Faulty ECU or Faulty CoP?
 

How likely is an OEM ECU to fail? Specifically the controller for the ignition system/circuit.

I have a Fab9tuning CoP on my 1997 (PnP) and it's been nothing but trouble. I have finally aquired an OEM system and that fires up and idles as it should.
The Fab9tuning unit barely idles and if it does catch, cyl3 is dead.

In the past, the symptoms would start with a random misfire being detected (very noticeable at speed) then would go away. About a week later, it will come back plus the car would barely start. Other rising issues at the time were hesitation/stumbling under hard acceleration.

Symptoms (when the car used to idle properly):
*From a standstill; if the throttle we're floored/launched, it stumble and backfire/misfire but if gingerly accelerated then it will hold it's own.
*If at speed and the throttle is stabbed/floored; it struggles and misfires. Very rare cases would the engine literally just shut off.
*Sometimes when coming to a stop, the car will stall and not be easily restarted (or stall if I clutch in, while at speed, and let the engine slow to where it's supposed to idle).

Things I've done:
*Cleaned all possible ignition related earth points that I could find.
*New plugs
*New coils (2011 CBR1000)
*Reseated all connections
*New thermo plug (ECU)
*New fuel filter (thought the old unit was clogged)
*Pulled each coil while the engine was running to confirm a dead cyl3 (at idle).

Have I missed another check or is my ECU (or CoP) simply kaputz?




tl;dr/rant:
Now I'm scratching my head.
Fab9tuning are no help at all and frankly I feel like they're crooks at this point.
You can't call them because their line is perpetually busy and emails sent might not get a reply for 3wks only to not have your question answered.
However, when I go to leave a 1/5 rating on their website, the ratings have to be approved. Suddenly he replies to me out of no where then goes right back to weeks of no reply and/or 'we're so busy' excuse.
Seriously, wtf???

If you have a DMM it is very easy to check for signal to the coil.

If the OEM coils work fine, and the Fab9 setup does not, this tends to exclude the ECU as a culprit.

I would suspect the external igniter module from the Fab9 system. No idea where they get those from, does it have any obvious markings on it as to manufacturer and part number?

This particular unit does not that I could readily find.
Back when I first got the system and having these issues, I was tinkering with the idea of benchtesting it to rule that out as a weak link or not.
The chaps at the local AutoZone managed to cross-reference it to a V.A.G. part which is the extent that I got with that before attempting to consult Bryan again.

The fact that only coil 3, and not coil 2, produced a reduction in RPM further points to either the wiring harness or the igniter. (I'm assuming you've swapped coils 2 and 3 and that the problem stayed in the #3 position.) Remember that the OEM ECU has only two ignition outputs which are doubled-up at the igniter, so if the ECU were at fault, coils would fail in pairs. (1 & 4 or 2 & 3).

Can't help you troubleshooting the wiring, that's all on you with a multimeter and the Mark One Eyeball.

I'll double check that it's indeed only #3
Wait a tic, 2 and 3 fire in unison? I always thought it was 1/3-2/4! Learned something new about these engines every bloody day! Lol

And ja, I swapped coils multiple ways.
Using the CBR1000 coils I got from the local bike shop;
*Pulled #3 then swapped all the FAB9 denso labled ones (trying each one on #3)
*Tried rearranging the CBR coils

I then tried the same test with the Fab9 coils in place of the CBR1000 units.


With a DMM, could I just ohm test the coil loom rather than having the engine on and testing for a signal or is that the only surefire way? I'd rather not have the engine on with a dead cylinder anymore unless I absolutely have to.

Side note: Bryan did send me 2 ignitors. The first one was some flat unit that was significantly different from the original Fab9 unit (that was actually branded Fab9tuning). This flat unit is the one that I took to AutoZone to have cross-referenced when it went wrong, that's when I found out it was a VAG related part.
Bryan finally sent me one that was a mirror image of the original, saying there was indeed a problem with the skinny/flat ones and that one other chap was having the same issues.

With inline-4 engines, where the 1 & 4 rod journals are on one side of the crankshaft, and the 2 & 3 rod journals are on the other) the firing order is almost always 1-3-4-2. 1-2-4-3 is also possible, but exceedingly uncommon (think Soviet-era), and equivalent for our purposes. (I always have to double-check this, as I grew up wrenching on air-cooled VWs, and the flat-4 engine is a whole different animal. I still have the 1-2-3-4 firing order for those engines ingrained in my mind. For those engines, 1/3-2/4 is indeed the correct wasted-spark pairing.)

If 1 and 3 were paired, then when the #1 plug fired at TDC, the #3 plug would be firing at BDC just after the intake stroke, and with a cylinder full of uncompressed fuel-air mixture. That'd be dumb. (Ditto for the other pair.)

By pairing 1-4 and 2-3, you ensure that when the cylinder which is at TDC at the end of the compression cycle fires, the cylinder which gets the wasted-spark is at TDC at end of the exhaust cycle, so there's nothing burnable in the cylinder.



