DIY LTCC or similar system for LT1s

[kur4o]

All 8 coils hooked up for the test.
There were total of 12 led events before it fired off and I got a led on second or third crank TDC which is pretty good.

It only takes some time to fire up only on first start, after it have been sitting for some time[First cold start]. I suspect fueling since I was seeing coil firing on the leds. After the first start it fires off almost immediately[0.3s range]. After it get soaked on the first short try. It fires off immediately on the second try. I might play with FIC constant or first prime pulse.

[spfautsch]

It only takes some time to fire up only on first start, after it have been sitting for some time[First cold start]. I suspect fueling since I was seeing coil firing...

I personally think you have a badly leaking injector, fuel pressure regulator, or bad check valve in your fuel pump. But it could be anything. Unfortunately without having known perfect fuel injector data, it's going to be very difficult to surmise where the problem is.

Whatever the case, for the benefit of everyone else watching this thread, kur4o sent me a PM of a video of his start attempt. Out of respect for his privacy I'm not going to share it. But I do plan to go over it frame by frame. My initial impression is that I have no reason to doubt it. So I think we have a winner!

kur4o let me know via PM how I can get the prize to you and in what form you would like it (if other than 100 USD).

More importantly let me know how it goes when you try to re-start it after cool-down, etc.

I think I'm going to go start mine if only for amusement purposes. My neighbor already refers to it as "the engine" so what harm could come of some more punishment of their aural senses?

[spfautsch]

Here's another release.

0.9.41 - 9/22/2020

Removed all references to arduino digitalRead() and digitalWrite()

Added EST passthrough in EST rising edge interrupt handler for faster first dwell

Increased MAP vs RPM multipliers slightly in cruising range

I've driven this about 275 miles without major incident. Mileage on the last fill-up was 24.3 mpg, which was somewhat surprising since I haven't exactly been babying the go pedal. Cold starts have been a little slobbery, but the ambient temp has been dramatically lower in the mornings the past week or two. Then this morning it set a DTC43 when starting, which cleared on it's own after a few seconds of runtime.

I was somewhat concerned with this DTC when I hooked up my laptop to see what it was, because eehack has a description of "ESC Failure" for it. But EagleMark's original thread [here] describes it as a "Knock Sensor (KS) Circuit". So not sure what to make of this. I haven't seen the CEL light with the engine running in around 2 years. I didn't even think the '95s were equipped to sense a disconnected knock sensor circuit??? Would love to have a better description of what this DTC trigger conditions are but haven't been able to locate where these live in the 2000+ pages of the FSM. Whatever the case I'll report back if it reoccurs.

Still haven't heard anything further from kur4o - hoping everything's well. I'd like to get you your prize money and hopefully try to see if we can further investigate the light acceleration stumble you mentioned. That is, if your rig is still in running order...

[NomakeWan]

Here's another release.

0.9.41 - 9/22/2020

Removed all references to arduino digitalRead() and digitalWrite()

Added EST passthrough in EST rising edge interrupt handler for faster first dwell

Increased MAP vs RPM multipliers slightly in cruising range

I've driven this about 275 miles without major incident. Mileage on the last fill-up was 24.3 mpg, which was somewhat surprising since I haven't exactly been babying the go pedal. Cold starts have been a little slobbery, but the ambient temp has been dramatically lower in the mornings the past week or two. Then this morning it set a DTC43 when starting, which cleared on it's own after a few seconds of runtime.

I was somewhat concerned with this DTC when I hooked up my laptop to see what it was, because eehack has a description of "ESC Failure" for it. But EagleMark's original thread [here] describes it as a "Knock Sensor (KS) Circuit". So not sure what to make of this. I haven't seen the CEL light with the engine running in around 2 years. I didn't even think the '95s were equipped to sense a disconnected knock sensor circuit??? Would love to have a better description of what this DTC trigger conditions are but haven't been able to locate where these live in the 2000+ pages of the FSM. Whatever the case I'll report back if it reoccurs.

