Sizing of coil flyback Zener for inductive kick

[ricko_uk]

Hi,
with reference to the attached schematic, I am designing a coil driving circuit using a flyback zener. Only one of two coils can be connected to it at any given time. Either a 40mH coil or a 160mH one. Both with a DC resitance of 25R. The maximum switching frequency is relatively low at 100Hz at most.

It is important that it the inductive kick (when the mosfet turns off) is as fast transient as possible (hence using the Zener instead of just a freewheeling). Is that the best way to create a large "as fast rising as possible" inductive kick?

Could someone please help with the following:

1) I assume the higher the zener voltage the faster the decay, correct?

2) what is the logic for choosing the zener voltage?

3) What is the correct way to choose flyback Zener power? I assume the maximum instantanous Zener current is the ~1A (24V / 25R). Correct? But there does not seem to be any pulse current ratings in the Zener diodes datasheet (even the power ones like SMC packages). How do I select the Zener power?

Thank you as always

[johansen]

Zeners dont handle this well in my experience.

If you are able to deal with slow solenoid response in the first handful of cycles, you can use a relatively large capacitor to absorb the energy using a diode to charge it Then use a zener to dissipate the energy while keeping the voltage at whatever voltage you want. Or build a buck converter to dump the energy back into the supply.

The voltage is a trade off between the cost of the switch and the performance expected of the solenoid.

You can use a 600 volt igbt and let the voltage rise to 500 volts but its diminishing returns past.. im guessing 4 times the nominal voltage used to energize the solenoid.

[Benta]

Use a protection/transorber diode instead, it's made for this job.
Place it across the MOSFET D-S.
Typical types are P6KExxx (600W) or 1.5KExxx (1500W). P6KE is enough here, perhaps you can find smaller types as well.
As for voltage: whatever your MOSFET will withstand. The higher, the faster your current decay.

1: correct.

2: see above.

[PCB.Wiz]

. How do I select the Zener power?

Spice would be useful here, as it can easily show the current curves, and also the powers at 100Hz

There are multiple clamp choices.
You may need to trade off power and time.

Direct across the mosfet gives triangle currents, but highest clamp power loss.
Across the coil lowers clamp power, but supply currents have sharper edges.
A zener + diode allows zeners below Vcc, a bidirectional clamp is fewer parts, but must be rated above max possible Vcc.

Addit : a rough ballpark is given by
0.5 * (.160 henry) * (1 ampere) * (1 ampere) * (100 hertz) = 8 watts
Spice will give more accurate clamping power numbers, as at 100 Hz reaching 1A may be difficult.

[ricko_uk]

Thank you all,
it looks like the TVS + IGBT might be the most robust solution.

Thank you

[langwadt]

. How do I select the Zener power?

Spice would be useful here, as it can easily show the current curves, and also the powers at 100Hz

There are multiple clamp choices.
You may need to trade off power and time.

Direct across the mosfet gives triangle currents, but highest clamp power.
Across the coil lowers clamp power, but supply currents have sharper edges.
A zener + diode allows zeners below Vcc, a bidirectional clamp is fewer parts, but must be rated above max possible Vcc.

if the mosfet can handle the energy you can also use a zener from drain to gate (with a gate series resistor)  like in self clamping fets

[ricko_uk]

Thank you.
Yes, also a good point/solution. But I think a TVS instead of a zener might be better (i.e. more robust) in this configuration too (i.e. across the switch).

[Benta]

A drain-gate Zener works fine, but is counterproductive here, as the goal is high speed turn-off.
This will only happen if you can burn a lot of power in a short time, which points to the drain.source TVS.
The drain-gate Zener will actually slow things down by keeping the MOSFET on.

[langwadt]

A drain-gate Zener works fine, but is counterproductive here, as the goal is high speed turn-off.
This will only happen if you can burn a lot of power in a short time, which points to the drain.source TVS.
The drain-gate Zener will actually slow things down by keeping the MOSFET on.

everything other than nothing will slow things down. A drain-gate zener will do exactly the same as a drain.source TVS
 

[Benta]

It won't dissipate any significant power. The TVS will.

[langwadt]

It won't dissipate any significant power. The TVS will.

if the voltages are the same the only difference is where the power is dissipated, the fet or the TVS

[Benta]

True.