Linkwitz Transforms

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi,

Thanks for trying to explain, but I still don't get it. :scratch:
Most power amplifiers are designed as power operational amplifiers, taking useage of voltage feedback. The definition of an opamp only describes the requirements regarding the number of inputs and their impedance, the number of outputs and their impedance and gain and bandwidth of the "black box". Power level is not defined or restricted to line level.
Its the nature of such a op amp to try to steer its output, so that the inverting input is a mirror of the noninverting input and hence the difference between noninv and inv becomes 0. Both inputs are high impedance points. The inverting input may be regarded as virtual gnd, but it is not low impedance. In a current feedback opamp it would be low impedance though.

jauu
Calvin
 
I think what we are dealing with is a semantic issue.

As already pointed out the name operational amplifier was originally used to perform operations in analogue computers, the classic one uses five discrete transistors.

The two stage differential amplifier/common emitter configuration of the standard discrete op amp, found its way into audio because it is a simple way to convert single ended inputs to double ended two rail outputs and by a.c. coupling the feedback to ground it has 100% d.c. feedback so it has d.c. stability, other direct coupled topologies need a servo to achieve this.

Power amps and later integrated op amps added a complimentary emitter follower to the output, enabling larger output currents and less quiescent current.

So essentially the appellation op amp came about as a historical accident as they were extensively used in military electronics and scientific/ industrial instrumentation to do mathematical operations in analogue computers, that were used in control systems and by scientists. to solve differential equations.

It is true that so called current feedback op amps are not strictly op amps by the formal definition, but neither are audio op amps because they are intended to amplify a.c. signals and not dc ones.
rcw
 
Hi,

Thanks for trying to explain, but I still don't get it. :scratch:
Most power amplifiers are designed as power operational amplifiers, taking useage of voltage feedback. The definition of an opamp only describes the requirements regarding the number of inputs and their impedance, the number of outputs and their impedance and gain and bandwidth of the "black box". Power level is not defined or restricted to line level.
Its the nature of such a op amp to try to steer its output, so that the inverting input is a mirror of the noninverting input and hence the difference between noninv and inv becomes 0. Both inputs are high impedance points. The inverting input may be regarded as virtual gnd, but it is not low impedance. In a current feedback opamp it would be low impedance though.

jauu
Calvin


Hi,

The difference between the two inputs is zero therefore the inverting
input is exactly the same voltage as the the output of a unity gain
gain amplifier.

The impedance of the inverting input is the the two feedback
resistors in parallel, in a power amplifier typically 1K and 28K
so just under 1K is very typical.

Op-amps simply don't support the voltage swing of power
amplifiers, and even if they did they cannot provide the
current or power to drive speaker loads.

Op-amps comes from operational amplifiers and analogue
computers based on mathematical operations where the
only issue is accuracy and this doesn't need high power.

Power amplifiers are not remotely op-amps.

400px-Sallen-Key_Highpass_Example.svg.png


Above is a high pass Sallen/Key filter with a unity gain op-amp.

I've described how to use it with a power amplifier with gain.
(You move R1's connection to the output to the inverting input,
you should be able to work out why its effectively the same.)

And FWIW assuming an ideal op-amp for the above diagram
the inverting input is zero source impedance as its connected
directly to the output, so your waffling bigtime on details.

rgds, sreten.
 
Last edited:
Hallo everybody,
And my 5 cents in the topic.
May we use Mr. Linkwitz tansform circuit with front loaded horns.
I mean Klipsch-horn, La Scala, Black widow and other with sealed back chamber, which suffer of lack of low bass.
Are all that accounts valid if we have front loaded woofer?
Of course, I think and about biamping the sistem.
 
The "Linkwitz transform" only applies perfectly to a single resonance, for example a speaker driver in a sealed box. Horn loaded drivers are a very different critter, with their fundamental resonance very heavily damped (hopefully!) by the horn loading, and the horn's own raggedy low frequency response dominating.

It will of course do *something* but it won't be predictable from measured driver terminal parameters, and it may not be what you want.

All good fortune,
Chris
 
Hi again!
Here you are the parameters of my K-33.
Say my, because they are vintage, long time working, nobody knows how.
And their parameters are slightly different from other K33 parameters, known in the net. Square magnets.

K-33
Fs = 24.54 Hz
Re = 3.40 ohms[dc]
Qt = 0.23
Qes = 0.26
Qms = 2.26
Mms = 76.93 grams
Rms = 5.237921 kg/s
Cms = 0.000547 m/N
Vas = 561.92 liters
Sd= 855.30 cm^2
Bl = 12.430875 Tm
ETA = 3.06 %
Lp(2.83V/1m) = 100.68 dB


As you may see, this driver is not the best for closed box.
But I lock it in 100L enclosure.
Because it is not suitable for sub-bass, I accounted the transform circuit for 39Hz; Q=0.81; +8dB
But something is wrong, now I have a big drop in the neighborhood of 100Hz.
My configuration is biamping, with 300Hz Linkwitz X-over:
Sourse-> X-ower(in-buffer; X-ower; output-buffer) -> Linkwitz Transform -> Amp
So, where is the mistake?
May be, somewhere phase-change:confused:
And again, let me underline, K33 is not in horn. It is in 100L box. And without transform circuit, there is no 100Hz drop.
 
Even if you can only answer one, or part of one, question I would appreciate it.

An LT is an EQ for sealed subs. It goes in between your preamp and poweramp. It both attenuates and applies boost to your preamp's line level signal your subwoofer's response into the lower frequencies.

If you look at the attached pic below... if the blue trace is your sub's close mic'd response, then you would build your LT to apply the EQ of the red trace to achieve the green trace's response. The result at listening position is a less peaky response that extends lower in frequency. I've included a spreadsheet that you can use to help you design your LT if you weren't already aware of it.

This is a fairly advanced project and it might take a few breadboard attempts to nail it for your system. I recommend putting your sub together first, hooking it up and do a few close mic measurements with the mic on axis, 1" from the dustcap. Then take a few at the listening position to see what kind of roomgain you are working with to make sure that you don't apply too much gain. You can then use the spreadsheet's graph to overlay over your close mic'd response and adjust the parameters until the graph of the LT curve is a good mirror image to your sub's response.

The more gain you apply, the more amplifier power you will need and the more throw your sub will need to have. If you are planning this with a ported driver, I recommend putting at least a 2nd order highpass filter on the LT's output just below port tune so you don't bottom out and damage your driver. You can also look into this project: Sub-Woofer Controller for a ported sub as it has a safe rolloff downlow and will be somewhat adjustable.
 

Attachments

  • LT.png
    LT.png
    74.4 KB · Views: 116
  • linkxfrm Labs.zip
    36.5 KB · Views: 58
Last edited:
The LT is more flexible than what Shred describes.

The new Fp can be either higher, or lower, than the original Fo
The new Qp can be higher, or lower, than the original Qo

You'll be better measuring the driver in-cabinet impedance and using Mr. Linkiwtiz equations to get f and Q.
You can use the Linkwitz site where all the design equations are given for ALL the LT variations. But you NEED to know Fo & Qo to get started.

Shred's spreadsheet does all of these variations and does it very well and you still NEED Fo & Qo to get started !
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.