much decided on going with the vented version of the box. I can easily seal it if it doesn't work out.
Here's a few graphs how it's looking like now.
Filtered response:
Red is HDS sealed with linkwitz transform and LR4 highpass in place
Grey is HDS vented with equalization and LR4
Yellow is XLS 12" in 40l sealed box with LT, subsonic 6dB @ 15 Hz and lowpass @ ~80 Hz.
An externally hosted image should be here but it was not working when we last tested it.
Group delay:
An externally hosted image should be here but it was not working when we last tested it.
Driver excursion: Vented is driven 80W/106-107dB and closed 25W/101dB (the program says, actually it's more because WinISD doesn't seem to factor LT-equalization in, dB figure is right though)
An externally hosted image should be here but it was not working when we last tested it.
Unfiltered/unequalized HDS vented:
An externally hosted image should be here but it was not working when we last tested it.
I'm still a bit worried if there's problems ahead with the vented I'm not taking into account... is the parametric equalization I'm using really panacea for all?
What about the roll-off towards the subbass, too shallow, too deep, wrong curve? The crossover curves from XLS to HDS clearly aren't perfectly aligned either, but I'm not too worried about that, at this point anyways.
Hi
I think of it that's a few times. i find 12" parameter.
i see the QTS. it's too low.It's difficult to design.
I try to simulate it with a vented box.it's not well.
I see the graph curve that you sent. xo at 80 hz i think
it's good respond curve for 12" . The group delay is high at
20 hz above. and the driver excursion is high.i think that you
simulate at 300 watt. I agree that you use subsonic 6 db.
and the LT4 for 12".The low frequency respond when you use
LT4.it's good.The low frequency slope is drop just a little.
It still has a deep bass.If the simulator is correct that
i can see. i think you can build it.....
/Thawach
I think of it that's a few times. i find 12" parameter.
i see the QTS. it's too low.It's difficult to design.
I try to simulate it with a vented box.it's not well.
I see the graph curve that you sent. xo at 80 hz i think
it's good respond curve for 12" . The group delay is high at
20 hz above. and the driver excursion is high.i think that you
simulate at 300 watt. I agree that you use subsonic 6 db.
and the LT4 for 12".The low frequency respond when you use
LT4.it's good.The low frequency slope is drop just a little.
It still has a deep bass.If the simulator is correct that
i can see. i think you can build it.....
/Thawach- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.