DIY LTCC or similar system for LT1s

[kur4o]

I know connecting plug wires to plugs in an f body is a pain with headers, and this is a cruel request. But my suggestion is to step back, disconnect the system and make your car start on the first attempt and run perfectly with the distributor. I've wasted an immense amount of time on my own setup trying to solve problems that I wasn't sure existed prior to the ignition system because I changed too many variables at once (engine rebuild, unknown fuel injectors, imperfect tune, etc.). I have no desire to do that again.

I did just that and I somehow got the freshest start from years - on key bump and 0.1seconds according to my log.
Actually #6 is the PIA, I had to crow under the car and it was almost impossible to reach.

Let`s make it clear first so we don`t miss anything. Car is running perfectly on stock single coil and opti.

Controller wiring and soldering is out of question since I managed to run with the COP setup for 10 minutes without issues. Sorry no video and logs were taken.

After letting the car cool down for 2 hours, and nothing touched except the switching from N to P and reposition the coil, I got all the problems I mentioned earlier.
Suspecting slower cranking speed freaks out the controller. I got the latest firmware with fixed crank spark table, coil selection and logging enabled.

JUST for the testing I unplugged the coils and start cranking while monitor the led output, so noise on opti wires is also out of question.

On the very last starting attempt I got some nasty popping from the engine, which might indicate that a coil was fired at the wrong time. I hope I didn`t bust a bearing.

[spfautsch]

Let`s make it clear first so we don`t miss anything. Car is running perfectly on stock single coil and opti

That directly contradicts this statement.

Started on the second try which is not unusual on my setup.

Controller wiring and soldering is out of question since I managed to run with the COP setup for 10 minutes without issues.

Not necessarily. I want to see your board to make sure the electrolytic caps aren't installed backwards, and that the right resistors are in the right places. Please take no offense, but simple mistakes can happen.

I also would like to confirm continuity between the coil driver pads on the board, and the correct pin on the coil connectors.

Sorry no video and logs were taken.

I want to compensate you for your trouble. If anyone else thinks I should pay you the reward I will because I trust you ran the engine on it. But if your setup wasn't starting and running perfectly to begin with this going to be a waste of time for both of us.

... touched except the switching from N to P and reposition the coil, ... Suspecting slower cranking speed freaks out the controller.

I've never heard anything indicating there could be a difference in cranking speed with the shift selector in P or N. The transmission (if it's also healthy) will be in neutral in both positions. The only difference is the parking pawl will be engaged in P.

Simple way to rule this out is to put trans in N and crank.

I got the latest firmware with fixed crank spark table, coil selection and logging enabled.

I really need to see what you're running. I don't mean to be insensitive but there is a language barrier we're working around, and I'm going to have trouble helping you without eliminating all doubt.

JUST for the testing I unplugged the coils and start cranking while monitor the led output, so noise on opti wires is also out of question.

I don't think anything is out of the question at this point.

I hope I didn`t bust a bearing.

It's more likely you would have broken the nose off your starter or the starter boss off the block. Bearings typically die from too much heat due to lack of lubrication.

I have things I have to be doing at the moment and I'm currently 4 hours behind schedule at the moment. So I may not respond to additional posts for two or three days.

[NomakeWan]

$#!1!!! I'm sorry! Did I short you any other parts?

Nah, just the microcontroller. Like I said, wasn’t a huge deal. Also I found the likely cause for my overheating—it appears my heater core bypass is leaking ever so slightly. Wasn’t a problem until 100F ambient days crept up and raised the cooling system pressure. The radiator was actually clean! I only noticed because on startup this morning I got the “low coolant” light for a few seconds. Sure enough the lower reservoir was almost empty.

Easy enough fix, gotta yank it off, tear it apart, and reseal all the joints. May as well flush the cooling system while I’m at it.

Best of luck to kur4o getting this thing to work. I’m sure it’s actually something super simple.

