DIY LTCC or similar system for LT1s

[spfautsch]

I suspect you have an air leak in the evap system. Canister, fuel tank, or somewhere in the middle.

That's certainly a possibility on a 25 year old car w/ 170k miles on the clock. When I have time I'll dig into it. Honestly I can't say I'm sure where the charcoal canister is on a y-body. I've already replaced all the soft vacuum lines on the car. Whatever the case, I'm hoping this might explain my very low TPS cruise surge / miss that I've been chasing.

When you switch on the ccp it doesnt get pig rich at all, It even leans out more. based on blms.

I'm glad you noticed that - I've never messed with the canister purge solenoid but that fits a few scenarios that could explain my difficulties.

When I turn on ccp on at idle it gets pig rich immediately and blm dropping to 108 and closed loop freaks out.

That's a much more pronounced change that makes me wonder if you aren't seeing raw fuel sucked past the canister, into the evap line and then the intake plenum.

I also scored some document with description of the dwell tables and scalars used with ls1 coils, regarding startup and coil temperature. It is freaking modelling with tons of variables to predict coil temperature, so it is indeed the coil temp related than coolant temp.

Yeah, as I'm certain I've mentioned this previously - coil temps != coolant temps, so the phrase "I told you so" comes to mind. I think I'd mentioned creating a multiplier variable or filter function or linear equation for the controller that would allow the user to describe the relationship between coolant temp and coil temp based on where the coils are physically located (on the block or fender, etc.). There's all kinds of under-hood airflow modeling and coil mounting minutia that comes into the picture as well, not to mention ambient air / intake air temps. It's a crap-ton more complicated than a+b=c. Although, on a side note I don't believe any of this is necessary to make a very functional system.

To post some positive news - over the previous holiday weekend, when I got tired of working on household maintenance drudgery I spent some time gathering photos for the PCB assembly manual. I also found a better quality DMM and uncovered another variable that fine tuning of improves the voltage sense logic. The results have been extremely promising on the bench. Hoping to have the PCB assembly manual in a rough draft form soon.

Meanwhile, it's 28F / -2C outside with intermittent flurries so I doubt I'll log any miles on the Green Monster this week, possibly even this month.

[NomakeWan]

That's certainly a possibility on a 25 year old car w/ 170k miles on the clock. When I have time I'll dig into it. Honestly I can't say I'm sure where the charcoal canister is on a y-body. I've already replaced all the soft vacuum lines on the car. Whatever the case, I'm hoping this might explain my very low TPS cruise surge / miss that I've been chasing.

Remove the rear passenger wheel well cover and gave into the abyss of the rear bumper. There it is, the big giant canister that makes replacing the #24 bulb in the passenger rear marker light an absolute pain in the you-know-what. ;)

[kur4o]

I got some good wheather, and did some cold start ups.

Fired great without issues on the first try.

I also played with the tune, adjusting transient fuel tables. It was a short idle test but I can definitely say there was some improvement noticed.

There are 2 tables boil time and some impact factor. The boil time needs to be increased, while the factor is reversed, lower values = longer to evaporate. I got insight from some b-body export calibration. Some more digging will be needed to fine tune this tables.

Ordered new plugs ac 41-110 and 180*F thermostat. I hope I can finally get rid of the idle problems and lifter issues with false knock is actually a tuning issue.

[spfautsch]

I also played with the tune, adjusting transient fuel tables. It was a short idle test but I can definitely say there was some improvement noticed.

Are we going to have to wait for you to write your book: "tuning antique fuel injection systems for dummies"? :-) I kid, but I'm definitely intrigued.

Meanwhile I've also had a bunch unwanted cold start tests, as the test mule has once again been forced into daily driver service due to a ruptured brake booster on the dirty diesel sucking all the fluid out of the system. Although I'm only having to go in 3 days a week, I had some pretty cold starts (one day in the 20s or -7c) and also had the distinct privilege of driving the beast in snow flurries. It hasn't missed a lick and fuel consumption has averaged about 23mpg. I'm also happy to report that the new Arduino based wideband controller "just works" although readings are off the chart lean until the engine (and I assume the exhaust system) comes up to full operating temperature.

I've been swamped with home improvements and emergency vehicle repairs (why do they always break down in winter?) so haven't made much progress on the assembly / install guides, but hopefully things will calm down soon and I'll be able to get back to that.

[kur4o]

I started the tuning season and finally managed to find some bug.

I exhausted all other option of failure but the controller.

