3
\$\begingroup\$

Say I have a design that switches a 5A coil using a MOSFET. In order to protect the MOSFET, I place a flyback diode parallel over the coil.

If the coil is switched off, a back EMF current will flow through the flyback diode. I can imagine that a flyback diode can have a smaller maximum current rating because the back EMF current doesn't last that long.

What I would like to know:

  1. Will the back EMF current be 5A, more or less?
  2. How long does the current flow through the diode on average before it dies out? I can imagine this is depends on the properties of both coil and diode, but can there be said something about this?
  3. Is there a rule or calculation to pick a proper maximum current value for a flyback diode?
  4. What type of diode could you use best as a flyback diode if the goal is to get the smallest PCB footprint?
\$\endgroup\$
3
  • 1
    \$\begingroup\$ 5 A coil? That's a lot. It is very large or is it the contact rating which is 5 A? \$\endgroup\$
    – winny
    Oct 12 at 10:30
  • 1
    \$\begingroup\$ It was only an example. The real things I want to control are double modelrailway turnout coil. There are even coils which even draw 9A of current. Though most are around 1.5A ~ 2A. The time that the coils are activated is typically 200ms or less but that is not of relevance for the diodes. \$\endgroup\$
    – bask185
    Oct 12 at 10:52
  • 1
    \$\begingroup\$ I see. Make sure you have more than 5 A peak current rating on your diode and preferably two times voltage rating compared to your drive voltage. \$\endgroup\$
    – winny
    Oct 12 at 10:53

3 Answers 3

6
\$\begingroup\$

The initial current will be 5A if that was the current through the coil. It will decay with a time constant no longer than \$\tau\$ = L/R where R is the coil resistance and L is the coil inductance. It will not be quite that bad because the diode has some voltage drop and some of the energy will go there, but for a higher voltage coil (eg. 24V) it won't be much different. If the coil is something like a solenoid with a moving magnetic circuit the inductance will change in a complex time-dependent way during operation, so measurment of a prototype may be the easiest approach.

The diode has to be able to handle the peak current, the die should not overheat during the current pulse (transient heat capacity and some effect from thermal resistance) and the diode must be able to dissipate the heat created by the worst-case switching frequency and on/off times.

All those requirements should be satisfied at worst-case ambient temperature and taking tolerances of parts into account. The inductor takes time to 'charge' so there will be a worst-case ton/toff that generates the maximum amount of heat in the diode. To find the minimum specification, the diode must meet all 3 requirements. Obviously a diode that can handle 5A continously under worst-case conditions would easily satisfy all requirements, but that would be overkill. You can run a simulation in (for example) LTspice and directly simulate the average power dissipation of the diode if you have a diode model and know the other parameters accuraely.

For minimum PCB area, the minimum dissipation is useful, so a Schottky diode with high current rating and the minimum adequate voltage rating will usually have the lowest losses and thus require the minimum of PCB area for heat sink. At higher voltages, like 48V etc., the advantages of a Schottky diode are less significant.

\$\endgroup\$
3
\$\begingroup\$

Will the back EMF current be 5A, more or less?

Initially, the current will be 5 amps and decay towards zero.

How long does the current flow through the diode on average before it dies out? I can imagine this is depends on the properties of both coil and diode, but can there be said something about this?

It depends on how much inductance your coil has. Strictly speaking, it takes a very long time to fully die out but, if you calculate the energy held in the inductance at the time of turning the MOSFET off, you can assume a constant voltage across the diode of 0.7 volts and get a reasonable estimate. Alternatively (and my preference), use a simulator as in this example: -

enter image description here

Is there a rule or calculation to pick a proper maximum current value for a flywheel diode?

You need to figure out what duty the diode has to perform. This is based on how often you activate and deactivate the inductor. Again, using a simulator, you can pretty accurately plot the power dissipation of the diode for various scenarios.

What type of diode could you use best as a flywheel diode if the goal is to get the smallest PCB footprint?

It depends entirely on how much heat it has to dissipate over time. A small diode may get too hot but, it boils down to the diode's duty cycle.

\$\endgroup\$
1
\$\begingroup\$

After the switch (MOSFET) is turned off the current through the inductor (coil) and thus also the current through the freewheeling diode is exactly the same as the current through the inductor before turning off. After turning off it decreases. The rate of decreasing depends on the voltage across the diode and possibly other components in the freewheeling circuit (e.g. resistor, Z-diode). The higher you allow the voltage to be the faster the current decreases.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.