ECU no longer firing 2 cylinders...

[jigga009]

1 minute ago, neil brown said:

Stock ignition system 

Yeh mine is set at 3.0ms all the way from 10-16v at all RPM levels.

[neil brown]

Just looked at dwell for Subaru ver 7 ign system

 

dwell ranges from 6ms to 1.7ms over the table 

[jigga009]

1 minute ago, neil brown said:

Just looked at dwell for Subaru ver 7 ign system

 

dwell ranges from 6ms to 1.7ms over the table 

I'm guessing this is what my tuner was referring to when he mentioned that I would have to make some adjustments within the map if I decided to switch back to the standard Subaru ignition setup; dwell settings...

I was under the impression though that dwell settings are a little redundant with a CDI setup though? Since the module itself controls how long it charges for?

[neil brown]

Are you running the latest firmware just in case there is bugs in software

Just read at hp academy 

 

that cdi ignition is good for when you require a really strong spark at boost levels at 2 bar plus

 

how much boost are you running 

[Adamw]

Note dwell time is irrelevant when running a CDI.  You cant have too much or too little, all it needs is an edge.

If you are worried that the ECU is over-heating you can log ECU temperature.  The max temp it has seen is also automatically saved in the ECU statistics register.  I think they shut down around 90deg C.

[neil brown]

Should it be set to 0 or 3ms as mentioned in his post

[Adamw]

It will need to be non-zero.  3ms is a good number - it gives the ecu a little headroom to stretch or shrink dwell if there are trigger errors etc.

[jigga009]

14 hours ago, Adamw said:

Did they state that the CDI box can actually fire all channels simultaneously?  I think I mentioned to you earlier that I have heard that many CDI's cant actually do this since either the transformer or capacitor is shared between channels.  

Adam, the CDI system is in fact able to fire both channels simultaneously, or close to simultaneously (in cases where the leading and trailing spark plugs fire close together) as each channel has its own storage capacitor...

I found this information in the instruction manual.

[cj]

What does you wiring diagram look like as far as +12v and grounds to the CDI boxes? Are they the factory wires that used to feed the COP's or something new? 

If you have a partially damaged ground or 12v feed to the CDI boxes, you essentially limit how much current they can draw from the battery, which in turn slows down how quickly they can recharge. The idea behind a CDI box is that it charges the capacitors comparatively slowly (one per channel according to your manual) and then discharges them into the spark plug when fired. If the capacitor isnt able to fully charge between firing cycles, it will either have a weak spark or no spark when commanded to fire.

so... my theory is that there could be a loose connection/bad crimp/partially cut wire somewhere in the circuit that powers the CDI's, and that is limiting how quickly the caps can charge. When you test fire it sequentially(cranking), your engine rpm will be about 180rpm , so with some basic math you have 3rpm per second, and with 2x channels per CDI, that leaves ~160ms between firing events (per cdi), and 330ms between firing events per channel. When running an ignition test you only have 100ms between firing events and each one fires both channels, so there is effectively 50ms per cap to recharge if at a fixed current (or 100ms for both caps but double the current requirement, depending on how you think about it). Following this theory through, it means about 1200rpm is also about 50ms between individual channel firing, and so if you get it running, it may well start to misfire somewhere below 1200rpm.

because you're seeing the same behaviour on both CDI's, and it used to run ok, you can probably assume any damage is in a part of the wiring that is shared between both boxes. Testing resistance with a multimeter wont show a problem because the current is so low that it wont care that you're wiring is only allowing say 20% of the current it needs to. What you can do though, is set your multimeter to volts mode, and measure from the battery positive to the CDI 12v input while you try to test fire it. If you have a substantial voltage drop between these points it will show up as + or - voltage on the multimeter and will indicate resistance in the wire when current is pulled through it. Be sure to test both + and - sides of the circuit, either one would cause the same issue. It may be easier to disconnect the wires from the CDI and connect a more "stable" but still high current load such as a head light bulb if the multimeter voltage jumps around too much while trying to measure it.

[jigga009]

7 hours ago, cj said:

What does you wiring diagram look like as far as +12v and grounds to the CDI boxes? Are they the factory wires that used to feed the COP's or something new? 

