Hi guys,
I finally got the car to start and idle, albeit roughly. It will only start and stay running with the CAS rotated max CCW (retarded). This would make since why it would jack up my EGTs, but if I so much as touch the CAS, it falls flat on its face and dies. If I try starting with the CAS centered, it won't catch.
I've cerified the CAS is stabbed correctly, yellow timing mark with pointer and CAS dot and mark lined up.
AFRs are in the 12.0 region at around 2000rpm idle. BAC unplugged, MAP is around 70kpa at that rpm.

The header is glowing red after only a few seconds of starting. I'm worried to let it even idle for sake of my seals. Everything else looks good though.

LINK TO MSQ

 

On phone so can't view msq, but what tooth #1 angle are you using? Do you have a log as well?

 

First tooth angle is 0 degrees BTDC. I used Aaron's writeup to setup all my ignition settings and base maps.

My real-time gauges in TS say I'm running around 25 degrees at ~2000 idle but it sounds not right. And the EGTs look like that could be wrong too. I will try to wire my Chinese EGT gauges up today and see if they actually work.

I will try and figure out how to save a log today and upload later. I suppose I could play around with the offset to

 

Are you SURE the CAS is stabbed correctly? It is very easy to get it one tooth off when stabbing it which will result in these exact symptoms.

You'll find that unless you rotate the CAS the entire way, the timing mark won't even be close with a timing light. Oh, also when you check base timing you need to make sure to lock the timing so the ECU won't try to advance it.

It can take several stabbings of the CAS to get this right. Take the top cover off and use your fingers to make sure it doesn't rotate while being inserted.

The other option is a VR sensor backwards. But make sure the CAS is physically correct before going into software and inverting trigger inputs.

 

I lined up the timing mark and pulled the CAS with the cover removed and it was lined up with the dot just like it should be. How do I check if the inverted is correct?

 

If you have a factory flywheel, remove the inspection platr near the exhaust and spin the motor until the flat side of the flywheel is visible. Thats your TDC.

 

I reverified that the cas was stabbed correctly. Egts were over 1300F. I cheated the cas over a few teeth and rechecked with a timing light and was able to get it close. It seems that the offset is wrong. For now, it idles and restarts reliably. I'll get a log up later.

 

I think I figured out the timing, but somethings weird. To get the timing light to line up with the yellow pulley mark, I had to cheat the CAS over a couple teeth. Whenever I set the CAS per the FSM, the engine would run poorly when using timing from table (Aaron's base table). Whenever I'd lock the timing at -5*, the timing light would be nowhere close to the mark. I made sure to use the leading plug wire for timing. By removing the CAS and reinstalling it slightly off-mark, I was able to dial in the correct timing.

I tried using "Basic Setting>Trigger Wheel Settings>Tooth #1 Angle (deg BTDC)" to adjust for the CAS, but you can only enter positive number. Positive numbers mean that it's actually more retarded than it thinks, which was already my problem.

Anyways, I got a pretty decent idle with it set this way and decent throttle response without messing too much with AFRs. Also, my $30 ebay Chinese EGT gauges were so helpful figuring out the timing.

Heres a walkaround of the car BEFORE I got the timing lined out. It idles much smoother now.

Here's a pic of the engine bay:

 

men make sure that your timing its 100% correct before you start going on boost our your going to be blowing engines like crazy.

 

Now check for inverted VR sensor wires.

You can swap the VR input polarity in software to test for this without having to dig into your wiring.

Have you done something weird like connecting the trailing ignition trigger to the leading? I've seen it happen (Haltech installation).

 

I will try the vr polarity in TS. I verified that the CAS wires were to the correct pins. It's been a while since I checked the coil wires, but I can do that today.

 

I was able to verify that all the coil wires were correct to the board. Also, I messed with the going high/going low settings trying "inverted" and couldn't get the car to run right. Even with the non-inverted setting the car is particularly hard to start. I also noticed today that I was getting a lot of sync drops. It made it nearly impossible to idle to car. below 1500 because a drop would cause it to stall. Using the advance **** on my snapon timing light, timing is still on the mark. Egts at idle are around 900F. Upload tune and log tomorrow.

