Split blms

[FTA1995]

Having some issues with fueling, fixed all the stuff that needed it, now im getting off idle blm splits

I could do the following;
forced open loop by raising open loop thresholds
speed density mode
keep it closed loop maf mode and figure it out painstakingly

I have done a bit of homework here, searched around and tried a few tricks with no results.
Everything mechanical and electrical is in good working order, as far as cl feedback goes.


suggestions..? I have not figured this one out yet, don't have a infrared thermometer for individual trim multipliers. As of right now, i just put the thresholds for cl way up to a temperature range i wont hit.

Thanks again to my favorite f body board

[Joe 1320]

It would be interesting to see individual cylinder exhaust temperatures. That's an expensive proposition though. For what it's worth, I was chasing that on my 97 WS6, I never totally fixed the problem.

[Craig 94 TA GT]

If you could measure individual EGTs or even have individual O2s for each primary, I would be the first one to request to know!

I've been fighting split BLMs for quite a bit of time. 160/128 at 16 and about 10 off for all other cells. Nothing has ever corrected it. I always assumed it was my driver's side header leaking, but it may not be...especially if you're getting it with long tubes - I assume the O2 bung is before the collector joint to y-pipe?

What throttle body are running? I heard that the AS&Ms are notorious for this due to the way IAC passages are. I tried adjusting the stopper screw in mine and got it to around 151/12x at cell 16.

[FTA1995]

running a ported stock tb, thats probably related, but i think it would be more affected at idle with poorly seated butterfly valves etc.
yea Joe, the egt's are a little to pricy at risk of understatement....
I'm going to try and get my hands on an infrared thermometer instead of using individual lamda's thats just tooo costly.
for now, it seems that keeping it in open loop and just running off the ve tables is working pretty well, plus makes tuning way easier since the block learns dont get in the way.

Well... back to the lab again, thanks guys if you think of anything let me know!

[Injuneer]

What do the long terms look like? How bad is the split? What makes you think it is a problem with individual cylinders?

[Joe 1320]

What makes you think it is a problem with individual cylinders?

Not sure who the question was for, my thought is that the computer averages each bank for fuel control rather than each individual cylinder. I've seen instances where one or two cylinders out of whack affect the blm as a bank average, effectively making the cylinders that were initially correct now out of tune. It's hard to pinpoint those problems without a long process of making runs, pulling and reading plugs ad nausium. Reading exhaust gas temps would be perfect, but each cylinder would need to be read seperate and putting 8 individual bungs in a set of headers is not within the scope of the average owner or tuner.

[Injuneer]

My comments were directed to the original post. He has a split BLM problem, and I was curious how bad it is, and why he had already talked about individual trim multipliers.

[FTA1995]

short terms off idle are about 126 and 135 sometimes, it really seems like its random and not consistent with any environmental variables, and at idle its usually plus or minus 1 from 128. I am going to auto zone to grab some oil for a quick change and warm up, i'll pull the data and see where those long terms are at, last time I was on the datalogger was not enough time for the long term trims to come into effect. When I did my intake a bit ago, fuel system off, a buddy of mine helped clean out/off the fuel rails and injectors, I don't think the injectors were put back where they came off, that would definitely be a reason for the split.

Known to be in good working order:
O2's
Opti
All ignition system components
fuel pressure and volume
MAF sensor
MAP sensor
fuel pressure regulator hold at spec and does not have a ruptured diaphragm


Today I will check the following:
Injector impedance
BIN file for tune to make sure there were no errors
TB position sensor (*might* cause a blm issue i've read)
If no problems are found, I'm going to disconnect all vacuum powered components and stop the ports up (ex. FPR for consistency)



ugh... damn cars... more info later...

[Injuneer]

Short terms are "random" by intent. They are dithered on either side of 128 to cause the engine to swing back and forth from rich to lean. That's required for the cat(s) to function. They have to average 128 over a long period of time. If they don't average 128, the PCM will adjust the long terms up or down to pull them back down to an average of 128. That's how the long terms are developed ("learned").

Why would switching the positions of the injectors cause a "split"? If they did, it would be very small, assuming the injectors are healthy. While they might each have slight variations in flow vs. pressure, those variations will be random, and there's just as much chance you reduced the bank-to-bank dofferences as there are that you increased them.