If you have a partially damaged ground or 12v feed to the CDI boxes, you essentially limit how much current they can draw from the battery, which in turn slows down how quickly they can recharge. The idea behind a CDI box is that it charges the capacitors comparatively slowly (one per channel according to your manual) and then discharges them into the spark plug when fired. If the capacitor isnt able to fully charge between firing cycles, it will either have a weak spark or no spark when commanded to fire.

so... my theory is that there could be a loose connection/bad crimp/partially cut wire somewhere in the circuit that powers the CDI's, and that is limiting how quickly the caps can charge. When you test fire it sequentially(cranking), your engine rpm will be about 180rpm , so with some basic math you have 3rpm per second, and with 2x channels per CDI, that leaves ~160ms between firing events (per cdi), and 330ms between firing events per channel. When running an ignition test you only have 100ms between firing events and each one fires both channels, so there is effectively 50ms per cap to recharge if at a fixed current (or 100ms for both caps but double the current requirement, depending on how you think about it). Following this theory through, it means about 1200rpm is also about 50ms between individual channel firing, and so if you get it running, it may well start to misfire somewhere below 1200rpm.

because you're seeing the same behaviour on both CDI's, and it used to run ok, you can probably assume any damage is in a part of the wiring that is shared between both boxes. Testing resistance with a multimeter wont show a problem because the current is so low that it wont care that you're wiring is only allowing say 20% of the current it needs to. What you can do though, is set your multimeter to volts mode, and measure from the battery positive to the CDI 12v input while you try to test fire it. If you have a substantial voltage drop between these points it will show up as + or - voltage on the multimeter and will indicate resistance in the wire when current is pulled through it. Be sure to test both + and - sides of the circuit, either one would cause the same issue. It may be easier to disconnect the wires from the CDI and connect a more "stable" but still high current load such as a head light bulb if the multimeter voltage jumps around too much while trying to measure it.

Hi CJ, thanks for your response! And I am thinking that your theory might be on the money...I'll explain.

The battery in my car has been relocated to the boot, since there is no room in the engine bay these days. I have a 2AWG wire going from the battery positive to the starter motor, and then another 2 AWG line from the starter motor to a "2 AWG in/2AWG out" distribution block with 150A inline fuse mounted on the side of the engine bay just underneath the main fuse/relay box in the engine bay. Coming out of that distribution block were 2 10AWG wires that feed each of the CDI boxes and an 8AWG wire that feeds the main OEM fuse box in the engine bay. All 3 wires were sort of scrunched together and inserted into the outlet of the distribution block. The car worked fine like this, but I wasn't completely happy with the solution since it always looked like the conglomeration of wires was always on the verge of popping out and looked to be full of dirt. The car ran around like this since it was put together a few years ago, with CDI boxes going bad every once in a while, but as I maintained before, i had always been able to fire all plugs at once in order to see which box had an issue. I was happy with the function of the car, but never really liked how the shop sort of twisted all 3 wires together and shoved them into the outlet of the distribution block.

A couple of weeks ago I made what I thought was a slight change with the intent of cleaning things up a little bit - I purchased one of these: https://www.bluesea.com/products/5045/ST_Blade_Compact_Fuse_Blocks_-_4_Circuits

And then removed the old power distribution block with integrated 150a fuse and connected the 2AWG wire from the boot of the car to the bottom post of the new fuse block pictured in the link using a 2 AWG set screw ring terminal, and then crimped some eyelets onto the 2 10 AWG wires that fed each CDI box, and one on the 8AWG wire that went to the main OEM fuse/relay box. I then installed all 3 wires onto the new fuse box. The fuse box was then mounted just underneath the OEM fuse/relay box where the old "2AWG in/2AWG out" distribution block was located before. 

Since the wire going to the main fuse box was 8AWG, I used a 40Amp fuse for that, and for the CDI circuits, they each received 30A fuses since they were 10AWG wires. This matched the 30A rating of the in-line fuses the CDI 12V power wire came with. I did crimp eyelets onto each of these wires, and while the crimps did feel secure when I made them a few weeks ago, I did notice on Sunday that they did not seem as tight when I was busy scrambling around trying to figure out what was going on with the car on the side of the highway. It did not occur to me though that they might be loose to the point of causing such an issue though, but I may have been wrong. I typically solder all of my connections (and all of the other connections mated the CDI boxes to the car when installed professionally were soldered), but for some reason I thought I would give crimping wires a go, since it seems to be an accepted way of making an electrical connection. So in essence, I introduced crimped connections to a setup that used soldering for all previously existing connections. It might be that the vibration the car has undergone within the past couple of weeks of use shook the connections I made loose.

