Have you by any chance inadvertently added the bass reflex alignment to the measurements in the FRD file? You don't have the diffraction response in the FRD file, which I assume is represented by the yellow dashed line. That's why your "bass reflex" response is rolling off at 36dB/octave instead of 24dB/octave.
Last edited:
Member
Joined 2003
Your providing a frequency response in 90dB range to a diffraction response input that's expecting 0-6dB range. Not the right way to do things, if you want to apply the enclosure model to your traced response, best to export the enclosure response and use the merge tool to splice it to the traced response.
Member
Joined 2003
On the surface this may look like a mess for impedance, but it tends to "pass through" the driver for the most part so place this before the driver and continue your cross, while keeping your eyes on the prize.. in other words the wanted response and phase.
Thanks! Maybe I'll use that in a later project again. For this project I'd need ~30mH inductances. 😀
Thanks for your response. Life has been busy so I'm just coming back to this.
Here's what my merge window looks like.
I've been following the "simulating without measurement" PDF, which was what guided me to this point. It advises:
Tonight I tried merging things around 4khz (well past the crossover point at about 2khz), and it looks more like I would expect. Not sure if I might be missing something since I'm merging things so much higher?
Hi, you have 85db LF -> HF step, this information is visible right side of the high frequency part. So you have to use the Scale on the nearfield data to make this step ~0db. Also, add cone area on the nearfield data row, or at least refer to manual what it is for as it relates to this gain matching. This seems to be something you are missing here.
Hey guys!
I have build these speakers 5 years ago. Its currently a "2 way speaker with build in subwoofers" as I'm crossing at 80 hz to the 10". There are TWO active 10" in each speaker - one on each side mounted back to back. The two woofers share the same volume. They are driven by a set of Hypex FA253's.
They sound great, especially the 20-100-ish Hz range/punch and rumble. Im very happy with that part. However, I belive I could do the integration of the tweeter and midwoofer better today. maybe also improving on the integration between the woofers and midrange.
I started using VCAD more since they were build and I think I am starting to get the hang of it. Therefore I was planning to redo all measurments of this speaker and get it simulated properly in VCAD. However, all other speakers i used VCAD for have crossovers in the 600+Hz range. So the crossover between the woofers and midrange in this speaker is challenging me.. add to that the dual side mounted configuration.
I am so locky that I can use a small anechoic chamber for the new measruements. The room is 2.4 x 4.6 x 2.0 m w/l/h (free space there are additional 3m of absorbtion all the way around that). No idea how deep in frequency this will let me go, but exited to see that. I use a Focusrite Scarlett 2i2 4th gen with loop back and REW for the measurements.
I have read the manual of VCAD from start to end as well as the REW specific help document (That one many times xD). However, Im still in doubt of how exactly to do the mesurments of especially my sidemounted woofers properly. From what I have read this is my best guess so far:
1. Tilt the speaker forward to make the baffel vertical
2. Place microphone 1 m from baffel at tweeter level
3. Measure tweeter 180 degree at 10 degree interval
4. Place mic at midrange level (stay 1 m from baffel)
5. Measure midrange 180 degree at 10 degree interval
6. Place mic a woofer height. Stay 1 m from baffel. Leave both woofers connected to amp.
7. Measure woofers 180 degree at 10 degree interval (this will mean they get significantly closer to the mic at the 180 degree position)
8. Do a measurement of both tweeter and midrange at 1 m from baffel placed at a Y value right between the two drivers.
9. Do a measurement of both midrange and woofers at 1 m from baffel placed at a Y value right between the two drivers.
10. Do near field measurements of midrange
11. Do near field measurement of woofers (with both woofers playing but one covered with pillows or whatever)
Then in VCAD I will set all drivers to Z=0 and adjust the Y value according to their location - relative to a vertical baffel. Then use the summation responses to adjust the delay of the drivers under the driver tab to ensure that the simulation matches the summed response.
Will this give a fairly decent representation of the woofers interaction with each other as well as the midrange driver?
I most likely will end up with a 80-is Hz crossover again based on previous experiments, but I was hoping I could maybe push it into the 150-250 Hz region as this would allow me to experiment with dedicated midrange drivers such as the 18M revelator(Which I'm very much in love with) or the like.
This was a long post but I jst wanted to give all the relevant information. Hope someone has a good take on how to integrate the woofers properly in the VCAD simulation.. Maybe i just need to rely on near field respones only?
Thanks in advance!
I have build these speakers 5 years ago. Its currently a "2 way speaker with build in subwoofers" as I'm crossing at 80 hz to the 10". There are TWO active 10" in each speaker - one on each side mounted back to back. The two woofers share the same volume. They are driven by a set of Hypex FA253's.
They sound great, especially the 20-100-ish Hz range/punch and rumble. Im very happy with that part. However, I belive I could do the integration of the tweeter and midwoofer better today. maybe also improving on the integration between the woofers and midrange.
I started using VCAD more since they were build and I think I am starting to get the hang of it. Therefore I was planning to redo all measurments of this speaker and get it simulated properly in VCAD. However, all other speakers i used VCAD for have crossovers in the 600+Hz range. So the crossover between the woofers and midrange in this speaker is challenging me.. add to that the dual side mounted configuration.
I am so locky that I can use a small anechoic chamber for the new measruements. The room is 2.4 x 4.6 x 2.0 m w/l/h (free space there are additional 3m of absorbtion all the way around that). No idea how deep in frequency this will let me go, but exited to see that. I use a Focusrite Scarlett 2i2 4th gen with loop back and REW for the measurements.
