IRFP450 Mosfet Question

[Ken1]

Hi, Can I substitute a higher current Mosfet for an IRFP450 without having to change any other circuit components? If so, can you suggest some part numbers? The IRFP450 has a current rating of 14A and I would like a substitute with at least a 20A rating.

 

[Pommie]

An obvious replacement would be the IRFP460 - 20A 500V.

Mike.

 

[Tony Stewart]

A bigger switch always puts a bigger Ciss load on the gate .

 

[ChrisP58]

You haven't said what how you are using the mosfet, so it's difficult to accurately answer your question. As Tony points out, there will usually be some other parameter that changes along the way. You need to know what affect other parameters may have on your circuit.

But if increased current capacity is your only goal, there is a big trap you need to look out for.

One important fact that everyone need to be aware of when choosing mosfets, is that the current rating has a fine print disclaimer that causes problems if not taken into account.

You will notice on the "Continuous Drain Current" line of the IRFP450 data sheet if gives two current ratings. The 14 amp rating is at Tc=25C. Tc is case temperature, not ambient. So it will only handle 14 amps continuous if the case is kept at 25C all of the time. Not as easy as it sounds.

Let's look at a few items from the datasheet.

1) The Rdson number is 0.4 ohms at 10Vgs. So when carrying 14 amps, it will be dissipating about 78 watts of heat. So, it will only carry 14 amps if you can keep the case at 25C while it is pushing out 78 watts of heat. There is no passive heatsink/fan combination that can do that unless you're running it in a freezer. Otherwise you'll need to resort to cold water or refrigerated cooling.

2) Junction temperature. The sum of the junction to case and the case to sink resistances is 0.89 Degrees per watt. This means that the the junction is .89 degrees hotter than the case for each watt of power. So, at 78 watts, the junction is 69 degrees hotter than the case. If the case is at 25C, the junction is at 94C. Lets take that number down to the graph at Figure 4. That shows us that our Rdson is really about 75% higher than we thought. We need to recalculate everything again based on the new info. etc, etc, etc.

It comes down to the fact that the current rating of most mosfets is a calculated number as to what current will put the junction temperature at it's failure point when the case is at 25C. Doable in a cryogenics lab, but not really a practical value for most real world applications.

So, if you're looking for a mosfet to continuously carry 20 amps, you will probably need a mosfet(s) with current capacity 3 to 10 times 20 amps. A project I am doing right now needs to switch 60Amps DC. I am using four 100 amp mosfets. Maybe a little overkill but, other than the copper plates that carry the current, there is no heatsink.

This is the datasheet I used for this discussion. https://www.vishay.com/docs/91233/91233.pdf

 

[Ken1]

I am attempting to repair a Powermax PM3-45 power supply unit. The unit is relatively new and hasn't seen much use. It uses 2 of these Mosfets in the SMPS part of it. I am assuming these are too light for the duty they are trying to perform and would like to try and replace them with heavier Mosfets. They are mounted on a large aluminum heatsink and the unit also has a cooling fan.

 

[gary350]

I replaced an IRFP460 with STP55NF06 it is rated 50 amps. My circuit is 15 volts.

**broken link removed**

 

[Misterbenn]

I am assuming these are too light for the duty

Why do you assume this, has the unit overheated or cutout?

 

[Ken1]

One of the pair of 18.2 ohm 1/4w 1% resistors that drives the gates of the mosfets has burned up and has also damaged the UC3825BN SMPS chip. Pins 11 and 14 are the 2 outputs of the chip. Pin 11 which the burned resistor connects to has a very low resistance to ground compared to pin 14. There is also a visible crack in the case of the same Mosfet. The unit belongs to a neighbour and he said it just quit working; he doesn't know if it was running hot before it quit.
I just checked the STP55NF06 Mosfet and it's rated at 60V. That's way too low as this power supply is powered by 120VAC which is fed through a bridge rectifier, then filtered with an 1800uF 200V capacitor then fed to the Mosfets through the SMPS transformer.

 

[ronv]

Often the input FETs are just the weak link and something else is the real culprit. Electrolytic caps are a prime suspect. Give them a good look, especially those close to the hot heat sink.

 

[Tony Stewart]

It seems this "entry level" charger" has very little margin before exceeding the component limits.