Still haven't heard anything further from kur4o - hoping everything's well. I'd like to get you your prize money and hopefully try to see if we can further investigate the light acceleration stumble you mentioned. That is, if your rig is still in running order...

As someone who owns a '95 and had that CEL himself, yes, they're wired the same way as all the older C4s (which is why the knock sensor part number is shared until 1996). The knock sensor on our cars has a resistor inside the sensor that the PCM uses to detect whether the sensor is physically present. When this resistor inevitably breaks or fails to be in the correct range, it triggers DTC 43. But this doesn't mean the sensor failed--it just means the resistor is no longer in spec. You can fix it by replacing the sensor, but I just disabled the DTC in my BIN since I could see in EEHack that the sensors were still working perfectly fine.

[spfautsch]

Thanks for the clarification. I remember reading something about the resistor for checking the circuit continuity but have a growing amount of information to keep track of and waning motivation to search thoroughly. Whatever the case I found it odd that the DTC cleared itself almost immediately, and these are 3 year old AC Delco O.E. sensors.

Since the last couple revisions I've had a couple attempts that didn't start on the first try, most recently a hot re-start this afternoon. These were both instances where I cranked around 1400-1800 milliseconds and then let off because it was obviously not going to run without intervention (I did not try opening the throttle). It felt like it hit on a few cylinders and then flooded (or possibly underfueled). Second, relatively long crank started and felt pretty normal once running. I'm going to try catching a log of one of these in hopes of finding the cause.

[spfautsch]

Here's another updated release. Made a minor modification to the dwell pass-through logic on first fire.

Since the last post I had one cold restart (naturally wasn't logging it) where it felt like a coil fired waaaaaaay before commanded spark advance (i.e. it sounded like the engine briefly came to a dead stop after a couple compression strokes, and then immediately started and ran perfectly for the next 57 miles). Hoping this change resolves that in the rare occasion it might re-occur.

I'm once again making a concerted effort to log the controller with every start to capture every bit of data possible. It seems like once running things are near perfect, as witnessed by the average fuel economy of around 23 mpg over the last 5-6 fill-ups. Aside from some A/C related surge and a bit of closed throttle bucking in 1st through 3rd that I've mentioned ad-nauseum, it seems like my tune is pretty decent and I want to believe none of my drive-ability issues are related to the ignition system. Due to my retarded brother-in-law the test mule has once again been pressed into daily driver service so I'm getting more test data than I normally would this time of year. Apologies to those who have an actual learning disability.

Anyway, after about 10 starts with this revision there's nothing to report. Over the next couple commutes the plan is to see how far I can stretch economy. This was my intention after the last fill-up, but a 15 mile interstate parking lot snarl made that an exercise in futility.

I've been hearing from kur4o via PM and things sound promising. When he's convinced it's mediocre, I'll be elated - over the moon.

[kur4o]

Finally have some time to update the thread.

New caps installed and new firmware loaded. Still using 0.9.40 version. Starts every time without issues. I also went for some long drive logs.

Car feels great. Overall better throttle response and much nicer low and mid range. Some of the low rpms, high load drone disappeared. Did some moderate wot pulls, but not hard enough to make some impressions. Still fighting with some mechanical lifter problems.

I feel some nice double platinum or iridium plugs will make it even nicer.

Here is the official video of the last drive with coils running the engine.
https://drive.google.com/file/d/1zr5...ew?usp=sharing

Ignore the coils attachment. It is still on testing fixture.

Also some eehack logs with the COP drives.


As a last note I want to Thank you for putting such an enormous R&D time, developing the controller. It is really amazing that it works so good.

The car feels so much better. Too bad I didn`t have enough time recently to conduct more tests. Some fine tuning of the dwell times, spark advance and trying some other plugs will make it even better.

I will contact you via pm about the reward.
I also want to do some more videos, so people can get more confident about the quaility of the work being done. I guess the best place to put them will be on some video sharing sites.

It is still too advanced for the average Joe but some install guides with a list of the needed parts will unleash demand.

This is to be considered the best mod on par with ls1 injectors patch.

As a side note project I might get access to a spark plug pressure chamber to test some dwell settings.