The less inductive the better.
So long as power is enough they should be fine.
So long as power is enough they should be fine.
Yes, and after this you put a big and expensive inductor at the output to protect the amplifier from capacitive load.
The source resistors inductance will slow down changes at amp output which you don't want.
The output inductor is usually just a few uH. I wouldn't call that big.
The Maplin lateral mosfet amp had enamelled copper wire wrapped around a resistor and was very cheap.
I have told you a billion times not to exaggerate !
The output inductor is usually just a few uH. I wouldn't call that big.
The Maplin lateral mosfet amp had enamelled copper wire wrapped around a resistor and was very cheap.
I have told you a billion times not to exaggerate !
In fact, you are the one exaggerating by recommending non-inductive resistors under the conditions in which you agree with the use of inductors at the output of the amplifiers. The inductance of a wire-wounds resistor is much smaller than the inductance that is usually added to the output of the amplifiers.
It depends how wire wound resistor is wound.
If its wound back on itself the inductance is negligible.
If its wound back on itself the inductance is negligible.
The coil used at the output of amplifiers are around 2-5µH.
Do you measured the inductance of a ,let's say, 0.1Ω/5W resistor?
I just measured one resistor like this and between 100Hz and 100Khz de inductance decrease from ~80nH to ~30nH. Measurement made with a VNA with 201 points, 10hz receiver bandwidth and 2.8Vpp signal amplitude.
For those who don't realize, the inductance of the resistor is at least 25 times lower in the worst conditions (2µH coil and 80nH resistor inductance).
Do you measured the inductance of a ,let's say, 0.1Ω/5W resistor?
I just measured one resistor like this and between 100Hz and 100Khz de inductance decrease from ~80nH to ~30nH. Measurement made with a VNA with 201 points, 10hz receiver bandwidth and 2.8Vpp signal amplitude.
For those who don't realize, the inductance of the resistor is at least 25 times lower in the worst conditions (2µH coil and 80nH resistor inductance).
Last edited:
But the source resistor inductance still affects HF response as its in the feedback path.
I recommend using a wire wound resistor wound back on itself to cancel out inductance.
At 100KHz even 80nH is 19 ohms ! And we have a 4 ohm amp !
I recommend using a wire wound resistor wound back on itself to cancel out inductance.
At 100KHz even 80nH is 19 ohms ! And we have a 4 ohm amp !
And the milihenri inductance of the speaker how big equivalent resistance will have at 100Khz?
You don't see the forest from the trees.
You don't see the forest from the trees.
And the milihenri inductance of the speaker how big equivalent resistance will have at 100Khz?
You don't see the forest from the trees.
Ps: the ballast resistor is outside of the feedback loop.
You don't see the forest from the trees.
Ps: the ballast resistor is outside of the feedback loop.
If you're talking about the ballast resistors that are needed when connecting multiple LM3886 in parallel, I'd go with a non-inductive type. The resistance should be chosen from the amount of standing current between LM3886es that you're willing to tolerate. Depending on your circuit, 0.1-0.33 Ω is about right.
The Zobel resistor (in the RC network from the amp output to ground) should be non-inductive. You want that network to present a load of R at high frequencies. I often use a metal oxide resistor there. A 2-3 W type is fine.
The resistor in the Thiele (L||R) network in series with the output can be an inductive type. I usually use a 2-3 W metal oxide resistor there. The resistor won't have any impact on the sound as the vast majority of the signal current flows through the inductor at audio frequencies.
Tom
The Zobel resistor (in the RC network from the amp output to ground) should be non-inductive. You want that network to present a load of R at high frequencies. I often use a metal oxide resistor there. A 2-3 W type is fine.
The resistor in the Thiele (L||R) network in series with the output can be an inductive type. I usually use a 2-3 W metal oxide resistor there. The resistor won't have any impact on the sound as the vast majority of the signal current flows through the inductor at audio frequencies.
Tom