I am starting to look at modifying my existing speakers.
I plan to try sealing the two ports to convert the vented bass/mid response to an upper bass/mid unit.
My plan is to add a low bass only speaker with an LR4 low pass. This is currently reproducing well.
And add an active high pass filter to augment the sealed roll-off to give the matching LR4 that I would need.
I can create the appropriate Q & F in the active filter to match the Q & F of the sealed box, if I knew what they will be.
I measure the new (sealed) resonant frequency (Fr).
What does this tell me?
I know the roll-off is high pass 2pole, i.e. 12dB/oct.
What is the shape of that 2pole roll-off?
Is it Butterworth Q=1/sqrt(2)? or something else?
What is the frequency of that sealed roll-off?
Is F-3dB = Fr (that I measured earlier)?
I plan to try sealing the two ports to convert the vented bass/mid response to an upper bass/mid unit.
My plan is to add a low bass only speaker with an LR4 low pass. This is currently reproducing well.
And add an active high pass filter to augment the sealed roll-off to give the matching LR4 that I would need.
I can create the appropriate Q & F in the active filter to match the Q & F of the sealed box, if I knew what they will be.
I measure the new (sealed) resonant frequency (Fr).
What does this tell me?
I know the roll-off is high pass 2pole, i.e. 12dB/oct.
What is the shape of that 2pole roll-off?
Is it Butterworth Q=1/sqrt(2)? or something else?
What is the frequency of that sealed roll-off?
Is F-3dB = Fr (that I measured earlier)?
You can find the Qtc and Fc of your newly sealed box in exactly the same way as you measure a driver for Qts and Fs. There is a guide on my website if you are unsure how to measure. You only need to know Qtc and Fc to design your filters.
Thanks.
I can manage that, I have the gear.
If the Qtc is highish, say around 1, then I am going to need a very low Q for the additional filter. Maybe Q<0.5 to achieve the LR4 roll-off.
I'll come back on that, once the soon to be sealed speaker is tested.
I can manage that, I have the gear.
If the Qtc is highish, say around 1, then I am going to need a very low Q for the additional filter. Maybe Q<0.5 to achieve the LR4 roll-off.
I'll come back on that, once the soon to be sealed speaker is tested.
Qtc is likely to be well below 1 - as the box will become oversized, and lower Q drivers tend to be fitted to reflex cabinets.
If Qtc was 1, your filter would need Q of 0.5 for Linkwitz-Riley response.
If Qtc was 1, your filter would need Q of 0.5 for Linkwitz-Riley response.
findings for sealed box speaker:
Fs 69.5Hz
Fh 110.2Hz
Fl 39.96Hz
Qtc or Qts? 0.372
Required filter Q is 1.344
Do I set the filter F the same 69.5Hz?
This should result in LR4 of F-6dB @ 69.5Hz
Is that right?
Your xls would benefit from a little calculator added to the first page.
Using 10r test resistor: input the mVac measured and the actual test resistor value.
Then using the predicted Rf, convert to the mVac required to find the two frequencies Fh and Fl. S
This would save having to use a calculator to calculate the mVac and for some, a bit of confusion on how to set Fh and Fl so that the two frequencies can be measured.
Fs 69.5Hz
Fh 110.2Hz
Fl 39.96Hz
Qtc or Qts? 0.372
Required filter Q is 1.344
Do I set the filter F the same 69.5Hz?
This should result in LR4 of F-6dB @ 69.5Hz
Is that right?
Your xls would benefit from a little calculator added to the first page.
Using 10r test resistor: input the mVac measured and the actual test resistor value.
Then using the predicted Rf, convert to the mVac required to find the two frequencies Fh and Fl. S
This would save having to use a calculator to calculate the mVac and for some, a bit of confusion on how to set Fh and Fl so that the two frequencies can be measured.
Last edited:
Your filter Q calculation is correct.