On some driving and a hard 2nd WOT pull, and let off, some random misfires started. Engine changed sound and rough running like you just broke off something.
The misfires were most pronounced at low speed low tps. ANd it almost died at idle, some heavy surging. I tried to isolate a cylinder but in the logs it was almost anywhere, Left right bank equal gain of oxygen in the exhaust.

I drive a mile with that condiftion before pulled off. Looks like after heavy deceleration it was more pronounced.

After 10min with engine off. It started fine and drove fine after that with no signs of anything bad happen.[I hope the bearings survive the punishment]

You can see the WOT pull at 1440 on log1 and after that till engine off it was having the missfires. The second log is after engine have been off for some time.

[spfautsch]

After 10min with engine off. It started fine and drove fine after that with no signs of anything bad happen.

Couple questions:

1) Did you attempt to restart it immediately, or were the 10 minutes needed before it would re-start?

2) Can / have you tried to reproduce?

I wouldn't blame you for not attempting, but I've never encountered this so I doubt I'll be able to reproduce as I've done plenty of WOT pulls to 6500 and let off. If you do, and can reproduce throw it in neutral immediately, get off the road, shut off and try to restart (immediately).

My guess here is that it ("it" being the controller logic) jumped out of sequence 90 degrees due to slop in the timing set or some other factor I'm not considering. Working from this assumption, I should be able to make the firmware so it continually verifies sequence so the worst that could happen is a couple cylinders miss before sequence is re-established. It's something I've intended to do for some time but have had other opportunities that required attention.

I'm not necessarily pointing the finger at them, but the number of DTCs you have coming and going through both logs is quite alarming. I'm not going to go in detail on this, but it's probably not a good idea testing an experimental ignition system in a vehicle with so many DTCs.

Off topic for kur4o's issue, but as you're all probably guessing I still haven't made much progress on the installation manual. I do intend to summon the motivation to make the project more complete and accessible, but the lack of interest by others who possess kits makes me start to question if there's enough interest to warrant it. Are you waiting on me to get off my duff and complete the manual or ??? Honest feedback would be truly appreciated.

[kur4o]

I didn`t try to start it immediately due to checking all possible cause for it. Coils, cables opti logging.
I didn`t try to reproduce, wanted to be sure for the cause first.
If I try to reproduce , better to have some logs with it.

The dtcs are false, you can ignore them. There is no DTCs set. I stream other data instead of the dtc bytes.

I too have made hard wot pulls so far without issues. The main difference here is that I let off immediately upon upshifting and there was very rapid deceleration with hard braking[It will give very fast reverse movement of the chain].

It didn`t feel like 90* off. It was more like on/off of all coils. Also definitely rougher running. The misfires were mainly on higher load conditions. Like weak spark or something.

You can see at the very end of the first log, map and rpm jumping up and down.

The logs have some higher rate of errors and some data drops. I suspect bad connection from the dyi serial cable.


So far I haven`t had any issues at all, it will be nice to have an installation manual, some pictures with connectors and wiring diagram will do it all. I have some temporary coil mountings and I am doing some fine tuning and testing recently.

Digging in the crap idle condition I have, transient fueling, cyl trims, EOIT, CL stability and so on. I do have some valuable data with logs, but no definite answer how and what.

Time to start part3 of idiots guide.

[spfautsch]

I didn`t try to start it immediately ...

If I try to reproduce , better to have some logs with it.

I can build a serial console command that will create the condition on demand. I won't ask you to test and possibly grenade your engine. Better let me do that with a $20,000 build with forged internals. Or not.

The dtcs are false, you can ignore them. There is no DTCs set. I stream other data instead of the dtc bytes.

Thanks for clarifying.

It didn`t feel like 90* off. It was more like on/off of all coils. Also definitely rougher running. The misfires were mainly on higher load conditions. Like weak spark or something.

Step back from your assumptions and imagine for a moment what a 90 degree v8 would act like if we instantaneously added 90 degrees of delivered spark advance on top of the PCM's commanded advance. Or say had a traditional distributor installed one or two teeth advanced. The spark is going to be lighting the mixture off just after the piston has started compressing it. At this point in time the mixture will be very inconsistent, with the fuel vapors concentrated near the piston and mostly air towards the top of the cylinder. Combustion (when it happens at all) is going to be extremely weak due to almost no mixing of the charge and very little compression.

Also, this scenario is really the only possibility I can imagine that hasn't already been encountered and resolved.

