I've searched the forum and can't seemto find the answer. Is there any difference between different manufacturers "IRFP240" FETs? For example here is the actual IR part :
http://www.digikey.com/scripts/DkSearch/dksus.dll?Detail?Ref=658044&Row=346147&Site=US
and here is a Fairchild IRFP240 :
http://www.mouser.com/index.cfm?han...uctid=410952&e_categoryid=277&e_pcodeid=51229
Thanks,
--Chris
http://www.digikey.com/scripts/DkSearch/dksus.dll?Detail?Ref=658044&Row=346147&Site=US
and here is a Fairchild IRFP240 :
http://www.mouser.com/index.cfm?han...uctid=410952&e_categoryid=277&e_pcodeid=51229
Thanks,
--Chris
Lotsa diffs..
Significant ones:
IR device EAS of 510 mJ vs FC at 250 mJ.
IR is capable of absorbing more energy in reverse breakdown, single pulse. Most likely, this is due to IR's better guardring design. Typical hits: no snubber switching, or inductive kickback.
Thermal resistance...FC better at .69 vs IR .83. I would guess FC is larger die.
The FC device looks a bit slower...on off delays and rise..
IR has 20 volt gate capability...FC at 30. FC has thicker oxide.
Transfer characteristics...IR has zero TC crossover at about 6.5 volts, FC is about 5.5 volts. Affects paralleling stability in linear or switching apps.
Cheers, John
Significant ones:
IR device EAS of 510 mJ vs FC at 250 mJ.
IR is capable of absorbing more energy in reverse breakdown, single pulse. Most likely, this is due to IR's better guardring design. Typical hits: no snubber switching, or inductive kickback.
Thermal resistance...FC better at .69 vs IR .83. I would guess FC is larger die.
The FC device looks a bit slower...on off delays and rise..
IR has 20 volt gate capability...FC at 30. FC has thicker oxide.
Transfer characteristics...IR has zero TC crossover at about 6.5 volts, FC is about 5.5 volts. Affects paralleling stability in linear or switching apps.
Cheers, John
Jneutron,
Thanks for looking at the DS!
These will be used in a class A audio amplifier so isn't it safe to assume the mosfet should always be on so the on/off rise time will not be detrimental?
Better thermal resistance =
About the zero crossing, could you elaborate on what you wrote :
Transfer characteristics...IR has zero TC crossover at about 6.5 volts, FC is about 5.5 volts. Affects paralleling stability in linear or switching apps.
I'm not very familier with the transfer charactersitic of mosfets, this will be my first project so if you could explain to me in detail I would really apreciate it
Thanks in advance,
--Chris
Thanks for looking at the DS!
These will be used in a class A audio amplifier so isn't it safe to assume the mosfet should always be on so the on/off rise time will not be detrimental?
Better thermal resistance =
About the zero crossing, could you elaborate on what you wrote :
Transfer characteristics...IR has zero TC crossover at about 6.5 volts, FC is about 5.5 volts. Affects paralleling stability in linear or switching apps.
I'm not very familier with the transfer charactersitic of mosfets, this will be my first project so if you could explain to me in detail I would really apreciate it
Thanks in advance,
--Chris
For your app, the differences probably don't mean a hill of beans..DIY_newbie said:
I think the FC die is larger, so it is able to dissipate a little more heat..the difference also isn't that big.DIY_newbie said:
Better thermal resistance =
DIY_newbie said:
About the zero crossing, could you elaborate on what you wrote :
Transfer characteristics...IR has zero TC crossover at about 6.5 volts, FC is about 5.5 volts. Affects paralleling stability in linear or switching apps.
I'm not very familier with the transfer charactersitic of mosfets, this will be my first project so if you could explain to me in detail I would really apreciate it
Thanks in advance,
--Chris
If you look at the transfer graph for the FC device, there are 3 lines...-55, 25, and 150 C operation. If you put 4 volts on the gate of the device while it is at -55 C, it will carry about 100 milliamps. If you raise the temp to 25 C, it will carry 600 milliamps. At 150 C, it will carry 2.5 amps..At this gate voltage, the current will increase as the temperature increases...a positive temperature coefficient.
Now, if you put 5.5 volts on the gate, look at the curves..all three are at the same current...20 amperes..this means that at that gate voltage, the device current will be the same regardless of the temperature. This is a zero temperature coefficient.
Note that above 5.5 volts, the curves have crossed..increasing the temperature there means the devices will carry LESS current. This is a negative temperature coefficient.
If you parallel two devices, and both are on with gate less than 5.5 volts, if one device heats more than the other, it will conduct more current, this is a thermal runaway condition.
If you do the same, but they are above 5.5 volts, the hotter one will carry less current..this is more stable at that gate voltage.
It is easy enough to design the circuitry to get around this also..
For a simple on-off application with two devices, just make sure you run them with a gate voltage above the zero point in the curves. For the FC device, that is 5.5 volts, for the IR, 6.5.
For your app, source resistors most likely will suffice if you parallel them.
Cheers, John
Remember too that you can hand identical devices to the
engineering and marketing departments of both companies
and still get different ratings.