Yes, you set the filter Fresonance the same as your box Fc. I have a program to work out values somewhere if you need one. The high-pass filter designer on my website works on a desired F3 rather than Fresonance.
Thanks for the comments, good suggestions and I'll look to add them into the spreadsheet.
Yes, you set the filter Fresonance the same as your box Fc. I have a program to work out values somewhere if you need one. The high-pass filter designer on my website works on a desired F3 rather than Fresonance.
Thanks for the comments, good suggestions and I'll look to add them into the spreadsheet.
It is pretty lucky that for most ported speakers when you seal the port, work out darn close to being decent satellites for use with a sub. Usually somewhere between 60 and 80- F3. Plug em up and see what you get before you decide you need to add eq.
Is there a reason you want to seal the cabinets when you add the LF speaker? The cabinet only effects the bass roll-off. The crossover will do that for you. Now, the lucky trick is a sealed box in the crossover region gives you a second order slope, ao add an AVR crossover that is usually second order and it will match to a fourth order on the sub nicely. Usually. Sometimes, If you are lucky.
BTY, I am very fond of low Q sealed cabinets. I find they are easy to blend into the room and overall forgiving.
Is there a reason you want to seal the cabinets when you add the LF speaker? The cabinet only effects the bass roll-off. The crossover will do that for you. Now, the lucky trick is a sealed box in the crossover region gives you a second order slope, ao add an AVR crossover that is usually second order and it will match to a fourth order on the sub nicely. Usually. Sometimes, If you are lucky.
BTY, I am very fond of low Q sealed cabinets. I find they are easy to blend into the room and overall forgiving.
sealing the box will raise the LF power handling of the little speaker.
Adding and active High Pass will reduce the signal sent to the little box. This too will aloow high SPL.
The combination with support from a bass only speaker should end up sounding better.
I remember when I had been listening to the AE1 for a while and went back to the big Tannoy that I had been missing the bass slam.
So I added a passive low pass to a separate amp to drive the big Tannoy as a bass only support for the AE1. This was my first foray into active speakers.
I now have a range of big Tannoys to use and a range of amps that are the start of an active system.
The activation of the AE1 is something I think worth trying.
I managed to cook the coil of another AE1 after an evening of louder music listening, on my own, no alcohol and no party. I don't want to blow another. So conversion to an upper bass may get me closer to what I want. I had a 2pole 150Hz high pass on the cooked AE1 and it could not cope with that.
Adding and active High Pass will reduce the signal sent to the little box. This too will aloow high SPL.
The combination with support from a bass only speaker should end up sounding better.
I remember when I had been listening to the AE1 for a while and went back to the big Tannoy that I had been missing the bass slam.
So I added a passive low pass to a separate amp to drive the big Tannoy as a bass only support for the AE1. This was my first foray into active speakers.
I now have a range of big Tannoys to use and a range of amps that are the start of an active system.
The activation of the AE1 is something I think worth trying.
I managed to cook the coil of another AE1 after an evening of louder music listening, on my own, no alcohol and no party. I don't want to blow another. So conversion to an upper bass may get me closer to what I want. I had a 2pole 150Hz high pass on the cooked AE1 and it could not cope with that.
I already have a T/S spreadsheet that I developed for vented speaker design.
It never occurred to me to use it to test a sealed box speaker. I have never designed a sealed box speaker before. All my speaker work is a development from "Bullock on Boxes"
If you have already cooked a speaker with 150Hz 2nd-order high pass then I don't think you plan is going to work, as the corner frequency will be lowered thereby allowing more LF energy to your little speaker.
I don't get how you are going to use your vented box spreadsheet to test a sealed design, unless you are planning to set Fb to something ultra low like 1Hz?
My offer of a program was to choose component values for the filter, you just need to input Q and F which you have already obtained.
I don't get how you are going to use your vented box spreadsheet to test a sealed design, unless you are planning to set Fb to something ultra low like 1Hz?