So far I haven`t had any issues at all, it will be nice to have an installation manual, some pictures with connectors and wiring diagram will do it all. I have some temporary coil mountings and I am doing some fine tuning and testing recently.

Thanks for the feedback, but you've already jumped off the cliff with me. I was mainly asking for opinions of others who have purchased kits. I'm seeing more and more folks on the corvette forums doing LS swaps on these cars and I would hate to spend a lot of time and effort only to sell another five or six kits that never get used for their intended purpose.

[kur4o]

I would disagree with the out of phase 90 degree theory.

I looked at some graph first case is spark to be delivered in an advance which will result in 20* before middle of piston travel upwards on compression cycle.The engine will run like total crap and likely will stall, the piston will move up while ignition will push it down. Total stall situation.

If the spark is retarder after tdc 90* it will be total power loss. I know how that sounds so it was definitely not the case.


Looking at the logs check 1587 to 1596. The engine is idling in failure mode[perfect idle with awesome map readings]. After that you can see some light throttle pushes. I was checking stuff here to isolate cause.
On each throttle push massive misfires with some knock from the engine. Like total spark failure, or very weak spark blown away by the pressure.

I can only speculate something freak out the dwell calculation. or the spark advance signal taken from the pcm got phased out in a death loop with dwell starting too late.

I also dont rule out some hardware breakdown, cables, wiring, coils got too hot. Not very likely[due to the recovery rate] but still a possibility.

[spfautsch]

Looking at the logs check 1587 to 1596. The engine is idling in failure mode[perfect idle with awesome map readings].

Ok, this rules out jumping out of sequence. Your initial description of how it was running afterwards seemed to indicate it was consistently bad. Sorry, I was busy entertaining a house full of people yesterday and didn't have a chance to look at the logs very carefully.

Take a look at your ignition voltage in that log. There's a very abrupt sag in voltage that starts @ 1442.6.

I'd be interested in seeing what the controller thought coolant temperature was here. But we'd need to be logging the controller to know that.

Which dwell table are you using again? Where are your coils mounted in relation to exhaust headers, etc?

I can only speculate something freak out the dwell calculation. or the spark advance signal taken from the pcm got phased out in a death loop with dwell starting too late.

All these calculations are done on the fly so your issue should have cleared up as soon as whatever variable responsible returned to normal.

I also dont rule out some hardware breakdown, cables, wiring, coils got too hot. Not very likely[due to the recovery rate] but still a possibility.

The coils too hot idea is another reason I'd want to know if a stop / re-start resolved the issue.

At this point I'm leaning towards a combination of a dead / weak charging system possibly coupled with an intermittent open ECT connection making the controller think coolant temp was pegged cold.

[NomakeWan]

Off topic for kur4o's issue, but as you're all probably guessing I still haven't made much progress on the installation manual. I do intend to summon the motivation to make the project more complete and accessible, but the lack of interest by others who possess kits makes me start to question if there's enough interest to warrant it. Are you waiting on me to get off my duff and complete the manual or ??? Honest feedback would be truly appreciated.

Thanks for the feedback, but you've already jumped off the cliff with me. I was mainly asking for opinions of others who have purchased kits. I'm seeing more and more folks on the corvette forums doing LS swaps on these cars and I would hate to spend a lot of time and effort only to sell another five or six kits that never get used for their intended purpose.

Hey there! Sorry for long time no response. I actually saw this post yesterday (I have email notifications) but I was also in the middle of a huge drive from one end of the state to the other. Now that I'm back home I can answer.

For me, I have all the parts I need (save for the opti passthrough I need to make because I don't want to splice wiring, and the actual spark plug wires because I haven't finalized a coil mounting solution). I was waiting on the code to mature a little, but when that happened, it was already a little "too late." Not because I don't want to do this--I very much do. In fact, now that Torqhead has pulled out of the C4 market indefinitely, the DIY-LTCC is really the only way to go. But because I ran into several critical issues with my '95 that required a lengthy teardown. The car is currently up on the lift in my garage in a disassembled state. I'm hoping to have it assembled by summer...I'll already be missing the registration deadline on it. Tons of things are changing on that car, and it was going to be my testbed, so since it's down so are my plans to test the LTCC.

My dad's '94 still runs (in fact it's the car we just took on that massive road trip), but it has issues that make it not a good candidate for testing right now. Chief among them is the noisy serial bus and, more importantly, the opti signal is currently doing something that defies logic. See thread here: http://www.gearhead-efi.com/Fuel-Inj...ion-at-low-RPM

And since the DIY-LTCC still relies on the opti signal to run, yeah.

