Is it safe to use the intrinsic diode of a FET as a .... diode?

[AndersJ]

I am considering using a SOT-23 FET, such as the 2N7002 or FDN5630, as a diode only.

I intend to short the gate and source,
and then connect the transistor "backwards", to use it as a diode.

My application is a 100-200 mA diode oring of two incoming voltages.

The underlying reason to do this
is to take one part type out of the BOM, and to learn and try something new.

Are there any downsides to doing this?

[mariusT]

I don't know why this shouldn't work, just have an eye on the maximum current and power rating. 2N7002 is only specified up to 115 mA diode current at 25°C ambient temperature (some derating may be necessary)

edit: typo

[Siwastaja]

Many designs rely on this, it's not wrong. It's only, the diode is not that well specified so you need to work with some assumptions. Like what's the maximum current, is it the same as Id? Although not that this really matters; power dissipation will be limiting the maximum current first, because MOSFETs rarely can achieve the rated Id even when turned fully on, with way less Vds than 1V. With the diode, you have more voltage drop, and can run even less current before exceeding Tj_max. But this is at least trivial to calculate, given even some value for body diode Vf.

[TimNJ]

You can treat the body diode as a "real diode", no problem...It's just that it's a rather crappy diode (voltage drop, and often recovery time/charge) when compare to a comparably sized standalone diode.

[TimFox]

Gate-channel diodes (drain and source connected together) in low-current JFETs are commonly used as protection diodes in high-impedance circuitry (such as input buffers for voltmeters).  The relevant specifications in the data sheet are gate-drain breakdown voltage V(BR)_GSS and leakage current I_GSS.  Maximum gate forward-bias current may be in low mA for small devices (10 mA for 2N4416, which is a higher-current device than typically used for protection diodes).
The original post said "FET", but there are still both JFETs and MOSFETs available for purchase.
Another useful diode for high-impedance protection is the C-B junction of small signal BJTs, e.g. 2N2484.

[David Hess]

The body diode in a power MOSFET is the collector-base junction of the parasitic bipolar transistor, so except for power dissipation limits, it has a current and voltage rating equal to that of the power MOSFET.

Recovery time is fast compared to a standard recovery time rectifier which can be microseconds, but sometimes not quite as fast as a fast recovery rectifier which could be 150 to 250 nanoseconds.  The datasheet will say if the body diode has been optimized for fast recovery.

The gate diode in a JFET is completely different and can be used as a low current low voltage low leakage diode.  Modern low leakage diodes are often JFETs in disguise.

[tszaboo]

The body diode in a power MOSFET is the collector-base junction of the parasitic bipolar transistor, so except for power dissipation limits, it has a current and voltage rating equal to that of the power MOSFET.

Recovery time is fast compared to a standard recovery time rectifier which can be microseconds, but sometimes not quite as fast as a fast recovery rectifier which could be 150 to 250 nanoseconds.  The datasheet will say if the body diode has been optimized for fast recovery.

The gate diode in a JFET is completely different and can be used as a low current low voltage low leakage diode.  Modern low leakage diodes are often JFETs in disguise.

Why would the power dissipation rating of the diode be lower than the FET? It is part of the silicon, distributed evenly.
In fact, for example, when I had to measure the thermal resistance of a FET, I used the diode as the dissipating element, as that was the easiest to setup.

[magic]

I recall reading an appnote about MOSFET avalanche recently where it has been determined that localized short term temperature rise from pulsed avalanche behaves slightly different (worse) than short term heating from pulsed linear operation (the familiar thermal impedance curves). Different effects may occur in different parts of the structure or potentially in more localized spots.

None of that necessarily means that the same is the case for body diode forward conduction.

[Terry Bites]

It depends on what the diode has to in your circuit. Recall its all one bit of silicon and the heat from the diode adds to the total MOSFET dissipation. You need to have a good estimate of the energy the diode will have to absorb per cycle in order to decide if you need an external one or not.

[David Hess]

Why would the power dissipation rating of the diode be lower than the FET? It is part of the silicon, distributed evenly.
In fact, for example, when I had to measure the thermal resistance of a FET, I used the diode as the dissipating element, as that was the easiest to setup.

It should have the same power rating as the MOSFET, however diodes are normally rated by current and at high current, the power dissipation may exceed that of the MOSFET when fully enhanced because the forward voltage drop will be higher.  At least this can be true for low voltage MOSFETs; high voltage parts have vary large area and high channel resistance so the opposite would be true.

For testing, the body diode can be used to measure the junction temperature.  Bipolar transistors are tested this way also.

[Someone]

The body diode in a power MOSFET is the collector-base junction of the parasitic bipolar transistor, so except for power dissipation limits, it has a current and voltage rating equal to that of the power MOSFET.

Why would the power dissipation rating of the diode be lower than the FET? It is part of the silicon, distributed evenly.
In fact, for example, when I had to measure the thermal resistance of a FET, I used the diode as the dissipating element, as that was the easiest to setup.

It should have the same power rating as the MOSFET, however diodes are normally rated by current and at high current, the power dissipation may exceed that of the MOSFET when fully enhanced because the forward voltage drop will be higher.  At least this can be true for low voltage MOSFETs; high voltage parts have vary large area and high channel resistance so the opposite would be true.

For testing, the body diode can be used to measure the junction temperature.  Bipolar transistors are tested this way also.

Its probably just your awkward wording "except for power" in the original, perhaps this is clearer: the diode is already considered in the datasheet limits, so the maximum voltage, current, and power limits all remain.
https://assets.nexperia.com/documents/technical-note/TN00008.pdf
There may be some odd devices out there with different characteristics in forward and reverse, but they probably make it obvious in their datasheet.

[David Hess]

I forgot to mention that beyond static power dissipation due to the forward voltage drop and forward current, reverse recovery losses can be considerable and destroy a diode if improperly applied even at lower currents.