Typically, R18 is 1W to 3W with low inductance, and R19 is 10W.
But R18 will often burn if the amp oscillates at full power.
But R18 will often burn if the amp oscillates at full power.
It depends. (Of course!)
If that network is on a 5W class A, then low power resistors will be fine, but on the end of a 500 W amp with a tendency to oscillate at 100Khz then they’d need to be big.
To get an idea work out the voltage at full output, then do some ohms law to see what power say R19 dissipates at that voltage. It won’t be quite that high at 20Khz, but it would be at higher frequencies.
At low frequencies neither will dissipate much.
As a rough guide on my blameless amp I think the zobel resistor is 3W and never gets warm.
If that network is on a 5W class A, then low power resistors will be fine, but on the end of a 500 W amp with a tendency to oscillate at 100Khz then they’d need to be big.
To get an idea work out the voltage at full output, then do some ohms law to see what power say R19 dissipates at that voltage. It won’t be quite that high at 20Khz, but it would be at higher frequencies.
At low frequencies neither will dissipate much.
As a rough guide on my blameless amp I think the zobel resistor is 3W and never gets warm.
Assuming that amp never oscillates and you use it for music rather than sine waves I’d suggest 1W would be plenty for each of them.
(15W is about 11V RMS into 8 ohms)
(15W is about 11V RMS into 8 ohms)
Then calculate the current through R18 + C6 at 11Vrms and 10kHz and P = R * I^2. For practical reason, R6 should be rated at 2W at least.
If the amplifier doesn't oscillate, and if you don't make prolonged stress tests at max amplitude , >20kHz, 1W is sufficient for R18, maybe 3W to be completely safe, and R19 can be as small as as practical: 0.25W for example, except the coil is often directly wound on the resistor body, meaning at least 1W