So at least for now, the project is shelved on my end until I have a testbed that can host it. That won't be until I get my '95 reassembled.

[spfautsch]

For me, I have all the parts I need (save for the opti passthrough I need to make because I don't want to splice wiring

If you have a male and female plug from a heated 4 wire O2 sensor you can modify them with a dremel for this application (the key is on the opposite side on the opti plugs). Unfortunately most of the O2 sensors use stainless steel conductor so it's nearly impossible to solder.

My dad's '94 ... the opti signal is currently doing something that defies logic. See thread here: http://www.gearhead-efi.com/Fuel-Inj...ion-at-low-RPM

I skimmed over that thread, definitely perplexing. Sounds like it's time to start looking for a new J520. Check your PM.

Thanks for the feedback. I'm somewhat in the same boat - mine is currently sitting in the garage with half the interior torn out. Found the warning chime (which has never worked) was draining my battery due to a cap that had developed an internal short. Lucky for me that can only be accessed by pulling the dash pad. While I was in there I decided to look at the PKE since it's been acting peculiar and replace any other caps in the head unit, etc. that tend to leak with age. As my luck runs DigiKey is four days behind filling orders, and they sent the wrong size caps for the tone generator and PKE so had to re-order different ones. I'm two weeks in what should have been a three day project, and the caps I ordered on Saturday still haven't shipped.

[kur4o]

I did measured voltage at battery tons of time. There is this phantom voltage drop inside the car. At battery and alternator is solid 13.6-14.2 volts.

There is some drop at very high rpms. Suspect a belt slipping, or alternator shut off. Anyway the coils and controller are wired through a relay and get juice from battery with good solid 5 mm2 cable.

They are on top of engine away from headers heat. They do get hot to touch as anything else under the hood.

To summarize possible causes as most likely to less likely.

1.Bug in the controller that cleared upon engine restart.
2.Coils overheated and some built in protection kicked in.
3.PCM freaked out and killed the est line, or noise on the est line.
4.Controller overheated and bugged out.
5.Leds overload the controller at very high rpms.


To clear some of the possible causes, the condition must be duplicated, and engine restarted immediately.
Some visual control of leds also can point to some problem. Frozen leds on will mean the coils overcharged.



When I try to duplicate the issue, I want to get some logging of the controller done. What would you suggest.

[spfautsch]

Before you go to any crazy extremes trying to reproduce, can you verify that the ECT pid from the controller generally matches what the PCM is reporting? To do this enable logging, reflash and then give the controller this request over the uart (followed by a cr+lf).

L123C56

This will give you the following pids once per second:

engine runtime in seconds, rpm lookup index, voltage lookup index, coolant temp in *C, map lookup index, dwell target lookup index

Field 1 will let you roughly correlate to the engine runtime pid in the .eedata log. If the coolant temp doesn't generally match what you're getting over the ALDL that's a big red flag.

Most likely the big voltage sag at high RPM is causing a condition where the controller is targeting quite a lot of dwell / charge time.

If the ECT sensing is also wrong / off / pegged at minimum this has likely caused a situation where the controller is trying to dwell your coils more than is physically possible (720 degrees of crank rotation minus spark advance and recovery time).

You have a serious issue if your PCM is seeing < 11.5 volts at WOT and high RPM. The PCM (as well as the diy-ltcc controller) is adjusting injector lag time based on this info. If it's inaccurate fueling will be inaccurate. Same with dwell. What's worrying is the PCM was seeing 11.3 volts when the cooling fans weren't even on (assuming these pids haven't been modified in your definition).

Edit: I forgot to specify, but just drive this setup "normally". No WOT passes recommended at this point. I just want to see what the controller thinks is going on.

[kur4o]

In preparation for some testing.

Updated to latest version of firmware, It turns out I was running v 40.
Enabled logging and engine position.

On key on engine off, I got -40 temp and 5 for voltage[pcm read 12.0 at the same time].

I guess they got picked at the time of starting.

Coils set to 514a.

I am also investigating that voltage drop issue. I suspect belt slipping or alternator switching off at very high rpms[around 6000]. I also find that alternator bearings are busted again. Installed new belt + new belt tensioner. I am really clueless on this[ changed 2 bearings sets in 1000 miles interval].

Also change the voltage byte stream to one of the injectors. Now it will show voltage measured at injectors. There is some good 0.5-1v drop inside the car where pcm reads voltage[fuel pump line]. I also bypassed fuel pump from the circuit on a separate fuse.