As you theorized, both CDI boxes receive their 12V power from the battery via the crimped eyelet connection to the fusebox in the link above, and the crimp connection may not be the best, I freely admit. The CDI boxes also receive a switched 12V source, and I was able to confirm with a testlight that each box was connected (soldered) to switched 12V supply. All other inputs from the ECU were tested and found present and functional. The only connection I forgot to scrutinize were the crimped connections I introduced to the system, which as you theorized, provide the main power to the boxes. They are the only non-soldered connections in the entire system, and also the latest additions to what was a working system. :/ 

I'll break out the soldering station and get them soldered tomorrow morning  

 

Thanks for your suggestion!!

 

 

 

 

 

[Stevieturbo]

Make life simple.....CDI Box, in the bin.

 

2001+ OEM Subaru COP's will easily take you to 800hp with zero issues.

 

And not sure what trying to measure frequency from the ecu outputs is supposed to tell you.

 

There will only be a frequency reading...as and when it is trying to charge/fire the coil, which of course will be a variable relating to rpm. To test the output proper you really need a scope...although you are happy now the ecu side of things is ok.

 

But really...CDI boxes are more hassle then they are worth, and for pretty much anything these days are in no way needed.

[jigga009]

46 minutes ago, Stevieturbo said:

Make life simple.....CDI Box, in the bin.

2001+ OEM Subaru COP's will easily take you to 800hp with zero issues.

As mentioned earlier, there is a reason why I had to go with a CDI setup over the stock COP; I was bumping into its limits, and the only way forwards in my situation was to switch to a CDI system. I would have been all too happy to stick with the OEM COP if its limits were a bit higher, and the transient demands it was putting on my electrical system (even with a 160amp alternator) was not dragging the system voltage dangerously low.  Lord knows I prefer the drivability associated with the stock COP over CDI, but the COP was showing signs of not being appropriate at the power levels being run.

I do agree though that they can be quite annoying from a reliability perspective, and with the longer duration of each spark with an inductive ignition system, low-down drivability/torque is actually better imo with a COP setup. With that said, CDI systems do bring certain benefits that are hard to ignore once you have outstayed your welcome on the stock Subaru COP system, and it isn't as if people don't have issues with stock coils acting up either. I do hear that Nissan R35 GTR coils might be a lot more capable than anything Subaru has though, so perhaps somehow retrofitting those coils might be a solution for reliability while being able to light off the plugs. 

 

Mini-update for the day:

I didn't have too much time today, but I was able to dismantle the old crimp connections I made a couple of weeks ago as per CJ's suggestion and properly solder on some eyelets instead. I unfortunately did not take any pictures as I went, but after looking online at proper crimping examples and comparing them to what I thought was good enough, I have to admit that I likely won't bother crimping anything again. I should have soldered these connections, but I was too lazy at the time to bust out my soldering station.

The other thing I noticed is that 2 weeks after installing the crimp connections in the first place,  all 3 wires (main fuse box 12V+, CDI box #1 12V+, and CDI box #2 12V+) were developing corrosion (green residue on the exposed copper strands within the crimps). I could not see this before now due to the plastic sheath on the eyelet ring terminals that covers everything behind the ring.  I cut off the affected wire and exposed fresh sections to solder into eyelets, cleaned them with isopropyl alcohol then tinned them up and used flux to allow for good solder flow into the eyelets. Everything has adhesive heat shrink tubing to keep moisture out.

One last thing I will do tomorrow before attempting to firing it up is to apply some Stabilant 22a to the eyelet connections within their fuse box.

I should be able to try firing the car up tomorrow. It's a bit late in the evening right now, and the car is rather loud.

[Adamw]

6 hours ago, Stevieturbo said:

And not sure what trying to measure frequency from the ecu outputs is supposed to tell you.

 

There will only be a frequency reading...as and when it is trying to charge/fire the coil, which of course will be a variable relating to rpm. To test the output proper you really need a scope...although you are happy now the ecu side of things is ok.

He was using frequency because he was broken down on the side of the road with only a few tools, a multimeter being one.  When the ignition outputs are in test mode they fire a 2ms dwell at 10Hz, so I suggested he can use his multimeter set to frequency mode to determine whether the ignition signal was present on all channels.

[Stevieturbo]

So you have a wiring or alternator issue ?

Just exactly how much current do you think the OEM coils pull ? Because it is an insignificant amount in general and in no way the cause of any problems.

Are you making 4 figure power or lots more to really necessitate stronger coils ? If so, IGN1A's are the obvious choice.

 

Bottom line, CDI is not necessary, no benefit and almost always more hassle than they are worth.