My rule of thumb is the impedance of each stage, where the ratio of impedances of successive stages should not exceed 100 and ought to be in the 10 range, so that RC dynamic losses don't dominate and exceed conduction losses.

This is determined by the V-I slope of the driver and RdsOn of the switch.
If the impedance gain or resistance gain is too high, you need more stages of MOSFET buffering to prevent overloading the gate resistors due to Qgs charge switch losses and switch frequency.

For example the UC3825BN is only 5 Ohms nom and 10 Ohm worst case. ( Use delta V/I )
But the IRFP450 is 0.4 Ω with a ratio of 12 seems ok but the RdsOn looks a tad high.

The battery charge may not have intelligent soft start and overstress the charger hooking up to a dead battery. Thats my guess.

 

[ronv]

Take a look at this one.
https://www.fairchildsemi.com/datasheets/FD/FDP20N50F.pdf
Much better all around. Nice power supply by the way. What are you doing with it?
Don't forget to check the caps.

 

[atferrari]

You haven't said what how you are using the mosfet, so it's difficult to accurately answer your question. As Tony points out, there will usually be some other parameter that changes along the way. You need to know what affect other parameters may have on your circuit.

But if increased current capacity is your only goal, there is a big trap you need to look out for.

One important fact that everyone need to be aware of when choosing mosfets, is that the current rating has a fine print disclaimer that causes problems if not taken into account.

You will notice on the "Continuous Drain Current" line of the IRFP450 data sheet if gives two current ratings. The 14 amp rating is at Tc=25C. Tc is case temperature, not ambient. So it will only handle 14 amps continuous if the case is kept at 25C all of the time. Not as easy as it sounds.

Let's look at a few items from the datasheet.

1) The Rdson number is 0.4 ohms at 10Vgs. So when carrying 14 amps, it will be dissipating about 78 watts of heat. So, it will only carry 14 amps if you can keep the case at 25C while it is pushing out 78 watts of heat. There is no passive heatsink/fan combination that can do that unless you're running it in a freezer. Otherwise you'll need to resort to cold water or refrigerated cooling.

2) Junction temperature. The sum of the junction to case and the case to sink resistances is 0.89 Degrees per watt. This means that the the junction is .89 degrees hotter than the case for each watt of power. So, at 78 watts, the junction is 69 degrees hotter than the case. If the case is at 25C, the junction is at 94C. Lets take that number down to the graph at Figure 4. That shows us that our Rdson is really about 75% higher than we thought. We need to recalculate everything again based on the new info. etc, etc, etc.

It comes down to the fact that the current rating of most mosfets is a calculated number as to what current will put the junction temperature at it's failure point when the case is at 25C. Doable in a cryogenics lab, but not really a practical value for most real world applications.

So, if you're looking for a mosfet to continuously carry 20 amps, you will probably need a mosfet(s) with current capacity 3 to 10 times 20 amps. A project I am doing right now needs to switch 60Amps DC. I am using four 100 amp mosfets. Maybe a little overkill but, other than the copper plates that carry the current, there is no heatsink.

This is the datasheet I used for this discussion. https://www.vishay.com/docs/91233/91233.pdf

Definitely worth reading twice at least. Gracias.

 

[Tony Stewart]

A bigger switch always puts a bigger Ciss load on the gate .

What's your budget ? $10? here's a good one with same low Ciss

https://www.digikey.ca/product-detail/en/IXFQ60N50P3/IXFQ60N50P3-ND/2650882


Drain to Source Voltage (Vdss) 500V
Current - Continuous Drain (Id) @ 25°C 60A (Tc)
Rds On (Max) @ Id, Vgs 100 mOhm @ 30A, 10V
Vgs(th) (Max) @ Id 5V @ 4mA
Gate Charge (Qg) @ Vgs 96nC @ 10V
Input Capacitance (Ciss) @ Vds 6250pF @ 25V

vs prev sugg. your choice $ or RdsOn
FDP20N50F, FDPF20N50FT < $4.
Drain to Source Voltage (Vdss) 500V
Current - Continuous Drain (Id) @ 25°C 20A (Tc)
Rds On (Max) @ Id, Vgs 260 mOhm @ 10A, 10V
Vgs(th) (Max) @ Id 5V @ 250µA
Gate Charge (Qg) @ Vgs 65nC @ 10V
Input Capacitance (Ciss) @ Vds 3390pF @ 25V
Power - Max 250W