Suppose I build a D-amp, and elect to attach it very close to the input terminals of the speaker. The inductance of typical speakers is much greater than the recommended inductors for the output of the amp. Can I ditch the inductors and let the speaker be the inductor?
If the amp output is BTL (bridged) then the speaker won't see the squarewave current, as it's common to both wires. The speaker sees only the difference, which is mostly audio.
However the speaker cables will broadcast RF like crazy - that's why you want to be close.
As to how the switching output transistors will cope with the load, its hard to say. Some app notes recommend this setup, for small power - but single driver, no crossover. The crossover may present a difficult load for the RF output.
I've always wanted to try it, but never got 'round to it. Certainly someone here has tried it.
However the speaker cables will broadcast RF like crazy - that's why you want to be close.
As to how the switching output transistors will cope with the load, its hard to say. Some app notes recommend this setup, for small power - but single driver, no crossover. The crossover may present a difficult load for the RF output.
I've always wanted to try it, but never got 'round to it. Certainly someone here has tried it.
Hello
One thing to consider, especially for the non-BTL case is the increased power dissipation caused by RF heathing in the speaker. The voice coil is typically surrounded by a lot of plain metal parts, an environment eddy currents will relish.
It would probably be a good idea to measure the equivalent losses resistance at the switching frequency, and compare it with the DC resistance of the coil.
This will give an idea of the order of magnitude of the problem, and I suspect it won't be negligible.
LV
One thing to consider, especially for the non-BTL case is the increased power dissipation caused by RF heathing in the speaker. The voice coil is typically surrounded by a lot of plain metal parts, an environment eddy currents will relish.
It would probably be a good idea to measure the equivalent losses resistance at the switching frequency, and compare it with the DC resistance of the coil.
This will give an idea of the order of magnitude of the problem, and I suspect it won't be negligible.
LV
Hi!
Does somebody have scheme of the layout that is added in RAR on first page?Maybe author of the PCB.I wuld be realy grateful.
Thank you.
Does somebody have scheme of the layout that is added in RAR on first page?Maybe author of the PCB.I wuld be realy grateful.
Thank you.
fdeck said:
Absolutely not! The inductor and capacitor form a low pass filter which is utterly critical to the operation of class-d. If you remove it your speaker will be subjected to square waves and not a lot else.
Filterless class D is only useful for low powers (low voltage) and small speakers. For example, 12Vpp at 250Khz square wave could be applied to a small 0.5 inch voice coil without much problem, but not 180Vpp to a large 3 inch woofer voice coil. Passive crossovers are not suitable at all for filterless drive either.
If the amplifier is BTL, EMI depends on modulation approach. Standard Class AD (both halves complementarily driven) produces more differential mode EMI but no common mode EMI when idle, so it's the desirable approach. Class BD drives both halves in phase and cancels differential mode EMI at the expense of the highest common mode EMI so it should be always avoided (it's one example of marketing driven engineering).
If the amplifier is BTL, EMI depends on modulation approach. Standard Class AD (both halves complementarily driven) produces more differential mode EMI but no common mode EMI when idle, so it's the desirable approach. Class BD drives both halves in phase and cancels differential mode EMI at the expense of the highest common mode EMI so it should be always avoided (it's one example of marketing driven engineering).
- Status
- Not open for further replies.