Here are the relevant wiring diagrams from Mazda for...


'90-'93:

https://cimg3.ibsrv.net/gimg/www.mia...ec9b37266e.png




'94-'00:

https://cimg6.ibsrv.net/gimg/www.mia...801b5f9f5a.png




'01-'05:

https://cimg7.ibsrv.net/gimg/www.mia...d5305efff4.png

(Ignore the numbers in the '01-'05 drawing; these are order-of-ops for disassembling and removing the system, not references to cylinders numbers. The front of the engine (cyl 1) is at top-left)







With a DMM, testing the loom on the bench is the only way. You can't look at the ignition pulses on a running engine with a DMM, you need an oscilloscope for that. (And, really, cheap multi-channel scopes are so cheap these days that anyone who is touching the electronics on their engine should have one.)





Then run the OEM coilpack you have until you fix this.




Get him to tell who builds the igniters. I can guarantee you that FAB9 isn't building those, they're just buying some random OEM part and sticking a label on it. (Edit: this isn't an indictment of Fab9, it's how smart engineers who are designing to a price-point do things.)

Any cheap DMM will read a square wave tach signal (coil, injector pulse, etc.) on the AC Voltage setting. Should be between 1 and 6 volts and increase with engine RPM.

Agreed, you will see some kind of signal. But if you want useful diagnostic information (dwell, saturation time, peak amplitude, the ability to compare timing between multiple ignition channels, etc), you need a scope.

Which he doesn't have.

Isn't knowing signal is getting to the coil useful diagnostic information? No signal at the coil confirms the issue is upstream.

From all the forum and interwebz digging i've done, Fab9 appears to use a pretty common Audi/VW ignition mod for their setup: <!--td {border: 1px solid #ccc;}br {mso-data-placement:same-cell;}-->4D0905351 (unless something has changed in recent times). They're everywhere on ebay, may be worth spending $20 or so to rule it out.

I had a dead coil when I first installed FAB's COP kit on my car, luckily Brian sent me a new coil and that resolved the issue. I had another misfire issue a year later that turned out to me the ICM, bought one of the $20 ebay ones, issue solved. I actually pulled the old one apart after and one of the transistors let the magic smoke out.

Sorry for the silence, medical rubbish popped up.
20usd for a replacement ignitor on eBay???? That's pretty steep from the prices I saw before of 130+ (albeit, from the likes of typical franchise auto superstores.).

But getting on with it; I tested both looms (before and after the ignitor) and both of them checked out. So, looking at the ignitor as the potential culprit (yet again) is there a method to bench test it or is it simply 'it may be broke, chuck it in the bin.'?
If what the chaps are saying about the Ebay route is true then I'll try that once payday is here.

So if that's the case, is it simply a matter of our OEM ECUs overpowering the units or/and just crappy ignitor units?


When all of this began happening (literally a few weeks after I first purchased it) the Fab9 branded one had a pin that would wiggle backward into the module, break connection and cause a random misfire code; that issue was strictly mechanical but from the short-lived experience, the system seemed to work just fine aside from that minor setback. I messaged Brian at least 2x about it (pulled it back out with needle plyers as he instructed; turned out to not be a permanent fix) and he finally sent me a new module. The new one he sent was the flat module and not even a few days later, all the mentioned symptoms began happening and getting worse with time.
He sent another one after that and the same result, drove ok for a few days, then starts dropping cylinders and non-existent acceleration from a stand-still or hard pull while at speed.

I'm not sure to be honest. When I was having issues with mine some googling of the PN showed up a few Audi / VW forums with similar issues on the car's they are meant for, so I'm inclined to think it's just a poor design of the igniter unit. They also seem fairly vulnerable to heat, so I've now bolted mine through the cowl / firewall and put some thermal compound between to keep the module cool. SO far, so good (Fingers Crossed) :x:

The concept of the OEM ECU "overpowering" the unit it more or less impossible. It's a mis-matching of concepts.

These are high-impedance, logic-level circuits, and power transfer just isn't a factor. (In other words, if you have a device that expects a +12 active-high trigger and sinks 10ma from it, you could drive it with a truck battery capable of delivering 1,000 amps, and the circuit would still only draw 10ma.)

It may be that the igniters being supplied are of poor quality, it might be that they are inadequately rated for the coils which they're being used to drive, or it might be that they're rated for operation at a far lower dwell than they are experiencing in this application. (This last option is highly improbable.)

Not if the stock ECU is dwelling the igniter for 25-50ms during cranking and 5-12ms during running. The igniter then runs super hot and exceeds it's thermal design envelope.

Hint: the numbers above are not random.

Which is why I said "or it might be that they're rated for operation at a far lower dwell than they are experiencing in this application"

Admittedly, I qualified this with an "it's unlikely" disclaimer. I've scoped several OEM ECUs ('90, '92 some random 1.8 NA, '00, and '04), and I've never seen them dwell at any more than 5-6ms when running. Where are you getting 12ms from?