My offer of a program was to choose component values for the filter, you just need to input Q and F which you have already obtained.
I already have a variety of active filter programmes and recently stumbled on one that predicts the Q value after inputing the component values.
I will try the simple 2pole @ 69.5Hz and see how it works.
If that is successful, then I may load the AE1 to reduce internal volume to get the final Fs upto around 100 to 150Hz and Qtc nearer 0.7, but this time with a 4pole.
If that is successful, then I may load the AE1 to reduce internal volume to get the final Fs upto around 100 to 150Hz and Qtc nearer 0.7, but this time with a 4pole.
Hi Andrew,
If I understand you correctly, you want to seal the enclosure to obtain a HP2 acoustic response and then add an active HP2 filter that will give you an overall acoustic LR4 response to your bass/mid drivers.
Unfortunately, it is only possible to do this if both responses, the sealed enclosure response and the active filter response, are 2nd order Butterworth (BW2) with both responses having the same Fs and Q.
I hope the attached two graphs will help clarify this. In graph HP4.JPG, you can see that I have entered the data you gave in your post #5 into a simulation program, which gave me your two filter response curves, which I then cascaded as shown in the Green curve.
If you compare this Green curve with the Red curve showing the cascaded response of the two BX2 filters in the attached LR4.JPG graph, you will see that the magnitude and phase (phase not shown) of these two cascaded responses are only the same at Fs and at the two frequency extremes either side of Fs. So your sealed enclosure + active HP2 will not sum to a flat response with your sub at frequencies close to Fs.
What this all means is that an LR4 response is uniquely defined by two identical BW2 filters in cascade and cannot be constructed by any other pair of 2nd order responses.
However, all is not lost, because there are a few other ways that can work, but in all cases you will have to force the sealed enclosure response into a BW2 response first and then add your active BW2 filter.
One way of forcing the sealed enclosure response into a BW2 response is to add a suitable peaking filter (BP type) to boost your Q from 0.372 to 0.7071 at Fs. You can then add an active BW2, which will give you an overall LR4 response. If you have, for example, a miniDSP, this could be implemented quite easily with only 2 biquads. Implementation in the analog domain is doable as well. You will need to accurately determine the coefficients of the peaking filter in order to force the sealed enclosure response into a BW2. Another way is to use a Linkwitz transformer to force the sealed enclosure response to a BW2.
Yet another approach which may work is to add resistance in series with the driver which will increase the Q. The drawback with this is obviously the loss of power if implemented passively. But if you implement it in an active form, you won’t loose power. After this you just add a BW2 to get a LR4 response. Hope this helps.
Regards
Peter
If I understand you correctly, you want to seal the enclosure to obtain a HP2 acoustic response and then add an active HP2 filter that will give you an overall acoustic LR4 response to your bass/mid drivers.
Unfortunately, it is only possible to do this if both responses, the sealed enclosure response and the active filter response, are 2nd order Butterworth (BW2) with both responses having the same Fs and Q.
I hope the attached two graphs will help clarify this. In graph HP4.JPG, you can see that I have entered the data you gave in your post #5 into a simulation program, which gave me your two filter response curves, which I then cascaded as shown in the Green curve.
If you compare this Green curve with the Red curve showing the cascaded response of the two BX2 filters in the attached LR4.JPG graph, you will see that the magnitude and phase (phase not shown) of these two cascaded responses are only the same at Fs and at the two frequency extremes either side of Fs. So your sealed enclosure + active HP2 will not sum to a flat response with your sub at frequencies close to Fs.
What this all means is that an LR4 response is uniquely defined by two identical BW2 filters in cascade and cannot be constructed by any other pair of 2nd order responses.
However, all is not lost, because there are a few other ways that can work, but in all cases you will have to force the sealed enclosure response into a BW2 response first and then add your active BW2 filter.
