SXE-Probe VFP1000 Oscilloscope & Digital Multimeter

[nobbyclrk]

I should of noted both these captures are taken from the coil leads on each vehicle.

Oh what the hell! While we are at it, I shall set a few things straight.

First of all, this pattern is actually upside down. It's just the traditional way of looking at distributor type ignition systems. And that's the way it is.
Second. Where I have pointed out the coil has been switched on, what we actually see in the secondary pattern is the induced voltage created by the magnetic being built up by the primary circuit. But generally the primary pattern mirrors the secondary (and vice versa) although the voltages will be different of coarse.

And last of all, the first capture (Red trace) I put up is actually from a Toyota Corrolla with a similar ignition system. I used this because it has a lot higher sample rate than the one taken from a Discovery (Corrolla 2million samples a second, the Discovery is only two hundred thousand). The sample rate is the number of voltage measurements taken per second and can be seen in top right corner of each image.
I posted the image of the Discovery up for the benefits of any purists out there crying 'charlatan'. There is no notable differance in either capture. The same physics apply to make a spark, they will be, are the same. The only really noticeable differance is the current limiting on the Discovery is alot shorter but being a V8 it has to generate twice as many ignition events as the Corrolla. Obvious really.

Enough of secondary ignition basics. Yawn! I did say the detail you can see with the right gear is impressive. Some will be impressed by what is able to be seen so far (so hang on to your seats). With enough practice you can spot such things as burnt valves, low compression and weak or lean fuel mixtures. In the Discovery image above you see the result of a lean mix, the upward sloping sparkline as the cylinder runs out of fuel (fuel increase's conductivity therefore requires less voltage to sustain a spark). This Disco was actually back a few weeks later for a weak fuel pressure regulator.

Time to get the spy glass out. Note in the Corrolla capture where the time marker is, about half way along the screen at 276.3ms. Right where the voltage breifly reaches 9Kv as it builds the voltage up to ionize the plug gap.

[nobbyclrk]

Let's magnify that 1000 times (this is why I needed to use a high sample rate capture). You can see the magnification in the top of image, just below tools.



So what is this?
What appeared as a single time marker was actually two. Those two in the above image. And this image is of that 9Kv firing line.

I did say we would see detail. But what detail is there to see.

In this line we can see the voltage rise in the secondary circuit to a point that ionizes the plug air gap at 9Kv to make it conductive enough to allow current to flow, nothing new there. This actually took 35us or 0.000035 seconds to achieve and that votage/time graph has just been stretched to fill the screen.
But not only can we see how much it took to jump the plug gap. We can now see how much it took to jump from the rotor button to distributor.
It's that notch half way up the slope at 5.43Kv and 23.29us into the event. Now I can see how good your rotor alignment actually is. Period.
Although you have just taken a very long way round to diagnose that. In the real world it's just not feasible or realistic to diagnose rotor buttons like this, but that's not what this post is about.

Now I'd call that detail.

[nobbyclrk]

And here's the proof.
The red trace is the coil lead to the distributor, the green is the plug lead from the dizzy.



I used to consider the length of time a spark lasted for was insignificant on a time scale. Now I know it's a lifetime in efi management.
Then I used to think it was incredible that I could see the firing line and rotor notch. But actually I hope to show engine management events that make the build of voltage in the coil lead look like a stroll in the park.

I'll try and get some images togeather over the next few day starting with an example of what real injection time is compared to injection pulse width.

[nobbyclrk]

I hope everyone is having an excellent Christmas break

Here is a capture of a saturation type petrol injector. Here I want to show measured injector on time compared to actual open time. This is actually off a 5 cyl S60 Volvo, but again it is (for the purpose of this post) the same as a Discovery petrol injector. The most likely difference will be whether or not the inductive kick is 'clipped' or not on the Discovery.
You will find most engine managements systems will operate in a very similar way although generally each manufacturer will have their own unique approach.

Once again I'll explain each componant of the waveform over the next couple of days.

[nobbyclrk]

There's not really a heck of a lot to say about petrol injectors. Especially this type (saturation) which is as basic as it gets. The PCM grounds the injector, the internal windings becomes magnetised and acts on the pintle which is being held in it's seat by spring pressure. Once the magnetic force becomes great enough to lift the pintle against the pressure of the spring the injector opens and fuel will flow. When the injector has been on for the required time the PCM open the circuit, the current flow stops and the magnetic field collapse’s and the spring forces the pintle shut and the fuel stops. Pretty simple really.
Just a couple of points regarding petrol injectors. The first is they are slow, incredibly slow (which I'll show another time). Even though it's lasts (hot, idle) on average only twice as long as a spark, about 2.5ms. The average spark last 1.3ms.
The second is the on time relevant to the open time which are two different things. Most of the time it's not an issue, if you see a pulse width greater than say 3.5ms you know it's running rich. If it's less than 2.0 ms you can say it’s running lean. So who cares about the relevance of pulse width to actual open times. To be honest, no one.