 

Here is a link to a datalog and my most recent MSQ

I'm not sure what the sync loss is about, but I didn't notice is on Saturday when i had it running well. Yesterday, I could hear the FP relay flipping on/off as the car was idling/stumbling.

 

If you're getting sync loss, take a composite log and send it to us. Sync loss means that there is noise or something else on your CAS signal that the MS doesn't like.

Ken

 

I am still not sure what's up with my timing. I reset the CAS back to FSM spec and my timing is still wildly off.

I locked the timing instead of using the table and kept adjusting the value until I got a nice idle. I ended up having 50* of advance. Granted, the timing light showed that this was not actually 50 degrees, but it runs and starts well with no sync losses. If anyone has a spare CAS they'd like to sell me, I think that might be worth looking into.

I tried "Falling Edge" Ignition Input Capture, I tried "Going High (Inverted)" Spark output. Neither made an improvement if I tried running off of Aaron's timing map. I have no doubt his map is fine, just that my MS doesn't think TDC is where it's actually at.
I tried using the "Tooth #1 Angle (deg BTDC)" To offset TDC, but I'm not sure what direction to go. It seems I would want to go negative to make TDC before where it thinks, but you can only put in positive numbers.

Any ideas?

 

Do you have both pots on the v3 board turned fully counter clockwise?

Have you taken a composite log so we can see what's going on with your sync loss?

Ken

 

I had turned the pots full CCW, but I will confirm this tonight.
I selected composite log and got ~1hr of run time last night but it seems that it only has
PrimLevel, SecLevel, Trigger, Sync, ToothTime, and Time. Seems like more should be there, but anyways, heres a link to that LINK

I actually drove the car last night. I have boost cut set to 105kpa and kept my eyes on EGT like a hawk.
I fiddled with Tooth #1 offset to try and get it to run on map. I ended up using a 270* offset, since you cant enter negative numbers, this should make it -90 effectively. I was able to get my timing light close to the mark when I locked timing at -5* with this offset. Throttle response was decent and idle was solid. I still noticed sync loss if I revved quickly in neutral.

While driving my tune 'disappeared'. The car just stopped running, no fuel, no ign. I pulled to the shoulder, grabbed my laptop, and power cycled the MS. When it reconnected with Tunerstudio, it showed on the change log that my TS tune settings (which were correct) were COMPLETELY different that my MS tune. It had all zeros in the tables and the trigger settings were default. I reloaded an old tune and away she went. I was still getting some sync drops, mostly at low load and low RPM. I locked the timing at +60 with the CAS installed default and it idled and ran okay. EGTs were around 1100 at cruise, which makes me thing my ignition is still retarded and nowhere near 90*.

If someone can give me pointers on how to capture an actual composite log, I will get one up tonight.

 

Something is not right here if your tune is disappearing. But you probably already knew that.

Start with the easy things. Grounds solid? Pins in the DB37s solid?

Are you getting processor resets logged or shown in TunerStudio?

How's the alternator? I've seen a half blown alternator superimpose AC on the output which caused all kinds of issues.

 

I'm not really sure how to check the db37 pins, but I have continuity between all 4 CAS wires and 3 ign outputs at the harness and connectors.
I'm not sure how to see if I'm getting processor resets.
I have all of the ground wires from the harness terminated at my fuse panel and a 10awg wire from my fuse panel to the same lug the battery is grounded to chassis with.
Alternator is putting out 13.7V and not making any particular noises. It's a Taurus alternator mounted where the AC comp used to be, but I've ran it with the stator and source wires unplugged and it made no difference.

I bought a new CAS, so I will see how that goes in a few days.

Also, I'm going to check and see if I have any other main pulleys to try out. After doing some reading, it seems the hubs/pulleys are generation specific. I'm using an S4 TII hub/pulley (I think) with an S5 front cover and RX8 eshaft... :-!

 

Just make sure that all the pins are tight in the connector, not bent and will make good contact with their mate. I have seen them fail, and had this once on a 1st gen, and it drove me batty trying to figure it out.

Processor resets and lost tunes would indicate either a problem with the processor board (loose in the socket? Pins pent?) or a massive power problem. Or something like a ground loop.

