Budget active subwoofer lowpass #2, buy, eval, mod

[bansuri]

Hi,

Could someone please check the schematic? I hope the pot connection is alright.

Thank you,
VJ

On IC1b there is a resistor missing from +in to gnd for the bias current.
Then the input inverts

 

[DonVK]

Happy to lend a hand.

Yes, swap your IC2A and IC1B so that IC1 (A&B) has the "BW3 HP" and IC2 (A&B) has the "BW2+BW2 LP" .

 

[vj_audio]

I have made the changes.

 

[bansuri]

Why don't you use a quad pot also for the last filter
Potentiometer - 20KLin Quad
Bourns also has quad ones

 

[DonVK]

In the current design there are 2 * BW2 LP filters. The first BW2 is adjustable (stereo pot) and is meant to create a subwoofer LP rolloff slope to match the slope of a closed box woofer (2nd order HP slope). The second BW2 is just an upper limit cutoff. At cuttoff BW2 * BW2 = LR4.

If the system woofers were ported ( HP is ~BW4 ) then a quad pot would be needed to create an adjustable 4th order filter to make the sub LP rolloff match. Unless you want to mismatch the slopes.

 

[bansuri]

If the system woofers were ported ( HP is ~BW4 ) then a quad pot would be needed to create an adjustable 4th order filter to make the sub LP rolloff match. Unless you want to mismatch the slopes.

I do not know why you have to match the slopes of the low end of a speaker and the upper one.
Maybe you mean the slopes of the x-over should match, from sub to midrange but the low end causes only a small phase change several octaves up.

For a sealed box you better bypass the 3rd order highpass to avoid the additional group delay, unless you use a turntable where feedback can occur because of mechanical vibration from sub to TT.

If you use a 4th order bandpass you can bypass the second Low Pass for a proper LR4. This sufficiently suppresses leakage of higher frequencies through the vent. You also do not need the HP.

Only a vented Sub will need all filters, but with a 4-gang pot for LR4.

 

[DonVK]

... maybe a drawing is better. The attached pic shows my system filters.

I match the order and magnitude of the slopes at the XO points to get better (smoother) summation (mag and phase). At low freq our ears are not sensitive to group delay so it's not a concern for me (actually tested it). The subwoofer HP@20Hz is to prevent exceeding Xmax at low freq where a ported driver will be unloaded. The subwoofer LP@100Hz slope mag matches the closed box woofer HP@100Hz slope mag. That is the basis of the subwoofer bandpass filter mods in this thread.

 

[jimk04]

I just measured the R values of my board as discussed here and maybe unsurprisingly they differ from DonVKs example.
Some will be swapped out so they are a non issue but for R5, 9, 10 and 11 I have 10k versus DonVKs 8.2k. Should I bother changing mine to 8.2k or are they 'near enough.' Sorry but I do not understand their function or impact on the operation of the board.

 

[DonVK]

Not surprised there are changes as there are no build revision markings. No promises made, so none broken

The change to 10K lowers the Fmax from 130Hz to 107Hz. This should not be a problem, unless you want your sub to XO above 107Hz.

There is a good filter calculator at Filter Design and Analysis and you can select the OpAmp configuration, plug in values, press calculate and view the bode plot.

 

[jimk04]

Thanks for that reply.
I am happy with it as it is then...107Hz is fine.
Which of those active filters in the link you provided is the same as the one on our board? I can't seem to sussnout which one it is!

 

[DonVK]

All the filters are unity gain S-K (see pic) except the input stage. The components are placed a little differently but if you follow the nets, they are the same.