[kur4o]

Since your keep questioning engine health, here is some recent log plus the logs from COP testing.

cop1 test log is from the successfull idle.

At cop2 test the @920 is with stock ignition hooked back.
@90 is the first attempt,@92 is the second attempt where the engine started on ?x? cylinders [3 or 4 max] and the leds shows that there was spark only to the running cylinders, @102 again started on less cylinders and runnig like crap, not all leds were flashing again, @111 it was like runing on 1 cylinder, Than 2 no start at all attempts where some of the leds freezed on. @430 I was doing some clear flood cranknig with coils disconnected, again watching the leds didn`t look rigth and some times there were some constant on some times all went off during cranking. @735 was the last attempt with coils hooked.

I am running 0.9.38 fixed with

#define LOGGING 1

#define DWELL_TABLE "D514a_dwell.h" // specify the desired dwell table based on coil type

#define R1_VAL 9850.0 // design 10000.0
#define R2_VAL 3750.0 // design 3900.0

The crank table is not fixed yet.

I can`t figure way to upload the sketch or download it from controller.

[spfautsch]

NomakeWan I'm really sorry to have left out the most important part. When I get settled into the condo Sunday I'll refund you half.

This gets at another reason why I'd like to see pics of your board kur4o. What if I gave you 4.7k ohm resistors instead of 470? I tried being extremely careful with all that stuff, but as witnessed above, mistakes happen.

Since your keep questioning engine health,

Again, not meaning to be rude. But if you say it doesn't always start on the first attempt with the distributor ignition, something's wrong. Since revising my injector flow constant approximation by around 1.5 lbs / hr and appropriately scaling my prime pulsewidth multiplier I've had the best starting reliability since 2017 before I put in the big injectors. It simply hasn't missed a lick.

I'll try to look at your logs tomorrow if things aren't nuts at work.

I'll have some debugging tools we can use to look at sequence detection, coil dwell and fire, etc. But that's not happening today. I need to be packing.

Is there any possibility you have a really, really sloppy, worn out timing set? If your cam is briefly turning backwards when an intake lifter is traversing the initial opening ramp during cranking, strange stuff could happen also.

Another thing - you didn't answer but was your EST line disconnected from everything except the controller when you were running it?

I am running 0.9.38

I don't think there are any significant bugfixes in it, but 0.9.39 is up on my site.

The crank table is not fixed yet.

You should do that. I don't remember what version it was fixed in, but there was an issue where strange stuff would happen if the ECT lookup table detection didn't work.

I can`t figure way to upload the sketch or download it from controller.

You can't download the microcontroller's flash with the serial bootloader. And even if you could it would be the raw intel hex with no symbols, which would be as meaningless to me as looking at a binary file in binary (1s and 0s).

On the other hand, if you're having trouble uploading to the board there's an electrical problem.

And if you want a place to upload stuff, PM me. I have a secure ownCloud instance you can use.

[kur4o]

All components were measured and matched with the print on the board and triple checked. Really there is no issues there. All build test were passed successfully.
ICM and coil were unplugged during testing.
The crank table at the temp range of testing almost matches the preset one so that also can be ruled out.

Chain and crank sprocket is almost new 1000 miles on it. Cam sprocket have some signs of usage.

The first start issue is usually when the car was sitting for a long time. It is some fuel issue that I still can`t fix and it is random. If it doesnt start whitin half second I stop cranking and it is starts right on the second try. This happens only if the car have been sitting for a long time.
You can see this in the log1 where the car starts and run on COP. On log2 it fires immediately after stock coil is plugged back and failed earlier on cop system.

I am in no hurry and will check the headers temp on stock coil, than will be able to test next week. ALso will setup some video equipment and will try to duplicate cranking conditions where led fails.

[In-Tech]

Hiya Scott and everyone,
I can't imagine a better tester than kur4o, when shiola happens and you're not there it can feel like a kick in the teeth.