One way of forcing the sealed enclosure response into a BW2 response is to add a suitable peaking filter (BP type) to boost your Q from 0.372 to 0.7071 at Fs. You can then add an active BW2, which will give you an overall LR4 response. If you have, for example, a miniDSP, this could be implemented quite easily with only 2 biquads. Implementation in the analog domain is doable as well. You will need to accurately determine the coefficients of the peaking filter in order to force the sealed enclosure response into a BW2. Another way is to use a Linkwitz transformer to force the sealed enclosure response to a BW2.
Yet another approach which may work is to add resistance in series with the driver which will increase the Q. The drawback with this is obviously the loss of power if implemented passively. But if you implement it in an active form, you won’t loose power. After this you just add a BW2 to get a LR4 response. Hope this helps.
Regards
Peter
Attachments
How strange, I have used the method Andrew proposes several times in the past and arrived at a LR4 reponse. I'll have to dig out some response graphs, I wonder if there's something we are doing differently?
Acoustic Energy AE1 loudspeaker 1992 Measurements | Stereophile.com
Hi,
Some measurements of the AE1 in the above link.
FWIW doing a x/o at around 70Hz you may as well
throw the textbooks out of the window as room
modes will severely complicate the issues.
rgds, sreten.
How did you get that number ?
Hi,
Some measurements of the AE1 in the above link.
FWIW doing a x/o at around 70Hz you may as well
throw the textbooks out of the window as room
modes will severely complicate the issues.
rgds, sreten.
How did you get that number ?
Last edited:
Hi Richie00boy,
You are not alone in not knowing this. There was a recent thread on the same subject that I think Speakerdave commented on extensively. IIRC, the thread extended for quite a while, so it should be possible to track it down. I suggest that you dust off the old simulator and check it all out using Andrew’s data. If I find the time, I’ll post some simulations showing the summation of the HP and LP LR4 Xover with misaligned filter Q’s.
Regards
Peter
You are not alone in not knowing this. There was a recent thread on the same subject that I think Speakerdave commented on extensively. IIRC, the thread extended for quite a while, so it should be possible to track it down. I suggest that you dust off the old simulator and check it all out using Andrew’s data. If I find the time, I’ll post some simulations showing the summation of the HP and LP LR4 Xover with misaligned filter Q’s.
Regards
Peter
Hi PLB
I see two alternatives.
Leave the cabinet as is, i.e. vented and move the 4pole active filter up to somewhere around 100Hz to 150Hz so that I have a "near enough" acoustic LR4. This is similar to what I did a few years ago when I adopted the active BW2 @ 150Hz.
or
keep experimenting with the sealed modification and add an inert volume inside the box to raise the measured Q to 0.71 and thus simulate a Butterworth using the acoustic roll-off. Then add an active BW2 at the new measured high frequency of the small box volume.
To be honest the former is the preferred if it works. It is simply 3 B1 buffers with the BW2 active filters around two of them (giving an electrical LR4). I already have the vero board version under test.
What do you think?
Anybody else?
I see two alternatives.
Leave the cabinet as is, i.e. vented and move the 4pole active filter up to somewhere around 100Hz to 150Hz so that I have a "near enough" acoustic LR4. This is similar to what I did a few years ago when I adopted the active BW2 @ 150Hz.
or
keep experimenting with the sealed modification and add an inert volume inside the box to raise the measured Q to 0.71 and thus simulate a Butterworth using the acoustic roll-off. Then add an active BW2 at the new measured high frequency of the small box volume.
To be honest the former is the preferred if it works. It is simply 3 B1 buffers with the BW2 active filters around two of them (giving an electrical LR4). I already have the vero board version under test.
What do you think?
Anybody else?
Leaving the box as is and adding electrical 4LR @ 150Hz filtering will give the most protection to your voicecoil.
Maybe I could shed some light but honestly I don't
quite understand what is it that worries you.
Do you own measuring equipment ?
- Status
- Not open for further replies.