Q&A - LM4780 type as bi-amped speaker driver

[FastEddy]

referring to:
http://www.national.com/images/pf/LM4780/20058622.pdf
http://www.national.com/pf/LM/LM4780.html

I suppose one could simply insert the crossover network in the negative feedback of each channels ... thus producing a two way speaker bi-amp ... What, what?

This chip has decent specs ... probably does not need the mute circuit, and +/- 35 VDC power supply might do it nicely for a "stand alone" 2-way box with healthy drivers.

I wonder if there is a balanced input version?

[jackinnj]

as the LM4780 is an opamp you "may" insert the capacitors as reactive elements for low or high-pass filter configuration, but I don't think it is going to be particularly stable.

you can configure an LM4780 as a differential driver.

If you download Texas Instruments Filter-Pro software there is a configuration for fully differential filters.

[FastEddy]

Yes, I remember now ... I made a gopher chaser once many years ago by using a low power op-amp with a notch filter in that feed back loop so it went into oscillation on start up, hooked that to a drive transistor and drove a motorola peizo-electric tweeter up in the 25k to 30k range, mounted the whole thing on a piece of 8 foot x 3/4" EMT and stuck it in the middle of the garden ... it got rid of the gophers ... and the ants too ... and the dog didn't seem to mind it too much ... wonderful toys, these op-amps.

[consort_ee_um]

If ypu look at this data sheet page 10 there is what you describe.
http://us.st.com/stonline/products/l...re/ds/1459.pdf
You could substitute one of the LM4780 amps for each of the TDA2039A amps
It is a 3 way design but the woofer frequency cut-off frequency could be made higher to make a 2 way design
Unfortunately real world cross-over designs rarely conform to text-book type active filter circuits, but it's a good starting point

[jackinnj]

Originally posted by consort_ee_um

Unfortunately real world cross-over designs rarely conform to text-book type active filter circuits, but it's a good starting point

If you are careful about the layout, use components with closely matched specs and choose the right opamp they actually "conform" pretty well to the equations. pick the wrong opamp and you're toast.

Where you'll get into trouble with using a high-power chipamp is when they start to run out of gain -- the GBW for the LM4780 is 2 to 8 MHz so it shouldn't be a problem. The TDA2030 is 100kHz so shouldn't be used as a high power filter. The early Bob Widlar LM12 didn't even specify in its data sheet. Remember that you are driving a reactive load, not a purely resistive load.

[consort_ee_um]

Jackinnj,
I agree that if you build a active filter correctly it will give you the desired response. What I was trying to say that if you design a standard active Linkwitz-Riley 4th order active cross over it may not give optimum results because:
a) the responses of the drive units have to be flat for an octave beyond the xover frequency
b) the drive units have to have minimum phase for an octave beyond the xover frequency
c) The acoustic centres of the drive units need to be coincident
Such conditions do not happen often in the real world.

What I would do these days is to design a passive crossover and then convert the design to active, since it is easier to tweak a passive crossover and there are tools like the excellent passive crossover designer spreadsheet to help you
That tool showed me that you can get a 4th order Linkwitz-Riley response from a 2nd order crossover filter, since the particular drive unit response fell off with frequency.