Measured frequency response of FR125S in sealed 9 litres enclosure stuffed moderatly and filterless:
As you can see, it's faily erratic above 1kHz (+/- 6db up to 20k) so I was curious what filters some of you have been using, and any measurements (with or without filters) would also be helpful.
An externally hosted image should be here but it was not working when we last tested it.
As you can see, it's faily erratic above 1kHz (+/- 6db up to 20k) so I was curious what filters some of you have been using, and any measurements (with or without filters) would also be helpful.
HobbyHiFi 1/2006
Re=7.1 Ohm
Le=0.39mH
Fs=44Hz
Qm=3.6
Qe=0.66
Qt=0.56
Sd=50cm2
Vas=8.7l
Cm=2.4
Mm=5.3g
Rm=0.41
BL=4.0
Re=7.1 Ohm
Le=0.39mH
Fs=44Hz
Qm=3.6
Qe=0.66
Qt=0.56
Sd=50cm2
Vas=8.7l
Cm=2.4
Mm=5.3g
Rm=0.41
BL=4.0
An externally hosted image should be here but it was not working when we last tested it.
A notch for the 1,5kHz bump might be one. I doubt the 3,5kHz and 6,5kHz are that bad. Mmm, maybe I should finally get a pair of those FR's, something to play with for the holidays.
I wonder if anyone has tried the FR/WR with this ScanSpeak aperiodic vent:
http://www.madisound.com/cgi-bin/index.cgi?cart_id=1334259.11089&pid=1189
I wonder if anyone has tried the FR/WR with this ScanSpeak aperiodic vent:
http://www.madisound.com/cgi-bin/index.cgi?cart_id=1334259.11089&pid=1189
Re: Re: FR125S Measurements and Filters
I've overlayed my graph to the same scale and it is very similar. The 2-4k anomaly is baffle diffraction rearing its ugly face I think.
I've overlayed my graph to the same scale and it is very similar. The 2-4k anomaly is baffle diffraction rearing its ugly face I think.
An externally hosted image should be here but it was not working when we last tested it.
I'm unsure if baffle step shows up with this technique: The measurements are combined nearfield and gated-on-axis spliced at 300Hz.Landroval said:
Hi Folks,
If... I had the FR125 (or WR) I would try the closed box with highpass capacitor. This allows a small box and good low frequency performance, including protection against too low frequencies.
I simulated the 9 liter closed box in AJ Horn (Demo) which is quite close to the measurements presented (red line). Then the simulation in a 6 liter closed box with a 400 uF in series with the speaker (black line).
Good and close enough is say a 330 uF bipolar rough, a 68 uF Bipolar Smooth and a 10 uF MKP.
Rick
If... I had the FR125 (or WR) I would try the closed box with highpass capacitor. This allows a small box and good low frequency performance, including protection against too low frequencies.
I simulated the 9 liter closed box in AJ Horn (Demo) which is quite close to the measurements presented (red line). Then the simulation in a 6 liter closed box with a 400 uF in series with the speaker (black line).
Good and close enough is say a 330 uF bipolar rough, a 68 uF Bipolar Smooth and a 10 uF MKP.
Rick
Attachments
Vikash:
I am not sure how/where you have the driver mounted on the baffle or if you are using rounded edges on your baffle, but I would bet that much of the uneveness in output is resulting from baffle edge diffraction . . . You could try to use the Tolvin (sp?) edge diffraction simulator to find an "optimal" mounting point for the driver. Also, if your driver is surface mounted and not recess mounted, your results will also vary . . .
RickNL:
I think your curves don't look right for a sealed solution. The black line in your graph approximates what I simulate in WinISD for the FR in a .5 cubic foot ported enclosure tuned to around 50 hz. Also, the sealed solutions all seem to start rolling off just above 100 hz and usually have an F3 around 90 hz . . . Try downloading WinISD and using that - most people seem to like it for basic simulations and it is a free download.
Regards,
Andrew
I am not sure how/where you have the driver mounted on the baffle or if you are using rounded edges on your baffle, but I would bet that much of the uneveness in output is resulting from baffle edge diffraction . . . You could try to use the Tolvin (sp?) edge diffraction simulator to find an "optimal" mounting point for the driver. Also, if your driver is surface mounted and not recess mounted, your results will also vary . . .
