I already have the datasheet, it states it can handle 50 volts at 50 amps but I'm worried about spikes since I'm using it to switch the primary side of the transformer at 70 Hz and it will be a square wave signal, so my concern is won't it over heat quickly since I'm near the 50 Vds point? And won't the 3 amps just make things worse?
\$\begingroup\$
\$\endgroup\$
7
-
5\$\begingroup\$ A square wave means on and off. When the MOSFET is on there won't be 50V across it anymore. When it is off there won't be 3A flowing anymore. Dissipated power is the product of voltage across the device and current trough the device. \$\endgroup\$– UnimportantCommented Mar 20, 2021 at 18:46
-
1\$\begingroup\$ yeah, you can't push "voltage" through something. We might talk about current, but if current flows, then that voltage won't be 40 V. So, I think you need to get your basics straight. Maybe start by doing what we basically always need: Draw the schematic that involves your MOSFET, and clearly mark between which two points you are seeing 40 V, and through which cut 3 A flow. Then we can deliberate about things on a common basis! (Voting to close – lacking clarity without a schematic) \$\endgroup\$– Marcus MüllerCommented Mar 20, 2021 at 18:51
-
\$\begingroup\$ If you turn the MOSFET on hard, the resistance will be low (~0.025 ohms). The static power is fine, i^2R = 3^2 * 0.025 = 0.23 W. You may still have a problem if your switching frequency is high, more power will be dissipated during the switching. \$\endgroup\$– Mattman944Commented Mar 20, 2021 at 18:54
-
\$\begingroup\$ You need to specify the stored energy by L and DCR of the transformer to get or simulate actual Pd. You can’t expect Vmax and I max unless you have a max heat sink and are running DC only. \$\endgroup\$– Tony Stewart EE75Commented Mar 20, 2021 at 19:58
-
\$\begingroup\$ Primary side of a transformer means what? flyback? full bridge? Either way, a good 20% margin is advisable over the max expected voltage when the snubber is active. For non inductive loads, 40V at 3A is just fine for the IRFZ44/IRLZ44 \$\endgroup\$– IndraneelCommented Mar 20, 2021 at 19:58
|
Show 2 more comments
4 Answers
\$\begingroup\$
\$\endgroup\$
2
It can't handle 50 volts at 50 amps, that would fry it immediately. It can handle either 50 volts or 50 amps. 50 volts when turned off (Vgs < Vth) and 50 amps (in pulses only, or with extremely good heatsinking!) when turned fully on (Vgs >> Vth).
If you're using it at 40 V, it should be perfectly fine. 40 V at 3 A is quite a bit of power, however, and you'll probably need liquid cooling for that much heat--that's 120 watts, more than your average CPU! Most likely, what you mean is 40 V when the FET is off (which is well within its rated 50 volts) and 3 A when it's on, which is (very!) well within its rated 50 A. If turned on hard enough, which for this FET is at a Vgs as low as just 4 volts, the power dissipation will be \$R_{ds,on}·I_D^2 = 315\ \mathrm{mW}\$, using the \$R_{ds,on}\$ at Vgs = 4 V found in this datasheet. This is low enough that you could probably get away with not using a heatsink at all, especially if it's pulsed.
-
\$\begingroup\$ So if I use two in parallel, do you think it will be more than enough to keep things cool without a heatsink \$\endgroup\$– BELSmithCommented Mar 20, 2021 at 19:14
-
\$\begingroup\$ I don't think you need to use two in parallel, but you can if you want. 315 mW will result in a temperature rise, with no heatsink, of about 20 °C above ambient, which is well within the rated temperature and not even hot enough to burn you unless it's already a really hot day. And if you're driving the FET with 5 V, that power dissipation drops to a level where you're only going to see it heat up by about 14 °C. \$\endgroup\$– HearthCommented Mar 20, 2021 at 19:30
\$\begingroup\$
\$\endgroup\$
2
I already have the datasheet, it states it can handle 50 Volts at 50 Amps
The data sheet says it can handle 55 volts and the data sheet also says it can handle 47 amps but, it never says that it can handle both this situations at the same time for any length of time. So, when you say 50 Volts at 50 Amps you are misinterpreting things as well as getting the actual numbers wrong a little.
So, if you really wanted to know what it can handle you'd look at this graph: -
And figure out what voltage and current that can be simultaneously handled over what period of time and then degrade your expectations a little bit to account for anomalies and spikes.
answered Mar 20, 2021 at 19:03
-
\$\begingroup\$ I should have thought of talking about the SOA graph in my answer! \$\endgroup\$– HearthCommented Mar 20, 2021 at 19:31
-
\$\begingroup\$ The you would estimate the L/(DCR+Ron)=T to estimate the pulse width. Max and the XL(f) to estimate the excitation current or Amp-s I the case of flyback \$\endgroup\$ Commented Mar 20, 2021 at 22:29
\$\begingroup\$
\$\endgroup\$
The IRFLZ44N is a 22 mOhm FET. With 3A the power dissipation will be 0.2W causing about 12 degrees C temperature rise. So the current isn't a problem if the FET is fully on.
However you don't push voltage through a FET. The FET can stand off 40V when it's off, but personally I would use a higher voltage rating for more margin.
If you really want 40V and 3A with the FET in linear mode you'd better look for a different part or a really good thermal management system. (Heatsink, forced air cooling, liquid cooling, etc.)
\$\begingroup\$
\$\endgroup\$
1
Take a margin of 20% at least when dealing with Vdsmax of power Mosfet's.
Pick a more robust Mosfet if you can.
The leakage inductance of the transformer will produce voltage spikes on the drain of your Mosfet that might go well beyond 50 V.
I assume here that you are designing a flyback power supply.
answered Mar 20, 2021 at 18:52
-
2\$\begingroup\$ Using a 50 V rated part at 40 V is already a margin of 20%; I wouldn't be concerned about it especially because the datasheet I found actually has its rating at 55 volts. \$\endgroup\$– HearthCommented Mar 20, 2021 at 19:02