[nobbyclrk]

But pulse width is one thing, that’s the electrical side of it. You want to know if anything happened mechanically. You can see it all. The same way a tooth on an ABS reluctor rings disturbs the magnetic lines of force across the windings in the ABS sensor and generates a voltage, so does the pintle in the injector.
You can see the effect at point A on the current ramp (I’m not going into the physics on this). That is the point that the pintle is moving and disturbing the magnetic lines of force created by the electromagnet in the injector.
Also once the PCM turns off the power to the injector the magnetic field collapses. Then comes a point where the spring that seats the pintle starts to overcome the magnetic field and the pintle moves back in place, disturbing the magnetic field again. We can see that at point B.
Now we know how long the injector is actually open for, but as I said that in it’s self is of no use. But what we do know now is that it actually physically opened. It’s not stuck. And depending on the system, you can compare the individual injectors to confirm they are all reacting the same way for the same duration. They should be very similar waveforms.
A couple of other points to note is the electrical behaviour of the injector is very similar to the ignition coil primary. The ‘pintle hump’ is probably the biggest difference in the capture. The way we use the magnetic field is what differs between the two. In the injector we use the magnetic field to move something physically. Where as in an ignition coil we the voltage created (inductive kick) by the magnetic as it collapses over the windings once it is turned off to create a spark (together with the secondary windings of course).
In this injector pattern we can see that kick is being controlled (that is the flat top on the spike). Not all injectors are controlled this way. It’s done by using a zenar diode which works by allowing a voltage above a certain thresh hold through the opposite way it otherwise normally would (way better explained here [ame="http://en.wikipedia.org/wiki/Zener_diode"]Zener diode - Wikipedia, the free encyclopedia[/ame]). This is done to control the speed the injector closes. Slowing the speed down prevents the pintle from bouncing back open again and also reduces the hammering effect on the seat therefore reducing wear in the injector.

As I said, not really that exciting the basic petrol injector. Electronic diesel injectors and the management system is way more impressive. Although I know alot less about electronic diesel.

[nobbyclrk]

So I've shown a petrol injector I thought I may as well show a diesel injector from a TD5. We shall see it's controlled in a completly different way than a petrol injector.
So here's a couple of images to get started with. In both the (the second image is a zoomed in section almost halfway along the first) colours are as follows.

Blue - Crank angle sensor
Red - Supply to all injectors
Green - Signal side of injector #1
Brown - Injector (all) Amps


The first image is self explanitory. It's a breif overview of the PCM controling the injectors from start up, idle, fast acceleration, over run and back to idle.




This second image is a zoomed in portion of the above and shows one engine cycle or 720 degrees of crank rotation - we can see that we have six injector events and we can see the difference in Signal side of injector #1 at each end of the image compared to the other injector patterns.




Straight away we can see some differances between the diesel and petrol injectors. Voltage is a lot higher (approx 80 volts) and current reaches 7 amps and the shape of the waveform is hugely different.

[nobbyclrk]

And here's a close up. Note the differance's and similarities between these two events. This system is a lot more complicated electronically so may take a bit of thinking and my explanation skills can get a bit rocky at times, but hopefully we shall get there.

[nobbyclrk]

In this image I have just over one engine cycle or just over 720* of rotation.

In the crank sensor trace you can see 31 peaks and 5 open areas. Each peak represents the hole in the flywheel passing the sensor.
(Normally I would expect the the voltage to only rise if a tooth arrives, unless I have probed the wrong wire, and fallen as a hole approached.But in this case a signal was only available on one wire. All to do with the programming and engineering development).
Those open areas are the same duration as one missing peak, so 36 peaks for 360* thats one for each 10* of rotation. Which, as I confirmed yesterday, is what Rave has.
So now we know I have 1045* on the screen, accurate within a few degrees - easy.



One area of interest is the second hole in the group of nine, marked by an asterix. It's alot higher than the rest. This same hole was evident on another TD5. This indicates to me it is normal not a problem (comparisons of other examples always good to have). I suspect that hole is actually deeper than the rest. Weather or not the PCM uses this as some sort of sync I don't know.

Another point worth noting, is the hole arrangement. It's in five groups, 7 - 4 - 9 - 5 - 6. This lets the PCM know which cylinder is at TDC . But it can't work out weather it's TDC compression or TDC exhaust. Although by this point it has figured it out and we can see what group of holes lines up with which injector event. 7 holes for injector one, 9 for injector two and so on. These won't be the hole id's the PCM uses as it occurs after the event, in other words you can't count 7 holes total until after the injector has fired - too late! But it probably uses the six holes before. Clear as mud.

Again, what does all this mean? And again, by itself nothing. But now I can see where the engine is in it's cycle fairly accurately. What injector should be firing, weather it fired or not (although an alternator does a far better job of that great for diagnosis) and RPM.

As anybody still reading this thread has probably worked out I enjoy looking at engine management componants this way. It gives a more definitive approach and you can see what is actually happening. Although when using a scope you can equally go on a very long trip for a simple fix. But you will probably learn a heck of alot too.

[nobbyclrk]

Hopefully what I have explained will show it is not impossible to figure out what the PCM is looking for under certain conditions. What I have here is the engine starting and you actually see the PCM trying to work out which cylinder is on it's exhaust stroke and which is on the compression stroke and should be having fuel injected.
I just want to say at this point, this is my perception of the fuelling stratergy and could well be incorrect. Alot of the times you have to make judgements and can get it wrong. For example I thought the crank signal indicated teeth, it never occurred to me it could be holes.

This first image is an overview so you can see where in the start up sequence the following images fit.