95 LT-1 Idle Cell Comparison - Humidity?

Agreed with you there. Guy wanted to sell my friends and I a Mercedes Benz V12 Biturbo. For an unbelievably low price, because he couldn't get it to run. So we went over thinking, sure, why not, let's see what we're in for. Well it turns out there's no physical key for the ignition, so you can't start the car without the fob. But even more hilariously, even if you have the fob, you can't start the car if the car's been left with a dead battery for too long, because then the backup battery for the security system fails and the only solution is to have the vehicle towed to a Mercedes dealer along with you and all the paperwork proving you own the car so they can install a brand new security computer and give you new keys.

Needless to say we walked away and wished him luck.

Here's me wishing you luck that the slipped rotor has been the only issue this whole time. :)

I was thinking that you could add 20 degrees advance across the board to your timing tables to confirm your suspicions about the distributor without taking stuff apart. Turn the key and see if it liked it.

As far retarded as it was I think the best I could've done was lock timing at max (46 btdc) with eehack. That might have given me 16 degrees advance best case, and it would have still ran like crap that way. With the egr the cam generates it wants at least 28 degrees spark advance at 925 rpm idle.

It's kind of immaterial - I'll have it back together this weekend if the "baby" degree wheel shows up when it's supposed to. My good wheel interferes with the k-member by about 2 inches.

After moving the hub yesterday I feel like this had to have happened gradually over numerous driving sessions. As hot as I was comfortable getting the hub without burning the lube out of the bearing, it still took quite a bit of force (with a channel lock pliers no less) to turn the hub back.

https://www.blocklayer.com/degree-wheel.aspx

Here. Print out one of these babies on card stock paper. Works for me in a pinch. Lots of options, too.

Quote:

---

Originally Posted by vilefly

Print out one of these babies on card stock paper. Works for me in a pinch.

---

I'd have to be in a pretty big "pinch" to use card stock, but I guess it could work. The nice thing about the neat little 7" wheel I got off slamazon for $15 is that even though it's aluminum it's a nice high-density extrusion and when I'm not using it for timing an engine I can hang it on my drum kit as a gong for when I want to attempt Rush's YYZ.

Once again I got called into service as a used car salesman over the weekend, otherwise this would have been done last sunday. If I could have found an adjustable MSD dizzy for a reasonable price I'd have been all over that - the time it would have saved would have almost made buying one new worthwhile. After getting the hub back to zero it took almost 6 hours and numerous 3/32" cobalt drill bits to get a hole drilled in the hub to lock it in place. It seems like it's some kind of powdered metal and damn is it hard!

Anyway, I've got a couple odd things going on now - first log was with -2 degrees spark advance by way of setting the 0x012044 "Spark Reference Angle" to 2 because I thought my hub was a couple degrees more advanced than I wanted. Stock injectors with original characterization and individual cylinder trims. Judging by the idle quality I think the timing was actually spot on and my determining of geometric TDC with the engine in the car (and heads on) was off by +2 degrees.

Right bank is quite a bit lean especially before bumping idle timing back +2 degrees and resetting blms to 128. I'm going to pull my plugs from that side to see if one of the iridium electrodes "blew" off.

Spark plug discussion detour: I compared my ACD double platinums to a Bosch 9655 double iridium and a Denso 4713 TT iridium using the highly scientific "eyeball" method, and the Denso TTs gapped at 0.040" produced a far superior spark to all the above (even the copper NGKs I bought in desperation). Detour over, back to the main subject...

For the second log I pulled the 2 degrees of retard back out of 0x012044 and reverted to the individual cylinder trims I arrived at months ago. Right bank still pretty lean. I forced closed loop for a short time and other than having +10 blms on the right bank it wasn't idling too shabbily. I saw a few data frames with 41.5 kpa at ~9.8 g/s on the MAF. The IAC counts hit 24 which tells me I probably need to go to the next smaller IAC bypass orifice because today was another 93F scorcher with swamp-ass humidity levels. This seems to invalidate the usefulness of enlarging the idle air feeds in the manifold.

At this point I think I only have one other problem to locate. But...

In both of these logs it didn't start right away like it was before this problem manifested itself. So I have more troubleshooting to do, because before this started I never needed to crank more than 1-1/2 engine revolutions before it took off.