RickNL:
I think your curves don't look right for a sealed solution. The black line in your graph approximates what I simulate in WinISD for the FR in a .5 cubic foot ported enclosure tuned to around 50 hz. Also, the sealed solutions all seem to start rolling off just above 100 hz and usually have an F3 around 90 hz . . . Try downloading WinISD and using that - most people seem to like it for basic simulations and it is a free download.
Regards,
Andrew
I am not sure what the impact is of using 45 degree angles on the edge versus a rounded edge - seems like the 3/4" rounded edge has a strong following . . . I can't tell from the photos if you've already done this or not, but I know that many people also use the "golden ratio" to determine driver vertical and horizontal placement on the baffle to help smooth the response . . . It looks like you are already pretty far down the path of building those enclosures but you might want to get some recommendations on what else you could do to help improve that response curve. You might also try using this simulator:
link
Andrew
link
Andrew
Hi ABS
Yes the normal closed boxes with WR/FR125 end up around 90 Hz . Ans Yes I do have and use WinISD for reflex and closed enclosures.
But the 3rd order alignment with a series capacitor, can not be simulated with WinISD. The Demo version of AJ Horn is the only package I know that does this.
The 3rd order alignment works fine with chassis that have a rather high Qms and Qts with a low fr. Then a smaller enclosure than the standard Q=0.7 closed box gives a lower cut off frequency. Also the power handling is increased as there is a limiting for the subsonic input to the chassis through the capacitor. The impedance plot gives the explanation for the behavior. The sonics of this 3rd order alignment is also between the vented and the closed box. And for highish Q chassis often better than vented.
Rick
Yes the normal closed boxes with WR/FR125 end up around 90 Hz . Ans Yes I do have and use WinISD for reflex and closed enclosures.
But the 3rd order alignment with a series capacitor, can not be simulated with WinISD. The Demo version of AJ Horn is the only package I know that does this.
The 3rd order alignment works fine with chassis that have a rather high Qms and Qts with a low fr. Then a smaller enclosure than the standard Q=0.7 closed box gives a lower cut off frequency. Also the power handling is increased as there is a limiting for the subsonic input to the chassis through the capacitor. The impedance plot gives the explanation for the behavior. The sonics of this 3rd order alignment is also between the vented and the closed box. And for highish Q chassis often better than vented.
Rick
Attachments
Hi Andrew,
There is not much room to play with driver positioning on a narrow baffle, but vertically it is placed off centre. The FR characteristics are consistent above 5kHz with HobbyHifi measuremements. These appear to be diver break-up modes. The peak above 20k is also consistent with my extended measurements.
I'm going to play with the stuffing a bit, and try some crude experiments of extending the baffle to confirm the 2k and 4k dips are baffle related.
Contrary to what the measurments might imply, they don't sound nearly as jagged to me in-room. Perhaps even overdamped in the midrange which I'm addressing at the moment. The bass and overall level balance throughout the SPL range is astounding and huge pluses.
There is not much room to play with driver positioning on a narrow baffle, but vertically it is placed off centre. The FR characteristics are consistent above 5kHz with HobbyHifi measuremements. These appear to be diver break-up modes. The peak above 20k is also consistent with my extended measurements.
I'm going to play with the stuffing a bit, and try some crude experiments of extending the baffle to confirm the 2k and 4k dips are baffle related.
Contrary to what the measurments might imply, they don't sound nearly as jagged to me in-room. Perhaps even overdamped in the midrange which I'm addressing at the moment. The bass and overall level balance throughout the SPL range is astounding and huge pluses.
RickNL:
Okay, now I am tracking with you - in your first post it seemed like you were talking about sealed and not ported boxes although I agree that the inline capacitor will provide the benefits you describe in either case. The only issues I see with this approach are the phase shifts created by the capacitor and the fact that the entire frequency range will now be running through that capacitor which means that the capacitor now becomes a critical component to the overall sound of the system. As a side note, I have tried similar simulations as this in WinISD - there is a tab in each project called EQ/Filter where you can add a 1st order (or higher) high pass filter of various types at some frequency if you wish . . .
