D
Deleted member 550749
That suggests its OK bias wise.
There are a few reasons and you will have to work through each one to eliminate it. It should be so hot at just one watt output.
1/ If you are looking at average voltages with music and using your meter on AC then the result will be hopelessly inaccurate. Use your scope to look at the voltage across the load and look what the peak voltage is. If you are worried about the scope causing problems then use a series ten ohm with the ground connection.
2/ The amp is becoming unstable when playing music. You need the scope again to check and the scope need to have a wide bandwidth of at least 10Mhz to see what is going on.
3/ The bias current is poorly controlled and the amp goes into thermal runaway. Easy to check. Monitor the bias current (voltage across the emitter resistors) when idling, then get the amp hot and suddenly cut the music while monitoring the current. It should not be excessively high when hot.
How do a measure the current across the resistor do you mean voltage across the 0.1 ohm resistor?
On the scope I don’t see much it’s clean frequency response what should I be looking for with the scope and do you want me to monitor the output ?
Yes voltage across as I mentioned here:
Oscillation would show on the amplifier output. It can take many forms but should be obvious if it is occurring.
im guessing im checking in DC voltage across the resistor correct? and what kind of voltage should i look for or do i just compare the idle voltage to when the amp is really hot !
Measure across the resistor. 5 millivolts across 0.1 ohm would be 50ma bias which seems a reasonable value although the exact value depends on what the designer chose and set.
50ma with a -/+ 35 volt supply is 1.75 watt dissipation per output transistor.
50ma with a -/+ 35 volt supply is 1.75 watt dissipation per output transistor.
If the amp designer has done a good job the bias should not be too excessive when hot. Compensating for bias current vs temperature is a basic design task with any amp.
For example if the bias is 50ma idling and goes to 70ma when hot then that is OK, if it goes to say 100ma or higher then the designer has done a poor job with the vbe multiplier configuration.
For example if the bias is 50ma idling and goes to 70ma when hot then that is OK, if it goes to say 100ma or higher then the designer has done a poor job with the vbe multiplier configuration.
D
Deleted member 550749
If it's possible then remove biasing resistor of D669 ( base to emitter) and put a 1k pot and control ur self what u want , just ad only 2 pin like this image
Note that class B amp like this one that uses a Sziklai pair or Complementary feedback pair has very low bias current compared to EF output stage and this means it runs very cool at idle. However, it doesn’t reach its peak efficiency until the power output is very high, which means at low power it is terribly inefficient. Thus it gets hot at quite low power output. Thus is a 50W amp so you do need a decent heatsink.
That said, it will run hotter than expected if it is oscillating, even if the oscillation doesn’t stop it producing audio. CFP output stages do not handle oscillation well as they have limited turn off speed and you get some cross conduction where both transistors remain on for too long causing large pulse currents to flow resulting in high current draw and excessive heating.
If you output wires are too close to your inputs this can happen. Likewise, poor earthing can cause oscillation.
Hope that makes sense.
That said, it will run hotter than expected if it is oscillating, even if the oscillation doesn’t stop it producing audio. CFP output stages do not handle oscillation well as they have limited turn off speed and you get some cross conduction where both transistors remain on for too long causing large pulse currents to flow resulting in high current draw and excessive heating.
If you output wires are too close to your inputs this can happen. Likewise, poor earthing can cause oscillation.
Hope that makes sense.
I mean yes that makes sense but man I’m only pulling like 1 to 2w from the amp it’s acting like I’m drawing 100w it gets so dam hot can hardly touch it like maybe hold my finger on the heat sink for 5sec
How are you determining this 1 to 2 watts figure? What measurements are you using to derive that and with what test equipment.
I just a meter to check output voltage into 4 ohms I can do it with a test tone if it doesn’t hurt the speaker to be more accurate
It might be worth trying. Use a frequency your meter can accurately measure... most are ok up to at least 1kHz. A reading of 2 volts AC at 1kHz is 1 watt into 4 ohms. That won't hurt the speaker but it will sound very very loud.
I have a 8ohm 100w resistor but last time I tried them it popped the output or maybe that was the oscilloscope ground I can’t be sure
An 8 ohm resistor should be perfect and maybe an easier load than the speaker because it is not reactive.
Try the resistor and turn the volume up slowly 2.83v would be 1 watt into 8 ohm.
Try the resistor and turn the volume up slowly 2.83v would be 1 watt into 8 ohm.