My own experience with 240 types is that IR has a little better
matching in a lot, but I haven't noticed any other difference
worthy of comment. On the P channel devices, the Harris
wins by virtue of the more constant transconductance in the
mid-band audio frequencies.
engineering and marketing departments of both companies
and still get different ratings.
My own experience with 240 types is that IR has a little better
matching in a lot, but I haven't noticed any other difference
worthy of comment. On the P channel devices, the Harris
wins by virtue of the more constant transconductance in the
mid-band audio frequencies.
peranders said:
don't know the parts in particular but cmos doesn't usually generate a lot of current so if the circuit in question requires a lot of current out of the IC, the cmos version may not function.
I would also expect any circuit that assume IC-specific impedance doesn't work directly when migrating to the cmos version. kind of like the bjt and cmos versions of the 555 timer.
In this case, the cmos version doesn't work right away because the circuit had pushed the operating parameters out of the envelope of the cmos ic - which is a different circumstance from what we are discussing here.
I will look into the lmc situation a little bit closer.
jneutron said:
John,
1: it's not important parameter for an audio designer, not even a SMPS designer as I have bean for a big telecom company.
FC have higher repetitive avalanche energy than IR and also better SOA which is what counts here.
2: This doesnt sound to belong nowhere especially in audio designing, or I'm missing the point here.
3: I agree with you here, FC seem to have "larger die" probably and that's good.
4: Take once more a closer look in the datasheet you will see that IR are testing with a Gate resistor well below half of that FC uses, they are probably quite equal in speed, further FC have much lower Qg/Cgd and should be even faster than IR, or at least having a smaller impact on distortion due to the unliner capacitance behaviour in this aspect.
5: Not relevant information in any aspects I would say.
In SMPS one doesnt use much more than 15 Vgs.
6: Some differences here but no big issues between these FET's, just that the quite similar thermal behaviour of FC and IR is at different Vgs. Both still have a huge positive tempco that has to be supervised and is a bigger challenge.
All in all FC wins in many cases, how it works as an audio amplifying device giving euphonic pleassure to the listener is an another thing.
Nelson Pass said:On the P channel devices, the Harris
wins by virtue of the more constant transconductance in the
mid-band audio frequencies.
What does that mean?
Cheers Michael
May I throw one more into the mix? This device is supposed to be Motorola's version. How do you think it compares and can it be used in the same circuit as the IRFP240?
Thanks, Terry
Thanks, Terry
Ultima Thule said:
When you setup a measurement of transconductance in
Common Source mode with a given resistive load and then
sweep the audio band, for IR P channel types you see a
shelving to a lower gain in the midband. This generally
leads you to consider whether the part is appropriate in
CS applications. It can be dealt with, but you want to know
it's there.
Ultima Thule said:
John,
1: it's not important parameter for an audio designer, not even a SMPS designer as I have bean for a big telecom company.
FC have higher repetitive avalanche energy than IR and also better SOA which is what counts here.
2: This doesnt sound to belong nowhere especially in audio designing, or I'm missing the point here.
3: I agree with you here, FC seem to have "larger die" probably and that's good.
4: Take once more a closer look in the datasheet you will see that IR are testing with a Gate resistor well below half of that FC uses, they are probably quite equal in speed, further FC have much lower Qg/Cgd and should be even faster than IR, or at least having a smaller impact on distortion due to the unliner capacitance behaviour in this aspect.
5: Not relevant information in any aspects I would say.
In SMPS one doesnt use much more than 15 Vgs.
6: Some differences here but no big issues between these FET's, just that the quite similar thermal behaviour of FC and IR is at different Vgs. Both still have a huge positive tempco that has to be supervised and is a bigger challenge.
Cheers Michael
The question was what were the differences..this is what I gleaned from the datasheets.
I did not address how those differences were relevant to audio or smps, merely pointed out where there were differences. I left it to others like you to clarify application..
Reverse breakdown single pulse energy absorbtion is a direct indication of the robustness of the racetrack isolation region of the die, as the bulk is not where the avalanche occurs. IR has a very strong history in this regard, and tailors the breakdown very nicely. I've had IR dice PASS all specifications even after the die was broken in half, with the polysilicon bridging the crack at the hinge point. Darndest thing I ever did see. (it broke during eutectic scrubbing, there was a particle under the die on the source side, and the die was pushed down in the middle, breaking it. Wirebond operator didn't notice the dihedral.
FC is using more linear length to accomplish what IR did with less. Basically, what IR does is just economically better, more die to a wafer.
Vgs..while design normally avoids overstressing the part in this respect, the difference between 20 volt flashover and 30 volt could be significant in smps should the drive or snubber be adequate for the 30 but not the 20, here, substitution could give problems..
Cheers, John
Ultima Thule said:
It's a mysterious characteristic. Charles Hansen related that
someone explained it to him, but he forgot the explanation.
I regard it like the weather - there it is. Again, this becomes
an issue when you use the parts in Common Source, and I have
not seen a significant difference based on Gate impedance, but
there is a large dependence on Drain impedance, so if you want
to eliminate this effect, you can cascode the parts or drive a
low impedance load.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.