What is the problem of my freewheeling diode 1N5822?

​Hey, guys! I have a DC motor with a voltage of 12V and a power of 550W, using a freewheeling diode 1N5822, but now this diode is often burned. Check the information to see that its maximum instantaneous current reaches 80A, thus I think it can work properly in the circuit. Why this happened? What should this margin be? Is it related to reverse recovery time?

Is there anyone can give me some idea, comment here, many thanks.

You're putting 80-amp pulses through a 3A diode.... and you are wondering why it's burning out? Those diodes are nowhere near sufficient for this purpose.

Consider the coil as an inductor, inductor current does not change instantaneously. Due to this property it acts like a source of current that changes voltage in order to maintain the current. When the switch is opened the voltage drops and quickly becomes negative as the inductor attempts to maintain its current in the same direction. The relationship is V = iR + Ldi/dt where L is the inductance, and di/dt is the rate of change of current with respect to time. So as it struggles to maintain its current the voltage becomes negative and can spike to whatever it needs.

Putting a diode in reverse bias parallel with the inductor allows a path to dissipate this energy protecting the rest of your circuit. The diode acts as a shunt providing a short with dissipation based on R. Increasing R can speed up the decay but it also raises V by the relationship outlined in Ohms Law. The current is no higher than what the motor pulled to begin with.

there’s an exponential decay which is what we direct using the fly back diode. You can calculate this when you know the values of L and R and the current. There are other ways to reduce this using Zeners, MOVs and transorb devices.

If you have a 12v 550 watt DC motor, I’m calculating over 40A to run it. You have that and the high reverse voltage to deal with.

DrAzzy:
You're putting 80-amp pulses through a 3A diode.... and you are wondering why it's burning out? Those diodes are nowhere near sufficient for this purpose.

Its pulse rated for 80A, non-repetitive. This not the same as handling 80A pulses repeatedly...

If the load is around 40A, at 12V, the diode needs to be rated at least 12V (there is no voltage spike because
the diodes there), and able to handle repeated 40A pulses.

Reverse recovery is not relevant for this application, it is forward biased when doing its job.

Something more like a 10A or 20A TO220 diode on a heatsink would be better suited.

If the motor is being PWM'd then the pulses are repeated frequently and thus the power dissipation in the
diode will be significant, hence heatsink. During the time its coducting it will have instantaneous power dissipations of several tens of watts.

Non-Repetitive Peak Surge Current

1N5822

Read the fine print for the parameter : "Non-Repetitive Peak Surge Current"

Also figure 8 in the datasheet shows the repetitive peak current drops dramatically with number of repetitions: https://www.onsemi.com/pub/Collateral/1N5820-D.PDF