One person commented in 2021 about a zener's dynamic resistance makes me wonder if there might be more to this.
(1) Answer Log (quora.com)

The fundamentals regarding snubber circuits has'nt really changed much toward designing counter measures for dealing with noise.
The example page 8 followed by a comentary on application and useage of zeners.
"Seminar 900 Topic 2 - Snubber Circuits" (ti.com)

 

One person commented in 2021 about a zener's dynamic resistance makes me wonder if there might be more to this.
(1) Answer Log (quora.com)

The fundamentals regarding snubber circuits has'nt really changed much toward designing counter measures for dealing with noise.
"Seminar 900 Topic 2 - Snubber Circuits" (ti.com)

Hi, I enjoyed reading seminar 900 but it did not appear to include a wealth of info on snubbers comprising Zeners. I did learn, however, if my understanding is correct, that my original circuit was passive and dissipative. Whereas @Danko's circuit is non-dissipative and active. I'm not sure if active is fully correct because it piggybacks the load switch and doesn't comprise any switches dedicated to snubbing.

Would you please elaborate on your first point regarding Zener dynamic resistance? It is integral to the negative feedback circuit? I would like to understand further.

 

Danko's circuit has the advantage of the the inductor current not going through the Zener(s) so the Zener does not need a large current rating.
The feedback to the gate keeps the MOSFET turned on just enough to carry the inductor current until the current stops.
No.
The MOSFET does not turn all the way off.
The feedback is just enough to keep the MOSFET on and conduct the inductor current with the MOSFET drain going to near the Zener voltage.

Below is the LTspice simulation of his circuit, slightly modified to use the Zener models I had.

As you can see, when the IN voltage (blue trace) goes to zero at 6.5ms, the MOSFET gate voltage (green trace) drops just enough to allow the MOSFET to stay turned and carry the inductor current (purple trace) during the transient, due to the feedback through the diodes.

The MOSFET drain voltage (red trace) jumps up to the sum of diodes D1-B and D2 voltage plus the MOSFET gate voltage during this time.

The max. diode current (white trace) is <800µA, so a low power Zener can be used.

D3-A is actually not needed, since the gate voltage never goes above the input voltage.

View attachment 262459

@crutschow @Danko , is there any downside to this solution? I am curious why anyone would prefer to send the current ahead of the solenoid directly through a Zener, or to ground directly through a Zener? Sending it through the MOSFET seems like an obvious improvement to avoid the surge wattage seen by the Zener.

Also, is this solution documented anywhere I can read up on it?

 

@crutschow @Danko , is there any downside to this solution? I am curious why anyone would prefer to send the current ahead of the solenoid directly through a Zener, or to ground directly through a Zener? Sending it through the MOSFET seems like an obvious improvement to avoid the surge wattage seen by the Zener.

Also, is this solution documented anywhere I can read up on it?

Yes, it is documented, see attachment:

 

Whereas @Danko's circuit is non-dissipative and active.

Only half true.
It is active and dissipative.
You always need to consider where the energy goes since the Conservation of Energy Law means it doesn't just magically disappear, it has to go somewhere.
Here, the inductive magnetic energy is dissipated as heat in the MOSFET instead of the Zener.

Look at the MOSFET dissipation below (red trace) during the turn-off:
Although the peak power is high, the energy dissipated is only about 55mJ.
You multiple the pulses per second times 55mJ to get the average power dissipation (thus 10pps gives an average MOSFET dissipation of 550mW).

 

is there any downside to this solution?

None that I know of.
It's not always used because a discrete MOSFET is not always used to control a solenoid, and so the control circuit design may not be amenable to that approach for spike suppression.

 

Yes, it is documented, see attachment:

Excellent thank you. I'll read through that shortly.

 

Yes, it is documented, see attachment:

I am not quite sure why Vin Rg1 and Rg2 are arranged as they are in the spec. I understand that the drain rises and the current flows through the Zener stack but why does it say "thru Rg1 to ground" without mentioning it flowing through Rg2?
Why are Rg1 and Rg2 in series what is the purpose of their separation?
Why is the diode with the cathode facing the gate a Zener?
Why is the Vin node directly connected to ground? Wouldn't that be a short?
I like the way my circuit is behaving and now and I want to understand why the spec shows a different arrangement

 

I am not quite sure why Vin Rg1 and Rg2 are arranged as they are in the spec.

