SpeakerTalk

[speakerguru]

Sorry but I'm going away for a few days and have no time to look at your simulation or give a full answer atm. However, the plot needs to be in dB, i.e. log not linear, in the vertical axis to make any sense of it.

[theozaurus]

I've been visually matching the simulation data with a Linkwitz calculator. So far the values look like:

f0 = 57
Q0 = 1.1
fp = 26
Qp = 0.42

These values look seem possible to me. This is looking at the BEC stage in isolation. The input stage seems to have a 3db boost up to around 200Hz, and the contour section does more alterations.

The values for the Z domain look like:

a0 = 1
a1 = 1.991917852130950
a2 = -0.991929388490413
b0 = 0.999353476592651
b1 = -1.991895897210170
b2 = 0.992597866818542
Fs = 48000

There are also some analogue coefficients available (which make no sense to me):

d0i = 128265.378796557
d1i = 325.58323864476
d2i = 1
c0i = 26687.4103005456
c1i = 388.959090444451
c2i = 1
gn = 95999.7639175758
cci = 9253321340.49686

I'd love to see if these seem broadly right.

[speakerguru]

see EQUATION 32-2 on page 595 of http://www.dspguide.com/CH32.PDF for the biquad transfer function with variable s.


For the steady state response you substitute jw for s. This means that your coefficient d01 is equal to (jw)squared, where w=2(pi)f.
Up to 4 or so figures your d0i is equal to [2x(pi)x 70] all squared. So 70Hz rather than 57Hz is probably correct. I doubt that KEF would have equalised such a small speaker down to 26Hz so that probably is not correct either. However, If you can trace out the schematic and get me some BEC (Friend circuit) component values I can calculate the transfer function for you.

My Biquad "app" will calculate s coefficients from f and q values and vice versa. PM me if you want a copy (Windows exe file only).

[theozaurus]

speakerguru: Are you referring to the 104ab? I'm planning on using this with my 104/2's.

The values for the friend circuit are the same as the ones from the KEF Kube 200 schematic. Perhaps my simulations are slightly out as I'm using an ideal op-amp for the model.

The graphs can be seen here: http://www.partsim.com/simulator/#15762

Clicking: Spice -> Run -> Run
Then going to the "AC analysis" tab

[theozaurus]

Having trouble sending a PM, keep getting the error:

Failed sending email :: PHP ::

DEBUG MODE

Line : 234
File : emailer.php

Is your app available to download anywhere?

[speakerguru]

Biquad sent.

For general info
the KUBE 200 BEC equalises a f=56 Hz Q=1.1 system down to f=26 Hz with a Q=0.45

[stelya]

Hi guys,
I have 107s but no KUBE and have few options:

MathLAB RoomEQ for Foobar2000 plugin
miniDSP https://www.minidsp.com/dirac-series/ddrc-22d
something else HW or SW based

miniDSP that i mentioned is all digital and goes between transport/PC and a DAC and it offers FR alteration in that domain. It also comes with a microphone with DIRAC room correction software
That is digital.There is analog miniDSP that i guess would go after the DAC and before the preamp?

Is there anything else on the market that would be a replacement of KUBE or an improvement?

given the similar price Id rather buy miniDSP and do room measurement and have added benefit of that, which I guess is not small.
Basically all I want to do is do what KUBE did... add around +6db 20-40Hz and leave upper frequencies untouched...without sacrificing the signal quality in any way.
Also...do it in the digital or analog domain?

Thanks!


corrected khz to Hz 22-11-19 SG

[Parts]

Many years ago I almost bought a Tact digital pre amp simply for its digital room correction with the thought of it being an ideal way to remove the Kube for my 107's.

I hesitated, watched the guys that formed that company split, then Lyndorf I think it was with pretty much the same thing was an option, but by then well confidence in the company went.

This was before I started a family, was also looking at replacing the 107 with Active ATC SCM50.

Should have spent before starting a family lol.