Hopefully it won't take me another two weeks to track down these problems. I'm about exhausted...

There is a lean condition on both sides, that evolve to one side later, that can be caused by numerous things.
The mechanical issues must be ruled off first. I would start at tapping the hole in the intake blade and load the last good ind cyl trims you have on record. The extra air from the hole can screw up alot of the fueling so as a first measure just make it temporaly tapped.
At this point make an idle log. If there is no improvement on lean bank to bank condition, swap injectors per bank. Now you can rule out a clogged or pissy injector. If the lean bank moves with the injectors the injectors are toast.
The last mechanical issue could be small exhaust leaks. It can affect 02s reading even if the leak is after them. Just for the record did you swap the 02s with delco ones.

If nothing of this help it is time to do some tuning. The maf table needs to be dialed first, than fine tune the individual cyl trims. Than tune the spark and again fueling since spark and fuel at idle are closely related and couple of degrees less or more advance can shoot your afr off chart.

I am having the same hard cold starting issue that developed out of nowhere. It will start right away after it has been running for a while cold or hot. The issue shows if it have been sat for a couple of hours no matter cold or hot is the engine. NO tuning helped at all so far, so i am thinking a mechanical issue could be the problem.

Spark reference angle offset all timimg tables by the amount specified. It doesn`t correct injector timing and any other opti related calculations.

Quote:

---

Originally Posted by kur4o

There is a lean condition on both sides

---

If you look at the first log the extreme lean condition on both sides improves on the odd bank when I manually increase idle timing from 27 to 29.

Another factor is the BLMs - I keep forgetting to increase the keepalive value to 128. I don't necessarily intend to continue running open loop, but that would be something anyone who is (running open loop full time) should do. Anyway, when I reset them to 128 the lean condition improves markedly.

Quote:

---

Originally Posted by kur4o

I would start at tapping the hole in the intake blade

---

There are no holes in the throttle blades and I've re-adjusted them back to fully closed so there should be very little air getting past the throttle at idle.

Quote:

---

Originally Posted by kur4o

... If there is no improvement on lean bank to bank condition, swap injectors per bank.

---

I believe my plan of attack will be putting my big injectors back in. I have the feeling the 150k mile stock Multecs have had the varnish in them solidify from sitting dry in my parts bin for 2+ years. I'm crossing my fingers this is my starting problem also.

Quote:

---

Originally Posted by kur4o

Just for the record did you swap the 02s with delco ones.

---

Yes, these logs were taken with the AFS75s. If you scan through the second log there's a ~30 period starting at 142.1 where I forced closed loop. When I switched back to open loop the odd bank BLM was still 128 but the even bank had climbed all the way to 142. This was with the factory closed loop parameters restored - not a single item has been touched, even 0x26fb Integrator / Transport Delay.

Quote:

---

Originally Posted by kur4o

If nothing of this help it is time to do some tuning. The maf table needs to be dialed first

---

I think my methodology is going to be leaving the MAF calibration alone until all other avenues have been explored. From what I've been able to find on the subject I don't think I'm going to be able to considerably improve on what GM did there.

Quote:

---

Originally Posted by kur4o

I am having the same hard cold starting issue that developed out of nowhere. It will start right away after it has been running for a while cold or hot. The issue shows if it have been sat for a couple of hours no matter cold or hot is the engine. NO tuning helped at all so far, so i am thinking a mechanical issue could be the problem.

---

I had the same warm restart issue before increasing displacement, and I was able to find a sweet spot between prime pulse and cranking VE. Unfortunately the prime pulse table lacks the resolution necessary for much fine tuning in the warm / hot regions, especially with larger than stock injectors.

Switched back to the modded 42lb/hr LS injectors. Still lean on the right bank, but if I force closed loop and let the BLMs adjust a bit it's much better and I'm seeing ~40kpa at idle once it gets up to temp. I think my individual cylinder trims need a lot of work here because looking at idle BLMs on some older closed loop logs the right bank has been 6-12 counts lean of left.

Initial start is still not great so I'm going to re-work the cranking fueling beginning with the base pulsewidth tables.

