Hi everyone,
I’m repairing a class D 3000W RMS amplifier, which uses a total of 12 IRFP064Ns in its power supply section. They are pretty hard to come by where I live. Can I use IRF3205 instead of them?
They look pretty similar on the datasheets, except for the safe operating areas.
I’m repairing a class D 3000W RMS amplifier, which uses a total of 12 IRFP064Ns in its power supply section. They are pretty hard to come by where I live. Can I use IRF3205 instead of them?
They look pretty similar on the datasheets, except for the safe operating areas.
The IRFP is a TO247 and the IRF a TO220 so mechanically they'll not be compatible. The TO247 package has twice as good case-to-sink thermals, that's what the larger package brings so even if the TO220 was made to fit, it would run considerably hotter. I'm guessing the original designers used TO247 for this reason - after all they did need 12 of them.
You are right about the package type but in the datasheets both have 200W of heat dissipation capability, meaning the actual silicone die I think is similar.
The only meaningful difference is in the SOA graphs at around 14V.
Yes from the capacitances and the current spec, Rds(on) etc. the silicon dies must be pretty close. But bigger packages exist for a reason otherwise every engineer would go for the cheaper TO220. If you read the IRF3205 DS you'll note they say this :
The TO-220 package is universally preferred for all
commercial-industrial applications at power dissipation levels
to approximately 50 watts.
Its rated at 200W but they only say 50 watts. Isn't that interesting?
The TO-220 package is universally preferred for all
commercial-industrial applications at power dissipation levels
to approximately 50 watts.
Its rated at 200W but they only say 50 watts. Isn't that interesting?
The marketing departments have fouled up many a power MOSFET datasheet with ludicrous max current ratings (in the small print its the die limit, not with bonding wires attached!!).
TO220 is good for 20A or so, max (you don't want the leads to unsolder themselves!). And 50W is pushing it some, 20W is more reasonable for routine use with long term reliability.
And be careful to distinguish pulse from continuous power ratings - look for the SOA graph for more detailed information, but bear in mind the heatsink is always assumed to be infinite in such graphs!
TO220 is good for 20A or so, max (you don't want the leads to unsolder themselves!). And 50W is pushing it some, 20W is more reasonable for routine use with long term reliability.
And be careful to distinguish pulse from continuous power ratings - look for the SOA graph for more detailed information, but bear in mind the heatsink is always assumed to be infinite in such graphs!