RusEFI Ignition Driver Circuits - Failed Coil Protection? Zeners?

Hardware inside and outside of the ECU
Post Reply
FerrariGuy
Posts: 9
Joined: Sat Aug 22, 2020 5:40 pm

RusEFI Ignition Driver Circuits - Failed Coil Protection? Zeners?

Post by FerrariGuy »

By way of explanation: *and the disclaimer that I probably don't know what the heck I'm talking about....*

Asking for thoughts:

A friend with a V12 w/ carbs has been bugging me about finding a system to run JUST ignition/COP on hs car (Electromotive apparently is in limbo for being purchased, so they're not currently adding development to their system). Besides using 1/2 of Proteus, my other thought ran to the abandoned MegaJolt/C project. I've kind of wondered about smashing some version of MegaJolt/C's hardware together w/ Proteus & run it off of RusEFI Code as a stand alone ignition system (Maybe call it the Zeus board?)... IDK... probably out of my league.

Looking into both - out of curiosity, I also happened upon this video which goes thorugh the diag of an OEM control box where the MOSFET inside has failed, and taken out the entire board with it, hence wondering about protection from coil-failure- regardless of type.


(Summary: He finds an unresponsive coil - does a back probe -sees no signal - finds a mosfet in the ECU that's failed shorted - replaces it - but it turns out it has burnt out the driver IC chip (in this case apparently directly the MCU).

As mentioned in the comments the thought is a better design could have saved the CPU and/or Mosfet control chips by adding a zener diode + small cap in parallel (faster slew rate) + resistors (already in RusEFI designs) to the circuit controlling the driver MOSFET.

A couple of the relevant comments:
Craig Sherman wrote: All they had to do is put a 3.3V zener from the output pin of the CPU/microcontroller to ground. Cathode (K) goes to the pin, anode to ground. Then use a 1K resistor from the from the pin of cpu where the zener diode is to the gate of the MOSFET. The input impedance of the MOSFET is so high that the 1K resistor would not effect it all that much - except for the turn off time. (Slew Rate) Due to capacitance between the gate and source, you may want to add a small ceramic cap of about 1uf across the 1K resistor. That would fix the slew rate issue and tighten up the off time. If a MOSFET shorts, the resistor would limit the current, while the zener diode starts to dump the current to prevent it from going above VCC of the chip. Think of the zener as a safety release valve. That would be a SMD zener and SMD resistor. Cost? Probably $0.50. The most current to flow through the resistor would be about 14ma.
Rick West wrote:I was thinking something similar. Put a 1k resistor in series with the gate to the processor pin. If the gate shorts, the current will drop voltage through acriss resistor as current flows through the input protection diode on the processor... assuming it's a standard processor. Most allow somewhere in the neighborhood of 25mA max diode current. Have a 3.3V, or 5V (whatever the power supply is) transient voltage suppressor will sink the current to ground. The TVS should be there anyway IMO. When I design circuits like this, I always include the TVS clamp to keep from destroying parts if an ESD event gets onto the regulated supply voltage.
Looking at the MegaJolt/C schematic, this is exactly what they did for their 74AC240 ignition output chips (except for the small cap in parallel):
(I"m guessing PESDxS4UD is a 3.3v zener out of the available ones in the series).
image.png
image.png (21.62 KiB) Viewed 500 times
The RusEFI system & Proteus more or less use a MIC4427 driver chip (instead of something like 74AC240). There are resistors on the backside of the IC between it and the MCU as well as on the Ignition coil side, but no protective zeners.

Neither PDF for either chip mentions any inbuilt protection so....
image.png
image.png (89.3 KiB) Viewed 500 times
So, what are the thoughts about protecting the coil driver ICs with Zeners? (I fully acknowledge I may totally misunderstand this whole thing...)


~Paul K.
Post Reply