MOSET selection confusion

[Benjam]

The DC bus voltage of my motor is 48V, and the parameters are as follows:

Code: [Select]

Continuous Current: Adc --> 12.00 | Arms --> 10.00

Peak Current: Adc --> 40.00 | Arms --> 30.00

Rated Cont Power: Watts --> 350

Speed at Rated Power: RPM --> 2400

Design Voltage(Rated voltage): Vbus --> 48.0 (Vdc) | Vac --> 34(Vrms)

Back EMF at Temp: Vpk / kRPM -->18.5 | Vrms/kRPM --> 13.0

Back EMF: Vpk / kRPM --> 19.0 |  Vrms/kRPM -> 14.0

Motor Constant at 25°C: N-m/√watt --> 0.28 | oz-in/√watt --> 39.5

Inductance: Lm --> 0.55

Then when I was designing the drive, I needed to choose an NMOS. I saw ON Semiconductor, but when I was reading the datasheet, I was confused. For example: the datasheet of this product FDMS86101DC, the first sentence says:

MOSFET-N-Channel,POWERTRENCH, DUALCOOL 56 Shielded Gate 100 V, 60 A, 7.5 m
(https://www.onsemi.com/pdf/datasheet/fdms86101dc-d.pdf)

It seems that it can meet my needs, but there is a data in the datasheet Id = Drain Current − Continuous TA = 25°C Ratings=14.5A. This data makes me very confused. Doesn't it mean 100V, 60A? Why does 14.5A appear again? Can this product meet my needs?

[Benjam]

Sorry for the picture above, it is wrong, add it again

[Le_Bassiste]

It seems that it can meet my needs, but there is a data in the datasheet Id = Drain Current − Continuous TA = 25°C Ratings=14.5A.

the table calls out footnote 1a. MOSFET current ratings come with a defined set of environmental conditions, where that ratings are valid, i.e., ambient temperature, heatsink size, drive voltage, and so on.

[Jeroen3]

60 Amp is only possible with liquid nitrogen cooling or a thermostreamer, or a very short duty cycle.
The mosfet will dissipate arounds 30 Watts running 60 Amps, which it challenging in this small package.
Check Figure 11. Forward Bias Safe Operating Area what you can do with this mosfet.

Check also the other pages of the datasheet for great thermal examples.

[exe]

Ha-ha, marketing .

If space allows, you can look into fets in TO-220 package (or any other through-hole part), those are easier to cool down because there are tons of heatsinks.

[Faranight]

While selecting a Schottky diode for a project of mine, I've almost learned the hard way that charts in a datasheet are more important than the listed rated values. Because those values are normally specified under given test conditions. In my case, I was looking at forward voltage of 0.45V that was specified under current of 100mA, but  I thought I could run the diode at 1A at that voltage. Fortunately someone then pointed me at the voltage/current chart in the datasheet that showed my diode would have a forward voltage of 0.8V at 1A current. Bummer. So yeah, be careful when selecting components by the values listed in the table - those are always given at predetermined test conditions.

[Jeroen3]

Ha-ha, marketing .

If space allows, you can look into fets in TO-220 package (or any other through-hole part), those are easier to cool down because there are tons of heatsinks.

These mosfets have exposed metal op the top for heatsinking which is actually quite niche.

Infineon also has some inverted mosfets for the same thing, you can attach a heatsink with a squeezy thermal interface layer, like thick foil.
CanPak iirc, and there are the shared packages like TOLT, ixys has those as well.
https://www.infineon.com/dgdl/Infineon+-+Product+Brief+-OptiMOS+in+CanPAK.pdf?fileId=db3a30433d346a2d013d443eeb6e4719

[magic]

The important thing is power dissipation, which is I²·R where I is current and R is RDS(on) of the FET.
For 12A it's only about 1W. This is probably doable with SMD parts and no heatsinking besides the PCB itself.
The 14A spec likely applies to continuous current with the FET soldered to some particular test board, see footnote 1a.
The remaining question is: how long are those 30A peaks.