Board logo

Misfire - Oscilloscope Analysis
gingerprince - 7/9/17 at 10:13 PM

My MK Blade started playing up a while ago, running on fewer cylinders on the road when hot, and idle misfire when hot. Swapped to different coils, new plugs, new HT leads, new plug caps, cleaned carbs, verified fuel delivery rate from pump, still running odd.

So bought a PC base oscilloscope to see what's going on, but don't really have much experience of what to look for. I know it can reveal ignition and/or fuel issues depending on waveforms, but it's all just a bit odd. Not sure whether it's cheap scope syndrome, or whether there's an identifiable issue here.

Can someone with scope diagnostic experience please take a look and let me know what you think? Cheers.


mackei23b - 8/9/17 at 07:35 AM

Hi there

Personally I would not have expected to see this amount of variation. No 1 there is also more variation with the spark duration as well....

It would be worth putting the scope on another engine just to see the difference to check the scope as well?

Cheers

Ian


gingerprince - 8/9/17 at 11:44 AM

Problem is finding another car that has HT leads and not stick coils. The variation is only really on the HT side so couldn't measure on any of our tin tops. I agree I need to find a way to determine whether the scope is sh1t though


mackei23b - 8/9/17 at 12:23 PM

Quite a nice clip here as well: https://www.youtube.com/watch?v=SUShO72Grq8&t=391s

Cheers

Ian


rusty nuts - 8/9/17 at 04:51 PM

Not sure but I think Picoscope's website shows some instructional videos which may be of help ?


jossey - 8/9/17 at 06:34 PM

Is the airbox missing?


martinnitram - 8/9/17 at 06:58 PM

Maybe a too obvious answer, earths ok from coil ?
Sure you already know that resistance rises as things get hotter.
Broken wire moving around when insulation gets hot ?

Do you think its related to temperature ?


obfripper - 8/9/17 at 08:56 PM

With your setup, every other ignition event captured is a wasted spark event, as your ignition fires the plugs as pairs, so only one is actually igniting fuel, the other is sparking on the exhaust stroke - this will give a different waveform to the ignition event.
Your captures may be better with 2 consecutive spark events on each, the lower kv one will be the wasted spark event.

Will your software allow you to overlay several captures so you can compare the burn time, peak kv and burn time voltage changes on the ignition event?
A couple screen captures from each cylinder with the full waveform shown may help to isolate your issues.


Dave


britishtrident - 9/9/17 at 08:02 AM

The varition between spark events on the secondary (HT) side of the coil is normal on a wasted spark system as every other spark is on an empty cylinder. I am short of time and will look at the vid in detail later. I wouldn't worry about the short spark duration as the coils are smaller the those on cars, most of the figures quoted for burn time are for US cars which tend to have much longer burn times.
You fuel air mixture looks okay but the ammout of hash at the start of some of the spark events is intersting.

I would suggest you post the the question on Paul Danner's ScannerDanner forum
Scannerdanner Forum link









[Edited on 9/9/17 by britishtrident]TextYellow

[Edited on 9/9/17 by britishtrident]


britishtrident - 9/9/17 at 10:41 AM

What I would do first is capture and screen shot a parade of the fspark events on each cylinder by increasing the time base to cover a least 4 spark events, this will give 2 firing events and 2 wasted spark events on each capture --- to see any detail on this might be pushing an entry level oscilloscope to the limit but it should give you an idea of what is going on. The idea is to quote ignition diagnostic guru Jim Morton "looking at the forrest not the trees" to spot any odd balls or drop outs.

This spark parade screen cap is from a negatively fired wasted spark on a Rover note every second spark is a wasted spark with lower (less negative) voltage.



When capturing it will also help to increase the oscilloscope trigger voltage (decrease for negatively fired cylinders) so it triggers only on actual firing events not wasted sparks.


Were the replacement coils new ? as I suspect a spark is jumping to ground outside the cylinder, also I would attach the scope to the power side of the coil and look at the supply for dropouts.


gingerprince - 9/9/17 at 02:57 PM

So had another half hour in the garage today, gathered more information, but getting ever more confused

To answer some of the above questions: -

Airbox is missing, but it's not the cause of the misfire.

Earths from coil are the transistor switched earths via the ECU. So if there is an earth issue, it's general wiring and I'll need to re-wire a fair bit. Could test with a couple of jumpers, but might find myself doing a full rewire anyway if I don't get anywhere with diagnostics.

Definitely temperature related, doesn't start major dropouts until fully up to temperature.

I'll have a look on that scanner forum.

So to pictures, the 1st one shows one of the coils LT and HT side on one plug (+ve fire). I can see the cylinder firing (1), and the wasted spark (2). Oddly there's also a blip (3) which lines up with when the other coil fires. Not sure if this is just noise from the scope, or whether something is leaking in the ECU/wiring, or whether it's normal? Also the gradual drop of voltage looks like nothing I've seen in other examples?

Description
Description


Next is HT on a +ve firing cylinder (both look similar), you can see the firing (1) and wasted (2). Has the strange voltage drop-off too.

Description
Description


Next is HT on a -ve firing cylinder (both look similar). The big thing to note is that the voltage slope off after oscillation starts MUCH higher (...lower, as it's -ve but you know what I mean).

The fact that it's the same on BOTH coils in the -ve side means that either ALL my coils are broken (I've tried 5 different coils), there's something at the ECU, or there's some other weird thing causing an imbalance. I've swapped cylinders +ve/-ve too so 3 with 2, 1 with 4, and the oddness moves with polarity, so I don't believe it's specific to cylinders.

Description
Description


Finally a +ve signal during misfire events - I believe (1) is a misfire. Simply don't see a high voltage. So is this a coil not charging properly so nothing to fire with, or a HT short circuit meaning there's no resistance or something?

Description
Description


Hope some of this makes sense to someone? Might have a bit more time to play tomorrow, but will have a look at the scanner forum too.

Thanks


jossey - 9/9/17 at 03:13 PM

Hi mate try a bigger battery.

Had similar issues when the rectifier went due to battery issues.


britishtrident - 9/9/17 at 05:12 PM

It looks like when the misfire happens the coil is not being fully grounded by the driver in the computer



gingerprince - 10/9/17 at 03:23 PM

Oddly that chart is normal running not a misfire event?

Found my spare ECU this morning, so swapped that out and still the same.

On the battery issue, it is a big battery but it has been deep discharged a couple of times and goes flat quicker than it used to - might be worth swapping anyway but not sure it would cause this? Have already run it with rectifier disconnected so running battery only and no difference.


obfripper - 10/9/17 at 09:48 PM

I think your ht pickup is loosing its zero volt point, and is allowing the detected voltages to wander.
The lt side should directly mirror the ht trace (with a good coil, i dont think you have 5 faulty ones), but it is not according to your traces, and has the coil sticking at 4kv after firing event and then -2kv after wasted spark event - this is unlikely to be occuring.

I have had similar things with dc amp clamps when a current higher than can be read is passed across - which offsets the zero point. It may be that the peak kv is higher than your pickup can detect causing similar issues.
Has your daily driver got accessable ht leads to get a comparative reading from a known good system?

From what i can see, you may have a drop in compression for your misfire events, some traces from the lt side while the misfire is occuring will suffice to check this. If you can upload the traces somewhere and point out what software is needed to view the traces, zooming into each ignition event is required to give more detail than a screenshot can give.

Dave