Circuit switches a 4mH coil. Is there a way to arrange the power switch so when it switches on/off it doesn't cause high voltage transients?

Ideally without adding any components.

 

No.
Sudden interruption of the current through an inductor will generate a back-emf voltage.
You can add a diode across the coil to catch the spike and limit it to <1V.
You can use a zener diode to allow a controlled voltage and so make the spike decay quicker.

 

"Without adding any components", the way to do it is to replace the power switch with a variable resistor. Then as the resistance is increased the current will drop and the voltage rise will be very small . That is the way without adding any parts, only changing one part.
An alternative is to place an inverse diode across the switch so that when it opens the voltage spike is returned to the power supply. Or you can add an indicator light across the coil, which will serve as a path for the current and avoid the big voltage spike..

 

No, that diode may keep the pulse out of other places, but the inductive pulse voltage is developed across the terminals of the coil when the current is changed, including being dropped to zero. When a switch is opened the rate of change of current is very large and so the developed voltage is also large. so the two ways to reduce the spike are to either slow the rate of change or to provide another path for the current to flow in. That is what the diode across the coil does. In the energized state of the coil the diode is reverse biased, but then when the switch is opened the voltage polarity reverses and the current flows through the diode, and so the high voltage is not generated. The current flows for a short time, so the magnetic field collapses more slowly. It takes several milliseconds for it to totally collapse. Usually that is not a problem.

 

Does that include away from the battery and regulator?

The diode will tend to keep part of the pulse away from those devices, BUT if the inductance rings even one cycle there could still be some voltage get through. But once the switch begins to open there is no connection except through whatever spark may be going on inside the switch. So a diode across the coil connections is the best way to go. You could also use a snubber, which is a resistor and capacitor in series, across the coil. The advantage of a snubber is that it is not polarity sensitive, either direction for connection is OK and works well.

 

I see two other solutions:

• Diode from Coil + to Ground.
• Let the Switch control Reset on U1.

But I also see that there are almost totally missing decoupling capacitors on various supplies.

 

I had not realized that the purpose of the coil was to generate the high voltage, so forget the advice about either diode. What you need is capacitors in parallel with that SMB device at the power input to the system. 0.1mFD and then some larger value as well. ALSO, quite important are the 0.1mFd capacitors across the regulator ICs, from input to common and output to common. Without those you can have some serious oscillation on the output of the regulators. I know this is true because it happened with a circuit that I built once. The frequency was several megahertz, so it affected everything in my system. So those capacitors are quite important.

 

When the coil current is switched off, the top end of the coil will swing negative, forward biasing the diode. It won't stop anything.

 

Added 1 ceramic and 1 electrolytic 50v rated capacitors.

Is this what was recommended to prevent transients to the battery and switch?

there is no high voltage transient but the electrolytic has high peak current. can it damage the capacitor?

The two capacitors serve different purposes. The lower valued one, 0.10 microfarad=100nf, serves to provide a low impedance for the fast edges of spikes, while the higher value electrolytic stablizes against slower but wider pulses. AND the 0.1 capacitors on the voltage regulator ICs are straight from the manufacturers application literature, and I have verified that they are needed most of the time.

 

The ceramic volt reg and electrolytic capacitors need to be laid out on the board as close to the volt reg, SMB and mosfet source?

The ceramic and electrolytic capacitors should be on the trace before other parts of the circuit. The requirement for being close is for the 0.10 microfarad caps for the voltage regulators. The applications notes usually say within 0.1 inch, about 2mm. But my experience has been that withing 0.25 inch of trace is OK. And those very small caps are simple to place.

Now I am also wondering about the purpose of the HEI ignition coil, since the main use is for running engines. What sort of interesting othern application have you found for that part?

 

What sort of interesting othern application have you found for that part?

Most of his threads are leading toward Tesla coils and HV. The HEI coil is just an ignition coil that has improved magnetics. Instead of the old style "can" type coils with no complete magnetic path, the HEI uses a E-I transformer core that has a complete magnetic path. The complete magnetic path adds to the coil output.

 

Most of his threads are leading toward Tesla coils and HV. The HEI coil is just an ignition coil that has improved magnetics. Instead of the old style "can" type coils with no complete magnetic path, the HEI uses a E-I transformer core that has a complete magnetic path. The complete magnetic path adds to the coil output.

OK, it is just that any time somebody has an alternative application for something with one intended use I find it interesting. Alternative approaches were a part of my career for a while.

 

Alternative approaches were a part of my career for a while.

They have for me too. where I served my apprenticeship, we did things with machine tools that other shops wouldn't even bid on because they couldn't think out of the box. I still do the same today with things.