Quote:

---

Originally Posted by kur4o

The prime pulse is complex as hell, in your case you will have to adjust for new injectors along with the extra cube. The good way is to convert the stock BPW readings to grams of fuels. than add some fuel for the extra cubes and convert from grams to bpw of the new fuel flow constant.

---

If you have a chance please look at the attached and check my math, particularly the lb/hr to g/s conversion factor (<lb/hr>*0.175070002). If this is anywhere close, my previous method of scaling arrived at numbers that were way off. There is scaling in this worksheet for cylinder volume changes and I also added another scalar to account for a delta in pumping efficiency due to cam swap, etc. Note that the cylinder volume scaling doesn't convert volume to air mass because I didn't see a difference when I calculated it outside of the spreadsheet.

per google conversions are

lb/hour to g/s x*0.125997881

g/s to lb/hour x*7.93664144


24lb/hr*0.125997881=3.02394 grams/s
42lb/hr*0.125997881=5.29190 grams/s

Lets take or example 50ms BPW and convert it for 383ci @42lb/hr
50ms[x/1000=0.05second]*3.02394=0.151197 grams of fuel

so 50ms BPW delivers 0.151197 grams of fuel at 24 lb/hour injectors.

346 ci to 383 ci is 10.7% difference.
Hope the 383 ci is correctly calculated and not marketing cubes like 350 is.

0.151197 grams * 10.7% = 0.1673751 grams needed or the new displacement.

SO for 383ci you need 0.1673751 grams /3.02394 =0.05535second BPw[55.35ms] at 24 lb/hr

or 0.1673751 grams /5.29190 =0.03162seconds BPW[31.62ms] at 42 lb/hr.

I doubt the cam will be a factor here since cammed only lt1 fires straight up with stock tables.

Quote:

---

Originally Posted by kur4o

lb/hour to g/s x*0.125997881

---

I'd used that initially but wasn't completely confident in the accuracy.

Quote:

---

Originally Posted by kur4o

346 ci to 383 ci is 10.7% difference.
Hope the 383 ci is correctly calculated and not marketing cubes like 350 is.

---

I was dividing the "new" ml/cyl by the ml/cyl from the original bin and it gives me 1.0999 so essentially 10%. I believe I used one of the online calculators to come up with 788.97ml/cyl, but the new setup is 3.750" x 4.040". I'll double-check this. The stock bin has 717.25 ml/cyl.

Adjusting the lb/hr to g/s conversion factor, a 50ms pw gives 32.87ms adjusted for injector flow and your displacement figures. This should be close enough in that region (8C / 46F). The real trouble spots are the warm and hot restart cells where I can only adjust between 0x02 and 0x03 (3.1 and 4.7ms respectively).

Edit: I noticed in your example you rounded all the precision out of the stock injector flow, which is closer to 25lb/hr than 24. Using 24 and 42 gives 31.4 to your 31.6 so more than close enough...

Quote:

---

Originally Posted by kur4o

I doubt the cam will be a factor here since cammed only lt1 fires straight up with stock tables.

---

Just trying to be thorough. I haven't dialed my VE in yet but the old setup had roughly 10% lower numbers in the low RPM regions compared to stock bin.

Either way it still doesn't want to start worth a damn. I drove it to work today and thought I was going to spend a half hour tuning next to the gas pump after filling it up. I'll repopulate the prime pulse tables with the current results but I doubt it will make a difference.

Attached are the logs from this morning and one from last night. At idle the right bank BLM is about %9 lean of the left bank. Once the throttle is opened the lean condition diminishes. In fact there's a short bit in the log from last night where I held the throttle at 4.7% tps for 10 seconds or so and the right bank O2 came right up and stabilized around 910mv.

The logs from this morning are mostly closed loop because I didn't want to chance stalling it at a stop light, and mostly interstate cruise in 6th gear. Here's the BLM analysis from the longest log.

Attachment 14660

Last night I ruled out a lower intake manifold vacuum leak - I got a noticeable puff of blowby after roughly 7 seconds. I also tried adding some fuel to cyls 2, 4 and 6 in the individual trims, but all that seemed to accomplish was making my eyes water from the unburned fuel.

Tonight I plan to try vilefly's bucket test to check for leaks on top but I highly doubt there's a problem there. The only other thing I haven't ruled out is an exhaust leak on the right side, but I would think I'd hear something if the header was leaking.