If you look at the un-simplified models, you will see the two resistors, which are apparently internal.
Obviously, from a circuit standpoint, since they are in series, they function as one resistor with a value equal to the sum of their resistances.

Why is the diode with the cathode facing the gate a Zener?

Those Zeners are internal and it likely just makes the fabrication easier to make them both Zeners, rather than to make one a Zener and one a standard diode.
Remember that a Zener in the forward direction acts just like a diode, so the two function the same as if the bottom one were a standard diode.

Why is the Vin node directly connected to ground? Wouldn't that be a short?

To show the operating state where the input voltage is zero, and the MOSFET has just been turned off and clamping the inductive load.

 

I think that operating point is valid and the clamp working. The zener diode's dynamic resistance can be calculated.
The shape of the graph is fixed however it translates to the left as temperature rises.
Since this shift is due to temperature, a snubber circuit using a zener diode would logically
use some kind of temperature compensation. The factors for stability would include the dynamic resistance
and the compensation needed to maintain during change in temperature. Data logging this could result in
finding a range which ends up somewhere on an I/V chart or operating point for a given load. When that circuit goes out of tolerance
it would be logical that it would be slow/not working. More specifically the diode's junction temperature for a given load would not exceed the ability of the compensation circuit to operate with a specified range.

 

How it works from -40 to +140°C :

 

If you look at the un-simplified models, you will see the two resistors, which are apparently internal.
Obviously, from a circuit standpoint, since they are in series, they function as one resistor with a value equal to the sum of their resistances.
Those Zeners are internal and it likely just makes the fabrication easier to make them both Zeners, rather than to make one a Zener and one a standard diode.
Remember that a Zener in the forward direction acts just like a diode, so the two function the same as if the bottom one were a standard diode.
To show the operating state where the input voltage is zero, and the MOSFET has just been turned off and clamping the inductive load.

This all makes sense now thank you.

 

How it works from -40 to +140°C :
View attachment 262608

Hi @Danko I simulated it and I see the temperature stepped -40 to 140 and the Drain voltage didn't change much. I am trying to understand the takeaway-
Can we conclude the temperature compensation sparky 1 mentioned is likely not necessary for the temp domain -40 to 140?And the changes to the negative feedback performance due to temperature are negligible?

 

None that I know of.
It's not always used because a discrete MOSFET is not always used to control a solenoid, and so the control circuit design may not be amenable to that approach for spike suppression.

I built this active dissipative circuit and there were oscillations and the dissipation of the kickback was very long. Do you have any idea what could cause this? I noticed the gate voltages are above the mosfet ratings.

Yellow is the drain and magenta is the gate. Gate trigger is 10ms, 3.3v on from the microcontroller, just like the sim.

This is without R4:


This is with R4:


Zoomed in; wave periods are shown at bottom, about 12 ns:

 

It looks like the inductor is oscillating with the stay circuit capacitance.
You might try reducing the value of R4 across the solenoid.

 

If the application can stand a sight sowing of the solenoid turn-off, stretching the switch-off time to 10 or even 20 milliseconds will reduce that inductive spike a lot.
But then I saw back close to the start that the target is a one millisecond switch-off and solenoid release, and then much farther in that it is a fuel injector driver circuit. And also that the TS does not want to change the circuit, only to adjust values. So now I am guessing that this may be a product design and certainly not a one-off project.
Changing the transistor to one with a much higher voltage rating will allow the same circuit to work without breakdown issues.
Another choice is to use a much higher voltage rated transistor and simply protect the output of the controller board. There do exist other ways around the problem. A 1000 volt device certainly is available.

 

Can we conclude the temperature compensation sparky 1 mentioned is likely not necessary for the temp domain -40 to 140?And the changes to the negative feedback performance due to temperature are negligible?

In your case small change (±0.6В) of V(d) does not matter.

I built this active dissipative circuit and there were oscillations and the dissipation of the kickback was very long.

Try to connect capacitor 0,01 μF - 0,22 μF between Gate and Source of MOSFET.

 

In your case small change (±0.6V) does not matter.
View attachment 262673

Try to connect capacitor 0.01 μF - 0.22 μF between gate and source of FET.

OK I will add the capacitor tomorrow, scope it, and report back. Thanks!

 

I was going to suggest a G-S capacitor but 100pf to 1000pf.

 

Electronic fuel injection has been around for many years. One option might be to see what others have done. Even some high-reving two-stroke motorcycles have used it, I think. Looking at one of those circuits may be very educational.