Analysing Impact of Voltage Drops on Ignition Coil Operation

Dmitriy Analyst Toronto, Ontario Posted   Latest   Edited  

Part 1: INTRODUCTION

This is a continuation of discussion between Rudy, Ray, and me that started in the "Oldie but Goodie" thread a little while ago. I have done some analysis and circuit simulations and would like to present results here.

Rudy captured the following primary ignition voltage waveform and was wondering whether it is good or not:

diag​.​net/file/f43u4i6lv…

It is not exactly ideal, with the initial voltage drop of 0.8V, increasing to 1.45V after 3.2ms. Ray quickly pointed that out by providing a known good waveform for the same engine that shows zero initial voltage drop: 

diag​.​net/file/f5yzzh52a…

But is it a problem that prevents the vehicle from running? Is this something that we should fix before chasing the P1300 code? Let's take a closer look.

Firstly, let's zoom into Ray's capture (many thanks to Ray for providing the psdata file):

diag​.​net/file/f165wprds…

Oh, so there is 1.2V voltage drop after 3.3ms. Important value, but the 400V scale was hiding that from us in the original picture. We will need it later.

Now, let's take a look at ignition schematics and see where voltage drops can occur:

diag​.​net/file/f2x1se1t6…

The logic module provides a signal to the Ignition Amplifier, and it, in turn, provides the ground to the coil, resulting in a build-up of high current through the coil within a short period of time. I think this schematic covers most ignition systems, though the Ignition Amplifier can be a part of the PCM (case B), or part of the Coil-On-Plug (COP) (case C). If you know any exceptions, please let us know in the comments!

Here we will assume that the power side is perfect and its voltage drop is zero: ΔV_P = 0. However, during the dwell period, the primary voltage is the sum of voltage drops across the Control Wire, Ignition Amplifier, and the Ground Wire: V_C = ΔV_CW + ΔV_IA + ΔV_G.

It is very important to distinguish between V_C and the voltage drop across the Ground Wire, ΔV_G (by the way, for schematic C this is the only value on the ground side that we can measure). I believe it is most important to keep ΔV_G low, while there is more leeway for V_C. Thus, here are my claims:

CLAIM 1: It is very important for ΔV_G to be small, and that includes reasons that are not directly related to reduction in the coil current.

CLAIM 2: If the voltage drop over the ground wire is negligible (ΔV_G=0) and V_C < 1.4V at the end of the dwell period (for example, 3ms) then the coil current is going to be at least 90% of the ideal current (the current achieved when there are no voltage drops at all). This holds even if the initial voltage drop is not zero.

In this post, I will address Claim 1, and Claim 2 will be demonstrated by circuit simulations in the Update post.

The Ignition Amplifier has electronic components inside, and significant enough ΔV_G will affect its operation. Firstly, a sudden increase in ΔV_G is equivalent to a sudden drop in the power supply for the Ignition Amplifier -- not something electronics likes. Secondly, the Logic module sends 5V (relative to a good ground) signal, but the Ignition Amplifier will see it as 5 - ΔV_G and may fail to trigger at all, or may trigger erratically.

I think this is what happens in another capture posted by Ray:

diag​.​net/file/f1flz5xbh…

There are three types of signals:

  1. Good, achieving 10.5A.
  2. Not good, with significant ground voltage drop, but not enough to confuse the electronics in the COP. The current achieved is 8A.
  3. Really bad, with the electronics resetting once the ground voltage drop reaches 3.5V, trying again and reaching only 4.3A.

Would you agree with this explanation? Anything that I am missing? 

Now, how is this related to Rudy's case? Rudy provided a separate ground to the Igniter and saw no change in the waveforms. So, we can assume ΔV_G=0 in his case. Phew!

Well, this is it for today. Claim 2 is not an easy one, but I really would like to hear your feedback about the setup in general and about Claim 1 first. Hope for a good discussion.

I will post an update with circuit simulations soon, so, if you are interested in the subject, please press 'Watch' to be notified by e-mail.

+9
Dmitriy Analyst
Toronto, Ontario
Dmitriy Default
 

PART 2: CIRCUIT SIMULATION Suppose we checked that ΔV_G=0, so the electronics inside the Ignition Amplifier works without any glitches: diag​.​net/file/f2x1se1t6… During the dwell period, voltage V_C will be the sum of voltage drops over the Control Wire and the Ignition Amplifier. These voltage drops can have different nature: - Resistive. Examples: resistors, wires, and…

+2 Default Ð Bounty Awarded
Dmitriy Analyst
Toronto, Ontario
Dmitriy Default
 

PART 3: PRACTICAL CONCLUSIONS Is all this just a mental exercise or there is a practical side to it? I hope the latter is true! Please help me formulate something that can be used in the field. Here is my attempt: It seems, voltage drops are not created equal -- some are worse than others. The examples above show that voltage drops over the ground wire of modules are particularly disruptive…

+3 Default Ð Bounty Awarded
Rudy Technician
Montebello, California
Rudy Default
 

LoL Nice!! Here I thought the difference in the ground drop captures was simply Rays superior scoping ability. Excellent write up and very thorough explanation

0 Default Ð Bounty Awarded
Dmitriy Analyst
Toronto, Ontario
Dmitriy Default
 

Thanks, Rudy, we might never know what is different inside those igniters, but they both seem to work. Now, there seem to be a voltage drop competition going on as Tanner B. has recently beaten your record by posting a "known good" primary voltage waveform with even more remarkable voltage drops: diag​.​net/msg/mw0okrk4zm… diag​.​net/file/f46muiohv… The primary…

0 Default Ð Bounty Awarded
Chris Diagnostician
Lansdale, Pennsylvania
Chris Default
 

Dmitriy, Excellent write up as usual. I think you touched upon a topic that has affected all of us at one time or another while scoping components: failure thresholds. Failure threshold has always been something that I struggled with; my capture is varied from known good, but is it enough to cause a fault and how best do I prove that out? This is applicable from ignition coils to injectors…

+1 Default Ð Bounty Awarded
Dmitriy Analyst
Toronto, Ontario
Dmitriy Default
 

Really good way to put it, Chris, it is all about identifying failure thresholds. Some of those come from physical processes involved, some -- from safety margins or requirements put in place by engineers. I hope I can contribute to such rules of thumb as there is usually some math involved to create them.

+1 Default Ð Bounty Awarded