TunerStudio will log processor resets. I believe there is a telltale light in the bottom right which will be red if the processor has reset.

13.7V is probably not enough. Alternators should charge at over 14V. It's worth checking to see if you have AC on the output.

Your grounds should go directly to the engine block for best noise resistance. Then the engine block should be wired to the chassis/battery with a nice thick cable. Preferably two for some redundancy.

If you have mixed and matched hubs and pulleys, then STOP. That explains all your timing craziness. A hub and pulley set are matched. If you've put an S5 pulley on an S4 hub then the timing marks are wrong. I would suggest getting a known good matched hub and pulley.

You can also find TDC and remark.

 

Just make sure that all the pins are tight in the connector, not bent and will make good contact with their mate. I have seen them fail, and had this once on a 1st gen, and it drove me batty trying to figure it out.

Processor resets and lost tunes would indicate either a problem with the processor board (loose in the socket? Pins pent?) or a massive power problem. Or something like a ground loop.

TunerStudio will log processor resets. I believe there is a telltale light in the bottom right which will be red if the processor has reset.

13.7V is probably not enough. Alternators should charge at over 14V. It's worth checking to see if you have AC on the output.

Your grounds should go directly to the engine block for best noise resistance. Then the engine block should be wired to the chassis/battery with a nice thick cable. Preferably two for some redundancy.

If you have mixed and matched hubs and pulleys, then STOP. That explains all your timing craziness. A hub and pulley set are matched. If you've put an S5 pulley on an S4 hub then the timing marks are wrong. I would suggest getting a known good matched hub and pulley.

You can also find TDC and remark.

 

I finally got motivated to start working on the car again. I am trying to rule out variables to see if I can narrow down a culprit.
So far, I've:
Swapped both coils for previously working ones
Swapped CAS for another that passes FSM spec
Sanded and cleaned my battery ground lug and MS ground lug
Tried with my alternator disconnected/connected (no difference other than load)
I also updated to 3.3.2 firmare.

I am able to get it to idle and start nicely without pedalling. EGTs are way down from before, around 700F at idle. I am using a #1 Trigger offset of 270* to get it to run at all. 0* offset won't even start. With the CAS installed per FSM, 270* offset matches up with the timing light perfectly with timing at -5* locked/zero split. Good spark from all 4 coils.

Here's where it's weird. It will idle great, but if it blip the throttle or slowly roll into it, once I get around 2000RPM, I get crazy sync loss and it sounds like an erratic rev limiter. Once the revs drop back down, it's fine. It might be similar to what's described at the bottom of this page. I have both triggers set for Rising Edge. I tried each and both at Falling Edge and saw no improvement. VR pots have been double checked.

At this point I'm wondering if I should just cough up the dough for a FFE trigger kit, assuming that CAS noise is my problem. Sync loss is rampant.

Below are a tune and several logs, sync loss, trigger, tooth, composite,ect.

Sync1
Trigger1
Tooth1
Composite1
Composite2
Trigger2

TUNE

Thanks for any help

 

is it just me or is this screaming of a polarity issue with one or both sides of the trigger wiring and or the trigger edge settings

i am no megasquirt tuner , but most systems use the falling trigger edge ( on reluctor signals ) for a good reason

 

Your trigger wires might be wired backwards that causes timing to be way off. I'm no expert eater but aaroncake talked about it in his Cosmo install video #24

 

IF I have one of the VRs backwards, could I switch the jumpers on the board?
DIYs instructions say to jumper the following from the main board to the Zeal card.
IAC1A >> VR2 +
IAC1B >> VR2 -

IAC1A and IAC1B are the pin inputs from the CAS in the harness. I can try tomorrow to switch these on the board. I am certain that from the DB37 connector to the CAS connector that the wires are correct. From the DB37 to the processor, the jumpers can dictate which CAS pin is actually going to the correct part of the VR circuit.

The onboard VR circuit is harder to mess up since it is all on the main board and the jumpers are better marked.
TachSelect >> VRIN
TSEL >> VROUT

Regarding Falling/Leading edge,
Aaron's writeup and DIY's writeup both specify Rising Edge.