Man, I sure wish I had a vehicle or test engine here to help. All of this is pretty exciting. I'd kick the dog for y'all but I think she's actually starting to like it :D lol

[spfautsch]

The crank table at the temp range of testing almost matches the preset one so that also can be ruled out.

Why not fix it, it takes all of 2 minutes? If the answer is because you're having trouble uploading to the controller you're going to need to figure that out. If you're using the bluetooth serial adapter I would recommend taking it out of the equation so we know we can talk to the board reliably.

Another thing I should mention here is that the FTDI cable I used when testing coils on the bench was extremely sensitive to the EMI generated by the coil firing. If the cable was too close to any source of noise it would crash something in the adapter and stop receiving data until the USB connector was pulled to reset it. I've never had any troubles with the ones I have in the car, but the UART wiring has some distance between it and the coils. Also the FTDI board is way over on the passenger side of the center console.

The first start issue is usually when the car was sitting for a long time. It is some fuel issue that I still can`t fix and it is random. If it doesnt start whitin half second I stop cranking and it is starts right on the second try. This happens only if the car have been sitting for a long time.

Good enough, we'll call it a leaky injector. But if there's even a remote possibility its related to the opti we're likely going to be chasing our tails. The word "random" gives me some cause for concern.

I wouldn't worry about making a video of the leds just yet.

Here's what I'd like you to try if you have time - get the latest build (0.9.39). Configure it for your setup and then uncomment the following options.

#define LOCATE_STRT_POS 1
#define LOGGING 1
#define DISC_ECT_DEBUG 1
#define PROFILING 1

Find and comment out (or just delete) these lines in the main sketch .ino file (diy_ltcc_0_9_39.ino)

#if defined(LOCATE_STRT_POS) || defined(DISC_ECT_DEBUG)
#error PROFILING cannot be used with LOCATE_STRT_POS or DISC_ECT_DEBUG
#endif

Test compile, upload, then disconnect the DTR line. Disconnect the COP coils and the factory ignition module, have the EST line connected to the COP system, and pull the injector fuses so you don't flood it. Connect with a serial console program and then crank the engine over for about 1 second or just slightly less - about a full crank rotation. Let off the key / starter and then send "p" over the uart (with a cr+lf) before turning the key to off. Do this two or three times, turning the key to off in between each. I should be able to at least get an idea from the logs what's happening.

Also, can you verify continuity from each of the coil driver pads (1-8) on the board to pin C of the appropriate coil connectors pls?

ls2_coil_schematic.gif

When I get back I'll try to find a dead side post battery to put in to test slow cranking starts. I ran mine down a few times working on the rewrite, and I don't recall having any troubles related to slower cranking.

[kur4o]

I have planned to do some initial testing and than fix all the small things, Everything seems smooth before the fireworks started.

I hope we figure it out soon. Bad news is I won`t be able to test at least late on Sunday.

Today I did some warm up with stock coil and headers were running cooler for sure. The idle wasn`t that smooth as with the cop system on. Engine seems quieter on stock coil and exhaust smell was much worse.
I suspect 2 reason for the extra heat just by the coils. Some lag at coil firing induce a spark retard. I still have to calculate lag time in ms vs rpm vs crank degree retard.
Reason number 2 is better mixture combustion by the fatter spark and more heat due to all the fuel burning.

I make a log and you can make a comparison of the blms. Stock coil vs COP. BLM comparison clearly shows the meaning of a good spark.

I use stock gm harness and connectors for the coils so everything is good and traced.
Ftdi usb is currently disconnected from controller and I have zero issues logging and updating firmware.

[spfautsch]

I have planned to do some initial testing and than fix all the small things, Everything seems smooth before the fireworks started.

I hope we figure it out soon.

You cannot fathom how much I'd like to figure this one out.

I beg you to focus on the debugging steps I mentioned with the 0.9.39 build. These may give us some hints as to why your controller isn't detecting or staying in sequence properly.