 

OEM Subaru coils are very powerful and will easily cope with 800hp. If you;'re having problems at much lower than that, then the problem is not the coils themselves.

 

and a good crimp is always preferable to solder. You'll rarely see solder in aerospace, milspec stuff.

 

if you really must solder, ensure the wire is well supported so it cannot fatigue and fracture over time

[motomattx]

Have you tried using the coils from another car? I have known cdi boxes to damage coils in strange ways that are not always easily apparent.

[jigga009]

8 hours ago, Stevieturbo said:

So you have a wiring or alternator issue ?

Just exactly how much current do you think the OEM coils pull ? Because it is an insignificant amount in general and in no way the cause of any problems.

Are you making 4 figure power or lots more to really necessitate stronger coils ? If so, IGN1A's are the obvious choice.

 

Bottom line, CDI is not necessary, no benefit and almost always more hassle than they are worth.

 

OEM Subaru coils are very powerful and will easily cope with 800hp. If you;'re having problems at much lower than that, then the problem is not the coils themselves.

 

and a good crimp is always preferable to solder. You'll rarely see solder in aerospace, milspec stuff.

 

if you really must solder, ensure the wire is well supported so it cannot fatigue and fracture over time

Stevieturbo, thanks for your response!

I don't believe it is an alternator issue, as logs showed system voltage to be utterly normal all the way to the end. I'm not ruling out an alternator issue though, but unless the log is telling me porkies, the log shows that the alternator was working, with voltage in the 13.7-13.8V range all the way to the end. In addition, I do have warnings on my dash set to warn me if and when voltage drops below charging voltage or above normal voltages. Nothing from what I can see in the log was out of the ordinary. Sadly, the shop I visited today with the alternator could only test it if it came out of a specific vehicle... Since it is not a normal production alternator off a specific vehicle, they were unable to program their computer with the details on the alternator in order to run the test. They did not believe though that it sounds like an alternator failure, given that data logs showed that it was charging all the way to the end. They thought I would notice other electrical-related issues such as issues with the radio. My battery did pass testing with flying colours though!

The whole CDI vs IGN1A inductive coil debate is old and about as divisive as talking politics and religion on a first date. I have my reasons for running a CDI setup, and I am well-aware of the capabilities of the stock Subaru COP, and have heard the arguments from those who are pro-IGN1A as well. I don't think anyone is grossly wrong at all. Both solutions have their pluses and minuses.  All I can say is that I'm not one to complicate things any more than I need to, and think long and hard before leaving the safety and reliability of OEM for anything aftermarket. Not to be cryptic, but putting one and one together, you already have an idea of the limits of the Subaru COP system, and I've already described what I experienced when I found that I was reaching the limits of the Subaru COP setup, so one might deduce that I am dealing with a slightly more serious setup than average. 

You are correct about the use of crimping in those industries  I learned about the space vehicle use of crimping connections from a former NASA technician who chimed in on a different forum I came upon, weighing in on the debate. In addition, all the videos I saw online made crimping look even easier than soldering, and I was feeling lazy that day also..  I'll freely admit though that I have been soldering for much longer than I have been crimping :s 

Regardless, my badly crimped connections were not the cause this issue I am seeing, as I tried starting the car again today. I am looking into other things now such as engine grounding connections, which might affect both CDI systems and also explain why it seems as if the starter motor is rather slow, even when powered with a fully charged AGM battery producing over 1000CCA. Even if grounding is found not to be the cause either, at least it gives me a chance to spruce those areas of the car up that I have not touched in a while.

 

[Stevieturbo]

I guess if for whatever reason you believe you must run a CDI system...then maybe opt for something proven like M&W, and get rid of the junk MSD.

[jigga009]

21 hours ago, Stevieturbo said:

I guess if for whatever reason you believe you must run a CDI system...then maybe opt for something proven like M&W, and get rid of the junk MSD.

I use MSD's coil and wires, but that's it.

[neil brown]

I would use Subaru COP set up as if good enough for Andy Forrest at 1200+ bhp on his Impreza as videoed by HP acadamey 

 

think you you will still be having issues with it constantly 

[Stevieturbo]

Andy has 6 cylinders though...( and perhaps a little more than 1200 on his faster runs ), so less cylinder pressure per cylinder.

But really, without any actual detail of the build or level he is at..who knows. He could be at 300hp....or 1300hp...or nowhere in between

Standard coils work exceptionally well and he hasnt stated any actual symptoms of a weak spark using them. But lots of problems with CDI, which is common.

 

Bin CDI lol.