I'm starting to think one side of the idle air plenum might be clogged with something. If nothing else shows improvement I might try opening the throttle blades some to give more idle air through the main plenum.

Things are falling into place, but it sure doesn't want to start cleanly and I can't help but think the lean right bank O2 readings aren't related to this.

Some more info I need. What are the headers your using. How long are the primary tubes, and are they equal lenght accross all cylinders. If not it is good idea to measure each tube`s lenght. The distance the 02s are mounted from exhaust ports. It can help dialing the initial ind cyl trims. If the 02s are mounted too far from engine than stock, there is some delay for the exhaust gas to reach the sensor and it must be fixed in the bin CL settings.

The new 02s perform much better now, still the cross counts are little too low.

The engnie runs too cold to my linking at 78 degrees. The thermostat might need attention or the coolant sensor is not good. At 85-90 degrees it runs so much smoother and better. Get some cardboard in front of the radiator and increase the fan points by 5 degrees. The fuel evaporates best even at 95-105 degrees coolant, you get a higher fuel milage and the engine is more efficient with all the shit from oil turning into vapors, but is too much for built lt1. The best compromise from performance and drivability is around 90 degrees.

I see you got higher blms everywhere on the right side meaning it is adding fuel off idle too. It does seems like exhaust leak to me. Try swapping the 02s and look very carefully for black sooth and for exhaust smell in the engine bayat cold stratup. Sometimes even the tiniest hole can bring much trouble and split blms.

The wideband also shows the lean condition so it is real.
If you suspect a cylinder s cyl cut test can point which ones is giving trouble. Set the afr to 12.5 and start cutting cylinders with 5-10 second delay between them. Record how much the wideband readings change when a cylinder is cut.

Did you tried to lower the rpm and the advance for more stable idle. Stock cubes cammed lt1 idles happy at 650 at Park and 550rpm at drive. I doubt the 383 can`t keep stable idle at 750-850rpm with 20-25 degrees advance. At lower rpms the surge from bad tune is much more pronounced.
How is the idle ovespeed/underspeed spark retard/advance tables set. Having 1-3 degrees there offload the iac motor jumping all the time.

I wasn't really intending for this thread to turn into a tuning workshop, but I appreciate the value of having your attention so what the heck. I want the tune to be perfect (eventually) but I want to know how to do it just a bit more.

These are the headers I have (link). Cats aren't currently installed, although I doubt they would do much of anything.

Quote:

---

Originally Posted by kur4o

... If the 02s are mounted too far from engine than stock, there is some delay for the exhaust gas to reach the sensor and it must be fixed in the bin CL settings.

---

I'm aware of the 26FB integrator delay / transport delay table, and until just recently I had the values almost doubled from stock. They're back to stock now.

I did a bit of "visual" measuring against the stock exhaust system before taking it to the scrapper for this purpose. I can't find any pictures at the moment but if I'm remembering correctly the left bank O2 bung wasn't considerably closer / further away from the exhaust valves than they are through the header primaries. On the right bank the O2 bung was several inches closer because they tucked the cat much closer to the exhaust manifold.

In the grand scheme of things I'm not sure there's a lot of difference in transport delay though, when you consider the velocity of gasses traveling through a 1-5/8 tube versus the huge log manifold and then into the 2-1/2" exhaust pipe. But I'm open to suggestions.

Quote:

---

Originally Posted by kur4o

The new 02s perform much better now, still the cross counts are little too low.

---

As mentioned the right bank has always been slightly lazy, but I'm starting to think this is a result of poor cylinder balance and improper injector characterization. Not to mention the old regulator producing higher than intended rail pressure.

To update the current state of affairs - since flashing the most recent prime pulsewidth tables using values from the correct conversion formulas, starts have been good so far. I suspect my old fuel pressure regulator has had me chasing my tail trying to find the injector slope "sweet spot" when in fact it was a moving target. I've been decrementing the 12B4C injector constant by 0x0001 with every flash and will continue until I see slightly rich OL numbers on the wideband. I suspect this will make the cross counts increase on both banks.

Quote:

---

Originally Posted by kur4o

The engnie runs too cold to my linking at 78 degrees. The thermostat might need attention or the coolant sensor is not good.