Today I did some warm up with stock coil and headers were running cooler for sure. The idle wasn`t that smooth as with the cop system on. Engine seems quieter on stock coil and exhaust smell was much worse.

I'm intrigued.

I suspect 2 reason for the extra heat just by the coils. Some lag at coil firing induce a spark retard. I still have to calculate lag time in ms vs rpm vs crank degree retard.

No offense, but I think you're chasing vapor trails here. The amount of time the coil per cylinder coils are taking to fire is likely on the order of a few hundred microseconds (us) and is also likely quite similar to that of the stock ignition coil. The controller itself is introducing perhaps a few dozen microseconds of delay. We've discussed this previously. The amount of spark retard this could be generating is negligible unless you're turning the engine > 10K-12K rpm. At idle speeds it's quite certainly irrelevant.

Perhaps there's another logical explanation that aligns with:

Reason number 2 is better mixture combustion by the fatter spark and more heat due to all the fuel burning.

First, let me say how disconcerting it is that you seem to be able to tell a difference in temperature of your exhaust headers when switching between the two ignition systems without instruments (i.e. by feel). In my experience, exhaust manifolds / headers should be hotter than all get out. With mine, within about 15 seconds of engine run time, you'd better not touch them with your flesh or you'll be nursing a blister very shortly. After more than a minute or two of run time you'll be instantly cauterizing the flesh, so blisters may not be able to form.

Working from that tangent, I have to wonder if there's another 3rd possibility: your stock ICM and / or coil have one foot in the grave.

I really, really want to figure out what happened, or at very least what the primary problem is. From your description it sounds to me like either:

a) something causing an immense amount of noise on one of the time-critical inputs

b) something is causing the AVR to reset randomly

c) a damaged AVR

I'm sorry to continue to ask, even though I'm sure based on the questions you've asked that you have a competent understanding of the basics. But I'd really like to see some pictures of the board and the install.

[kur4o]

Still nothing to report.

On the last test with stock coil,
I left hooked the controller and it does seems to function properly when the engine is run on stock coil. All leds flashing with proper operation. I even make a video of the leds when the engine was running with stock coil. Maybe you can PM me some details on upload server.


To rule out any lag in ls coil ignitor or even the coil itself, I plan to do realtime comparison of spark output and film it on tape. Not high speed but hope fast enough to get an idea.

The plan is on warmed up engine to hook #3 coil on the controller and remove #3 from opti. Put on both ends spark plugs and start engine. It will give exact visual confirmation of spark output and duration from stock coil and controller.



I was measuring the radiated heat from the headers by feel, didn`t risk touching them, and the stock coil is some msd unit. You can feel it loose steam on higher rpms but on low it must be strong enough.

[spfautsch]

On the last test with stock coil, I left hooked the controller and it does seems to function properly when the engine is run on stock coil. All leds flashing with proper operation.

PM sent. Out of curiosity, how close to the engine / ignition system (coils and high tension components) is the controller?

The plan is on warmed up engine to hook #3 coil on the controller and remove #3 from opti. Put on both ends spark plugs and start engine.

Another test (assuming this functions correctly with the controller driving one coil) is to run one cylinder from the controller and the other 7 from the distributor. Just make sure to put a grounded spark plug on the loose spark lead to control the severe EMI an unhooked plug wire will cause. Then use a cylinder balance test to see relative strength of combustion.

the stock coil is some msd unit. You can feel it loose steam on higher rpms but on low it must be strong enough.

What about the ICM? Is it also MSD or stock?

[kur4o]

Here is the requested log.

The first 2 attempts the leds show a valid sequence was detected so I tried to duplicate the fail conditions by giving some smaller durations crank time. I did manage to make some of the leds freeze on some of the last attempt. Too bad the phone camera switched off on the most important part. I will try to do another video.

