L20 AMP。use only two NJW0302G

[ljm_ljm]

To reduce the four than before to 2. But it is more low price and easy installation.

And NJW0302G power is bigger, overall output power 200 W 8 R (+ 65 V)
Would not reduce the output power.

Please note that blue PCB genuine original,

Designers LJM.

I will name L20SE life for it, to distinguish them from.

[ljm_ljm]

Each channel to use two NJW0302G NJW0281G.

Use B649 D669 + A1930 C 5171 as the drive pipe

[dantzig65]

No trimpot to adjust quiescent current?

[ljm_ljm]

output satage quiescent current Set to 10MA (1 transistor)

[AndrewT]

Two pair of plastic packaged To247 or To3p will be highly stressed driving an 8ohm speaker from +-65Vdc supply rails.
Two pair of metal packaged, high temperature, high power To3 will cope with this if you can keep them warm but not hot.

Reducing to one pair on +-65V supply rails is virtually guaranteeing early failure, even into 8ohms. Into 16ohm, low to medium reactance speakers, one pair might survive to a long life.

[ljm_ljm]

Originally Posted by AndrewT

Two pair of plastic packaged To247 or To3p will be highly stressed driving an 8ohm speaker from +-65Vdc supply rails.
Two pair of metal packaged, high temperature, high power To3 will cope with this if you can keep them warm but not hot.

Reducing to one pair on +-65V supply rails is virtually guaranteeing early failure, even into 8ohms. Into 16ohm, low to medium reactance speakers, one pair might survive to a long life.

CLASS AB Efficiency about 65%.
OUTPUT 200W POWER，Loss power about 35%. 107 watt.

They will share in these transistors TO-3P。NJW0302G+ NJW0281G

[AndrewT]

Originally Posted by ljm_ljm

CLASS AB Efficiency about 65%.
OUTPUT 200W POWER，Loss power about 35%. 107 watt.

They will share in these transistors TO-3P。NJW0302G+ NJW0281G

the 107W is an average power prediction for driving an 8r0 resistor load.

This 107W cannot be used to determine whether the transistors will survive driving an 8ohm speaker when the devices start to warm up or worse get hot.

I will give you free starter:
A two pair output stage driving 200W into an 8r0 resistor, will get warm. Calculate what temperature the device cases (Tc) will be for your average power dissipation in each device.
From the Tc that you determined, calculate the Pmax that the devices can dissipate.
Is this average dissipation more or less than the DC (continuous) power rating for the device operating at that Tc?
Is this average more or less than the 100ms one shot power rating for the device operating at that Tc?
Is this average more or less than the 10ms one shot power rating for the device operating at that Tc?
Is this average more or less than the 1ms one shot power rating for the device operating at that Tc?
Is this average more or less than the 100us one shot power rating for the device operating at that Tc?
Now that you are into short term one shot transients, you must consider not the average dissipated power but the instantaneous peak dissipated power.
Next stage is to repeat those comparisons using resistor loading and instantaneous Pmax, to all those temperature de-rated SOA limits.
Finally, the difficult bit, you need to examine what effect changing from a resistive load to a reactive load does to the average P and instantaneous Pmax and use this information to make all those comparisons again.
Manually this is a lot of work.
David Eather shows a manual method for doing this.
Benson gives a spreadsheet to convert the manual method to computer assisted for mosFET output stage.
There are other computer assisted methods available to Members.
Start with Eather so that you can understand what it is that you need to know to make the assessment that you originally asked for.

[stoc005]

With 65 volt supply and 8 ohms, worst case power in the output devices is around 132 watts, which occurs at less than full power out. I agree that 107 watts is around the power in the output devices at full power out. Heatsinks need to be designed for WORST CASE power dissipation, IMHO. One pair of devices will be highly stressed, like AndrewT said. More output devices tend to makes things better as long as the heatsink is big enough.

[sajti]

I use 4pairs of NJW0xxx with +/-65V. It's easily drive the 4ohm loudspeaker, and the distortion is significantly lower with 4pairs compare to 2 pairs (I use no nfb)

Sajti

[padamiecki]

Originally Posted by sajti

(I use no nfb) Sajti

Please show