---

I've had a 160F thermostat installed for several years. When I first bought the car it had numerous cooling system issues from all the dexcool sludge that had accumulated, so I put this in out of an abundance of caution. I see less need for that extra insurance now that I know the cooling system is 100% healthy. I understand the pros and cons of the lower temps, and I think I'll leave it alone until I'm happy with the state of the tune so a before / after comparison can be made when I install a OE spec 180F t-stat.

Quote:

---

Originally Posted by kur4o

I see you got higher blms everywhere on the right side

---

Sorry, I'm not sure I'm seeing that (everywhere). Looking at the big log from earlier today stacked with the one attached from this afternoon I'm only seeing averaged BLMs on the right that are higher than 129 in cells 2 and 16-18.

Edit: I see what you're saying though - the right bank is on average lean of the right.

Quote:

---

Originally Posted by kur4o

If you suspect a cylinder s cyl cut test can point which ones is giving trouble. Set the afr to 12.5 and start cutting cylinders with 5-10 second delay between them. Record how much the wideband readings change when a cylinder is cut.

---

That's what I've been doing minus dropping the AFR to 12.5. Nothing has really "jumped out" so far. I'll try to get a fresh balance test tomorrow.

Quote:

---

Originally Posted by kur4o

Did you tried to lower the rpm and the advance for more stable idle. Stock cubes cammed lt1 idles happy at 650 at Park and 550rpm at drive. I doubt the 383 can`t keep stable idle at 750-850rpm with 20-25 degrees advance. At lower rpms the surge from bad tune is much more pronounced.
How is the idle ovespeed/underspeed spark retard/advance tables set. Having 1-3 degrees there offload the iac motor jumping all the time.

---

I've never been happy with the idle quality below 29 degrees spark or 900 rpm - I like to hear the cam, but I don' t like to feel it. I'm sure some of this is a function of cylinder balance. So hopefully this is something that will improve as I find the right individual cylinder trims. Unfortunately MBT is happening around 29 degrees so there's not much room for the underspeed idle spark control to do it's intended job.

Thanks to everyone for all your help and suggestions. I feel pretty confident now that all the problems encountered were not a direct result of the experimental ignition system, and that's what I was most concerned with.

Here is a starter pic of the intake manifold for reference for individual cyl trims. You can clearly see why in the stock applications #7 is always set at 0.9.
WIth long tubes with different lenght primaries there is considerable difference in cylinder efficiency. The longer the tube the more air it draws. SO the difference will be how much readily available air is left for cylinder to draw from the intake passage and how much vacuum is made in the exhaust tube.

If you haven`t equalize the stock intake passage holes, you can start with the stock cyl trims settings and work from there. The idea is to find out which cylinders draw more air and which ones` less, Than adjust the fueling.

I always wondered why the CL worked for you at idle even at that cam and didn`t experience the dreaded enleanment at idle CL. When you mentioned that the 02s were down the pipe, I checked the bin settings for your calibration and it is seems like it is preset for some headers already. I will test them as soon as possible and will report how it does with midlenght slp headers.
You can still make some tweeks to get more crosscounts at off idle cruising.

Quote:

---

MBT is ideal at wide-open throttle (WOT), but not desirable when the engine is at idle.

---

per wikipedia.
Too much heat in exchange for really lean mixture to utilize the extra spark advance. Maybe that`s why you are getting that unburned fuel smell. The afr is too fat for the advance and the fuel can`t burn and it is dumped in the exhaust.

I agree and terminal_crazy also reported similar results.
I have also zeroed the underspeed tables and the result was a disaster in idle quality. So I increased it to a max of +-2 degrees. It really smooths and offload the IAC valve from constantly adjusting.

The only reason I see for idle rpm increase is if the engine stalls at take off or it gets unstable at near idle conditions and is trying to die. Also the lower map readings you get are because the higher rpms. I am sure that the map will jump around 45-50kp area if you lower the rpms by 100. It takes alot of tuning to get it right with no surging but the results is clean and efficient combustion.