Vikash:
Sometimes even small shifts side to side can make a difference in baffle diffraction - did you take a look at that simulator I linked you to? One thing I noted with the WR design I did is that a lot of mid and high frequencies can bounce off the rear of the cabinet and back out the cone of the woofer as well as out of the port. It may be that some of what you are seeing are reflections from inside the cabinet. If you haven't tried it already, you might consider putting some type of damping material loosely stuffed right behind the woofer and see if that helps at all.
IIRC the WR wasn't really showing any breakup modes until around 10K in my testing. Also, from the waterfall plots I've seen of the FR, there is very little energy storage in the upper octaves - that seems inconsistent to me with a "break up mode".
I am not familiar with your test software, but what kind of "gate" and "resolution" are you using when you do the response testing? If the gate is too large you could even be picking up reflections from the floor, ceiling or walls in your room.
Andrew
Okay, now I am tracking with you - in your first post it seemed like you were talking about sealed and not ported boxes although I agree that the inline capacitor will provide the benefits you describe in either case. The only issues I see with this approach are the phase shifts created by the capacitor and the fact that the entire frequency range will now be running through that capacitor which means that the capacitor now becomes a critical component to the overall sound of the system. As a side note, I have tried similar simulations as this in WinISD - there is a tab in each project called EQ/Filter where you can add a 1st order (or higher) high pass filter of various types at some frequency if you wish . . .
Vikash:
Sometimes even small shifts side to side can make a difference in baffle diffraction - did you take a look at that simulator I linked you to? One thing I noted with the WR design I did is that a lot of mid and high frequencies can bounce off the rear of the cabinet and back out the cone of the woofer as well as out of the port. It may be that some of what you are seeing are reflections from inside the cabinet. If you haven't tried it already, you might consider putting some type of damping material loosely stuffed right behind the woofer and see if that helps at all.
IIRC the WR wasn't really showing any breakup modes until around 10K in my testing. Also, from the waterfall plots I've seen of the FR, there is very little energy storage in the upper octaves - that seems inconsistent to me with a "break up mode".
I am not familiar with your test software, but what kind of "gate" and "resolution" are you using when you do the response testing? If the gate is too large you could even be picking up reflections from the floor, ceiling or walls in your room.
Andrew
With a 1/3rd octave smoothing, it looks like the driver would be +/- 3 dB from 60 Hz to 20 kHz. Pretty flat to start with... And they sound pretty neutral to me, at least. A lot flatter and more neutral than most full range drivers.
If you wanted to filter anything, I'd deal with the rise centered at 1.5 kHz - a Q=0.7, Fc=1500 Hz, G=-2 dB filter should smooth that right out.
Dan Wiggins
Adire Audio®
If you wanted to filter anything, I'd deal with the rise centered at 1.5 kHz - a Q=0.7, Fc=1500 Hz, G=-2 dB filter should smooth that right out.
Dan Wiggins
Adire Audio®
Andrew, I have seen the diffraction simulator thanks. In fact I have some very powerful software to hand, but just haven't used it for this design. The cabinet is already medium-heavy damped with polyfil.
Perhaps my terminology is wrong, and they're not energy storage issues, but those peaks and dips are consistent nevertheless.
I use Speaker Workshop for measurements. Nearfield taken to 300Hz and then gated on-axis up to 20kHz. Gating window is to 7ms which is when the first reflection occurs.
Dan, I absolutely agree that they sound natural
I'm questioning my measuring just as much since I'm quite new to this. My measurements are smoothed to 1/16 octave.
If others are smoothing to 1/3, then that explains a lot!
I'm sure I could get +/- 1 db if I use the right smoothing
Perhaps my terminology is wrong, and they're not energy storage issues, but those peaks and dips are consistent nevertheless.
I use Speaker Workshop for measurements. Nearfield taken to 300Hz and then gated on-axis up to 20kHz. Gating window is to 7ms which is when the first reflection occurs.
Dan, I absolutely agree that they sound natural
I'm questioning my measuring just as much since I'm quite new to this. My measurements are smoothed to 1/16 octave.If others are smoothing to 1/3, then that explains a lot!
I'm sure I could get +/- 1 db if I use the right smoothing - Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.