I have read the manual of VCAD from start to end as well as the REW specific help document (That one many times xD). However, Im still in doubt of how exactly to do the mesurments of especially my sidemounted woofers properly. From what I have read this is my best guess so far:
1. Tilt the speaker forward to make the baffel vertical
2. Place microphone 1 m from baffel at tweeter level
3. Measure tweeter 180 degree at 10 degree interval
4. Place mic at midrange level (stay 1 m from baffel)
5. Measure midrange 180 degree at 10 degree interval
6. Place mic a woofer height. Stay 1 m from baffel. Leave both woofers connected to amp.
7. Measure woofers 180 degree at 10 degree interval (this will mean they get significantly closer to the mic at the 180 degree position)
8. Do a measurement of both tweeter and midrange at 1 m from baffel placed at a Y value right between the two drivers.
9. Do a measurement of both midrange and woofers at 1 m from baffel placed at a Y value right between the two drivers.
10. Do near field measurements of midrange
11. Do near field measurement of woofers (with both woofers playing but one covered with pillows or whatever)
Then in VCAD I will set all drivers to Z=0 and adjust the Y value according to their location - relative to a vertical baffel. Then use the summation responses to adjust the delay of the drivers under the driver tab to ensure that the simulation matches the summed response.
Will this give a fairly decent representation of the woofers interaction with each other as well as the midrange driver?
I most likely will end up with a 80-is Hz crossover again based on previous experiments, but I was hoping I could maybe push it into the 150-250 Hz region as this would allow me to experiment with dedicated midrange drivers such as the 18M revelator(Which I'm very much in love with) or the like.
This was a long post but I jst wanted to give all the relevant information. Hope someone has a good take on how to integrate the woofers properly in the VCAD simulation.. Maybe i just need to rely on near field respones only?
Thanks in advance!
Member
Joined 2003
Steps 8 and 9 are not needed, loopback measurement includes timing so you don't need to fiddle with delay values separately. You may complete these steps as validation of offsets entered however, but if you measure correctly, summation should match with zero delay.
Since the baffle is sloped and you are tilting forward for measurement, Z offset and driver tilt should be entered in the crossover driver section. Z offset for the tweeter can be zero, and midrange it is the physical offset from baffle location on a slope, and tilt value is equal to the slope of the baffle. Woofers are entered with 9 deg rotation and Z offset from driver center to front of baffle.
Since the baffle is sloped and you are tilting forward for measurement, Z offset and driver tilt should be entered in the crossover driver section. Z offset for the tweeter can be zero, and midrange it is the physical offset from baffle location on a slope, and tilt value is equal to the slope of the baffle. Woofers are entered with 9 deg rotation and Z offset from driver center to front of baffle.
Yeah ste 8 and 9 was mostly also a validation.
Actually, the speakers are so small that the baffel is pointing at the listen position. So I intentionally wanted the simulation to be perpendicular to the baffel. That's why I wouldn't adjust z. Forgot to mention that..
Why 9 degree rotation on woofers? And rotated in which direction?
Actually, the speakers are so small that the baffel is pointing at the listen position. So I intentionally wanted the simulation to be perpendicular to the baffel. That's why I wouldn't adjust z. Forgot to mention that..
Why 9 degree rotation on woofers? And rotated in which direction?
I think 90 is ment, rotation to check the vcad documentation. The woofers are 90 degrees rotated viewer from the listening point.
Member
Joined 2003
Yes, 90 degree, sorry for the typo. Since you have 2 woofers, one rotated 90, the other -90, I believe X offset would be 1/2 of the front baffle width, and Z offset would be distance from the center of the driver on the side panel to the front panel edge. Since the operation is low frequencies with long wavelengths, the specific dimentions won't have a huge effect here, 90 degree rotation is the most important one.
Aah 90 makes sense... But, this is exactly what confuses me.. since they are sharing the volume I have to measure with both playing at the same time right?
At low frequencies the enclosure will be "invisible". So the woofers more or less act like one big Omni directional driver?
So if I measure 180 degree and then add two drivers in Vcad rotated 90 and -90 degrees I will get incorrect results I guess? The response may be fine but the volume will be too loud?
That's why I was thinking of just measuring with 0 deg being perpendicular to the baffel as the other drivers. Then the drivers would already be" rotated" in Vcad?
At low frequencies the enclosure will be "invisible". So the woofers more or less act like one big Omni directional driver?
So if I measure 180 degree and then add two drivers in Vcad rotated 90 and -90 degrees I will get incorrect results I guess? The response may be fine but the volume will be too loud?
That's why I was thinking of just measuring with 0 deg being perpendicular to the baffel as the other drivers. Then the drivers would already be" rotated" in Vcad?
Member
Joined 2003
Sure, treat the pair as a single driver is fine, especially since their operation is at low frequencies only. Mic pointed at the front speaker baffle as 0 deg angle, no rotation value necessary. Give it a go.
Trying to learn how to use VituixCad and find it a very hard start. Anyone knows good tutorials (text or video)?
I am aware of the manual, but it reads as an unconnected description of the elements.
Would be good to go through a process/workflow based one, maybe using example drivers/baffles/boxes/data files etc.
Youtube search returns limited results. 😶🌫️
I am aware of the manual, but it reads as an unconnected description of the elements.
Would be good to go through a process/workflow based one, maybe using example drivers/baffles/boxes/data files etc.
Youtube search returns limited results. 😶🌫️
It's actually very easy once you got the hang of it. As said the official tutorial on VituixCad is very helpful.