I also have 160* thermostat but never have seen under 80-81 C degrees cruising even at colder weather. I can suggest to leave the thermostat and increase the fan points. At colder weather block part of the radiator. Just for sake of experiments try the idle quality at 90*C and see how it calms and quiets down. I have never run it hotter but can imagine stock at 100*C. There is one very nice settings for fan temp with mph threshold. You can set higher temps with low speed cruising and colder settings with high speed performance trips.

Sometimes it is better to tune by feel and sound. The preffered method is also seat in the pants tuning.

Quote:

---

Originally Posted by kur4o

When you mentioned that the 02s were down the pipe, I checked the bin settings for your calibration and it is seems like it is preset for some headers already.

---

Am I correct to interpret that to mean the O2s are in the manifolds on other LT1s? That would explain a lot of the difficulties f-body guys have with headers.

Quote:

---

Originally Posted by kur4o

per wikipedia.
Too much heat in exchange for really lean mixture to utilize the extra spark advance. Maybe that`s why you are getting that unburned fuel smell. The afr is too fat for the advance and the fuel can`t burn and it is dumped in the exhaust.

---

I would suggest you search "spark hook test". Maximum Brake Torque is the spark advance where, with all other variables removed (speed and AFR) the most torque is generated, and consequently the lowest resultant manifold pressure. Adding more spark advance beyond MBT for a given mixture yields no additional torque. This is why you want spark at MBT everywhere except at idle, because the underspeed adder can't increase engine speed if you're already at MBT.

Quote:

---

Originally Posted by kur4o

The only reason I see for idle rpm increase is if the engine stalls at take off or it gets unstable at near idle conditions and is trying to die.

---

I've never attempted to get it to idle below 875. The cam lope shakes the entire car and I'm not a fan of it. If I can find any improvements with cylinder balance I may attempt to work on lowering idle speed and spark.

Quote:

---

Originally Posted by kur4o

Also the lower map readings you get are because the higher rpms. I am sure that the map will jump around 45-50kp area if you lower the rpms by 100.

---

I was under the impression lower map at idle was the goal, not higher. ???

Quote:

---

Originally Posted by kur4o

It takes alot of tuning to get it right with no surging but the results is clean and efficient combustion.

---

When you write a book I'll be the first in line to purchase it.

Quote:

---

Originally Posted by kur4o

cut test can point which ones is giving trouble. Set the afr to 12.5 and start cutting cylinders

---

Is this something that's only possible with your v4 patches? The fork of eehack I'm using has the cylinder disable button group greyed out when AFR is overridden, but when I commented out the line that greys them out and tested it seems like the AFR reverts back to the tune default when a cylinder is disabled.

As a workaround I changed the OL AFR target to 13.2 in all the cells I thought would be affected. I should have gone one temperature band higher, the result was 13.4-13.6 during the balance test. In the attached log it starts at timestamp 4410 and I disabled the cylinders in numerical order.

Attachment 14667

If you watch the narrowbands, it looks like the right bank is robbing fuel from #3 pretty efficiently as both sides go flat when it's injector is disabled. I'm not sure what this means (if anything).

Other than that I don't see any really obvious red flags except that MAP took a nosedive after #6 had been disabled for a few seconds.

At the end I played with idle speed and spark and managed to get a reasonable idle at 800 with 25-26 degrees advance (albeit at 13.4:1 OL afr). But then I let it revert to closed loop and couldn't tell much difference. I guess I need to try these settings out in a tune and see how it feels as it's warming up.

The attached log is about 80 minutes of mixed driving in closed loop until I got in the driveway and played with cylinder balance and then idle. If you look at the CL Performance tab on analysis in eehack note that my two primary cruising cells (2 and 6) seem to trade places - in cell 2 the right bank is lean, and in cell 6 the left bank turns lean. The BLMs almost trade places.

On an unrelated subject, I think I came up with a semi-reasonable method for working out the cranking prime pulse tables. I need to clean up my worksheet and write up a how-to for you to look at and see what you think. What I did is took the actual pulsewidths after the floats have been converted to hex and calculated the sum of the actual fuel charge for the first two pulses. Then I adjust one or the other until the summed fuel charge most closely matches the desired sum. It seems to be producing decent results. I've seen folks mention running 80lb Ford Racing injectors in this application, and I have to wonder how they've managed to get crank fueling dialed in with no more resolution than is available here.