filtering circuit for automotive pic supply

[Simon]

I'm designing a pic based circuit for use in a car. now I'm having trouble with the 78L05 regulator I'm using to supply the pic as it keeps blowing, the obvious reason being that its being caused by all the spikes in the car power system.

My plan now is to eliminate the spikes and dangerous transients etc with filter circuitry. but I'm unsure how to go about it,

something like: power in > inductor > 24 V zenner > inductor > capacitor,

somewhere in there I could ad a specific component to eliminate spikes etc (I'm looking into it), my main concern is the value of the inductors, I'm supplying up to 3 A.

[desolatordan]

I would just leave the inductors out, but then again I don't know any better.

If your supply is nominally 12V, I'd choose a lower zener like 18V. Your 78L05 is probably rated to 30V DC input.

[Dago]

A TVS might be a good idea, better (higher peak power) than zeners at least.

[charliex]

I'm using S10K30 for transient, S1G for reverse, BZG03C30 filter. But i'm not supplying 3A

[qno]

Please place a schematic here.

I usually use a simple RC filter to dampen the spikes to the "low current" part of the circuit.
Assuming the 78L05 draws 100 mA a resistor in the range of 10 to 27 ohms with a capacitor of 470 microfarad will have a corner frequency < 20 Hz and will loose only a few volt as input for the 78L05. As long as the 78L05 has more than 8 volt on the input you are OK.

If there are a lot of spikes these will be mostly dissipated in the resistor. Use a sufficient large resistor to handle the heat generated.

If you switch inductive loads use damping diodes and quenching circuits.

Did you measure some of the spikes you are dealing with?

 

[Simon]

The main issue here is bad quality supply caused by the car ignition system which being a vintage car is worse I am told, I have not measured the spikes yet but I would think they are quite fast and hard to detect.

The back emf on the feild coil with the original regulator is -70 V and is quite a long impulse lasting pretty much all of the off time. on my electronic version I have used a clamping diode but am also thinking of putting a resistor across the mosfet controlling the field coil so that the of period still supplies current to the feild coil but not as much and hopefully makes for a smother supply (although this is neither here nor there as the main issue is the ignition spikes doing damge)

[qno]

Maybe a fast recovery diode in reverse to protect your circuit for the negative voltage. Take care that the circuit has enough buffer capacitance to supply your circuit wit power.
Also layout of cables could help. If you connect the power directly to the battery.

[Simon]

the power is coming off the ignition switch, it may be put on the dynamo itself as that seems to work but it still blew then

[qno]

no not the ignition switch.
Use a direct connection to the + and - of the battery.
The battery has a very low internal resistance preventing spikes.

If your circuit then works you can look for the source of your problem.

[Simon]

the circuit needs to be switched on when the engine is started, I'm going to wind an inductor and use a 1000 uF capacitor to dampen spikes along with a TVS ans zener