Removing the vacuum compensation line from the fuel pressure regulator will hurt consistancy. The driving force on the injectors is the DIFFERENTIAL pressure between the rail and the intake manifold. That is held constant by the vacuum compensation line. Remove the vacuum compensation, and the rail pressure will be constant, but the DIFFERENTIAL pressure will vary all over the place, based on manifold vacuum. That will not help you in any way to analyze the problem.

[FTA1995]

Why would switching the positions of the injectors cause a "split"? If they did, it would be very small, assuming the injectors are healthy. While they might each have slight variations in flow vs. pressure, those variations will be random, and there's just as much chance you reduced the bank-to-bank dofferences as there are that you increased them.

Removing the vacuum compensation line from the fuel pressure regulator will hurt consistancy. The driving force on the injectors is the DIFFERENTIAL pressure between the rail and the intake manifold. That is held constant by the vacuum compensation line. Remove the vacuum compensation, and the rail pressure will be constant, but the DIFFERENTIAL pressure will vary all over the place, based on manifold vacuum. That will not help you in any way to analyze the problem.

Right, it's the relative positioning problem here, as in the injectors were not in their original positions and i have no idea which one is where, hence the injector pulsewidth multiplier or whatever its called calibration is now bunk.

And yes, I'm well aware that removing the vacuum from the reg will just peak pressure constantly behind the injectors, i was talking moreover checking all the vacuum harness's for leaks/fuel contamination.

Actually about to go check this stuff out right now, finally got some free time. I've been fiddling with the program and I seem to have gotten the thing to smooth out, if not run better in the right conditions before putting it in cl lockout. but, since I live in chicago where the weather never stops changing for one second, I really need to shoot for closed loop for the drivability in varying conditions.

[Injuneer]

Are you suggesting that the factory tweeks the individual cylinders based on the minor flow variations between individual injectors? I don't think that would be done in a mass production engine assembly line.

I have the ability to tune individual cylinders by A/F ratio, and for spark and injector timing (start or end of event relative to TDC), and even spraying a 300-shot of nitrous on a 500 HP NA setup, we could not find enough variation from cylinder to cylinder to make any trim adjustments. That was based on spark plug examination, under a 10X viewer, by someone who spent the last 20 years tuning dry nitrous systems (not me).

The only time I've seen that much effort put into tuning individual cylinders was an a 20+++ # boosted 383, making 1,350HP. For tuning purposes, each primary was fitted with an EGT. To avoid the common SBC problem of high temps in the siamesed exhausts (3-5; 4-6), due to the lack of cooling between them, the engine had a coolant line run directly to a nipple added to the block between the pairs of cylinders, to eliminate cylinder-to-cylinder variations resulting from coolant flow. That engine had adjustments to A/F trims and timing by individual cylinder.

In the SAE paper on the design of the LT1 engine, there is a chart showing the variation in cylinder wall temperatures, based on the position in the cooling path, and I would suspect that the factory cylinder-to-cylinder tweeks would be based on those temperatures.

[FTA1995]

Are you suggesting that the factory tweeks the individual cylinders based on the minor flow variations between individual injectors? I don't think that would be done in a mass production engine assembly line.

The owner before me did it since the injector batch he picked up wasn't the best, and I haven't had the right opportunity. I suppose I could turn all the multipliers back to 1 to see what happens?
Checked out what I could today, everything is within spec.
Just going to fiddle with oen loop tuning and a wideband maybe if I cant get the splits down with anyting on the tune...
Ideas at all?

[FTA1995]

After a great deal of logging and tweaking, I was able to get the system to hit a long term trim of 123 on one side and 126 on the other. It would seem that the timing was largely responsible, so I had to tighten up what I had previously to more appropriate values, basically all the lower midrange advance cells, mostly between 40 and 75 for a map value, between roughly 1000 - 3200 rpm.
I still think messing with the individual injector multipliers might fix the problem; when I 'zeroed' them at 1.00, the split changed what seems like just enough to show significance towards some type of system as mentioned before to get, most likely an infrared thermometer will do fine, much cheaper than an individual EGT sensor setup.

Fred; I can't seem to find the SAE document on the LT1 for some reason, at least one with the diagram you're referencing involving the cooling dynamics, do you have a link to the whole thing? I'd like to read through it.