[spfautsch]

The leds "sticking" on when cranking isn't the sequence being detected incorrectly, and it most likely isn't related to what you saw after the successful run where it wouldn't start / run.

This is merely an instance where the EST line hasn't switched low (likely because engine speed is below where the PCM will try to fire it) and the coil that was just dwelled is never fired. This likely happened in your first attempt here.

c @ 90 dwlTgt=7 dwlDgs=17 qIdx=1
d cyl=5 dwlDgs=17 ectTbl=0 qIdx=1
f cyl=5 saDgs=10
c @ 90 dwlTgt=5 dwlDgs=15 qIdx=2
d cyl=7 dwlDgs=15 ectTbl=0 qIdx=2 << coil # 7 was dwelled here but is never fired
c @ 90 dwlTgt=3 dwlDgs=13 qIdx=3
f cyl=2 saDgs=-1 << this is the EST line going low, likely after you let off the key
16,15,15,16,1,
c @ 89 dwlTgt=11 dwlDgs=21 qIdx=3 << my guess is this is the engine turning backwards from a cylinder that was on it's compression stroke
ready

This could be easily resolved by turning all the coils off in the stall routine. But after much consideration I decided not to. I don't think it's a serious problem because the driver (edit: the operator of the vehicle) is either going to shortly turn the key off or try re-starting.

I'm still picking through this, will likely have some more questions for you but here's the first one.

Can you look in your arduino compiler output when you build this and see if there's something along the lines of #pragma message "MAX_OVERLAP=?" please? If you aren't seeing this you may need to go to File > Preferences and set the checkbox next to 'compilation' for the 'Show verbose output during' setting.

I'm interested to see similar logs with one of the coils being driven. Bear in mind however that the EST line may get more noisy with both the controller and the O.E. ICM connected to it due to much lower impedance than design.

Edit: Here's another instance where a coil would have "stuck" on at the very end.

strtDegs=62 strtLRCnt=1 strtSeqDeg=0 invSeq=0 firstCyl=7
d cyl=7 dwlDgs=13 ectTbl=0 qIdx=2 << coil #7 is dwelled here but never fired
c @ 90 dwlTgt=6 dwlDgs=16 qIdx=3
d cyl=2 dwlDgs=16 ectTbl=0 qIdx=3
f cyl=2 saDgs=10
c @ 89 dwlTgt=6 dwlDgs=16 qIdx=4
d cyl=1 dwlDgs=16 ectTbl=0 qIdx=4
f cyl=1 saDgs=10
c @ 89 dwlTgt=7 dwlDgs=54 qIdx=5
d cyl=8 dwlDgs=54 ectTbl=0 qIdx=5
f cyl=8 saDgs=10
c @ 90 dwlTgt=7 dwlDgs=54 qIdx=0
d cyl=4 dwlDgs=54 ectTbl=0 qIdx=0
f cyl=4 saDgs=10
f cyl=4 saDgs=-1
16,13,16,16,3,
c @ 88 dwlTgt=4 dwlDgs=51 qIdx=1

I can only guess why the EST line isn't driven low after #7 is initially dwelled, but if the engine had started it would've fired when #7 came back around. This I might be able to address, though I still don't believe it's a serious problem.

[kur4o]

Some part video of the first log[the first 2 cranks], and than from the second log[full video. You can time stamp it with the logs for better picture.

Also spark to spark comparison from #3 run by stock coil and ls coil. The weaker spark is the stock. The plugs looks different so I switch the cables on the second run to exclude any fd up spark plug.

I looked the video at some low speed and it looks like there isn`t any lag at idle I also give some throttle at second video. Some better player can give better resolution than x0.25.

https://drive.google.com/drive/folde...sr?usp=sharing


I noticed some pattern where some coils will stay on all the time. You crank it for some time, than let it off without turning ign off, than crank it again and it will definitely lit some of the coils full time. You can see it in the video.