Tuning from scratch

[brian617]

Cylinder Volume.
Injector offsets.
Returnless VS Vacuum Ref. (Flow rate vs KPA)

[Jim Blackwood]

Please see if these figures look like they are within reason. I only changed the cylinder volume.

Cylinder volume is 712cc
EOIT Boundary: 585
EOIT: 230-500
Inj PW voltage adj: .06-9.68
Inj voltage correction: 1
Min PW: 1.277
small pulse threshold: 3.997

Couldn't find the returnless/vac reference but my fuel system uses a return and is referenced to MAP.

Thanks,

Jim

[kur4o]

The way you set the inj flow rate, requires constant 60psi fuel pressure.
Step 1 to fix this, remove the map reference to the fuel pressure regulator.
or Make inj flow rate constant across all ranges and leave the map reference.

I am not quite sure how the ve works in ls1 but it must have some max value that will be close to 100% ve. Looking at the values at the right bottom of the ve table will give you the max value the pcm can use. If you max the ve table the only way to tune will be increase the cylinder volume scalar and than rework the ve table.

Dont forget that all of this ve calculations are represented as airflow to pcm in g/s. Monitoring that value will give you a clue if you manage to change the values or they are capped.

Don`t tune warm up, there are some other adders for that purpose. Tuning ve table must be done on fully warmed engine with all the others turned off.

[brian617]

Couldn't find the returnless/vac reference but my fuel system uses a return and is referenced to MAP.

Thanks,

Jim

"Flow rate VS KPA" is how its listed in HP Tuners, not sure what it will be named in your XDF. Also VE table should be a whole number from 0-100 for instance the file I'm looking at ranges from 17-97

What injectors are you using?

[Jim Blackwood]

At the risk of sounding obtuse, how do you ever tune the ve table if you can't idle the engine? Seems to be a necessary first step for getting it up to operating temperature, I don't see an easy way around that. Any suggestions for initial cold enrichment settings and where to make them? I'm all ears, I'd love to hear your suggestions and will try them out straight away.

My injector flow rate/kPa chart only varies by .22 gm/s but I have changed it to 4.4 across all values. I can't see that making much difference, certainly not enough to fix my lean condition but that's done.

My FPR has a hose that connects to the intake plenum so that it maintains 40psi in reference to the intake regardless of vacuum or boost conditions. There is no other modifier to the fuel pressure. This is consistent with most return-type fueling systems and is completely independent of the engine controller.

The max entered value in my VE table is 2.3980 which is at 4800 rpm and 170 kPa. At 100 it is 2.3227@4400 rpm. I am also accustomed to using values from 1-100 but that doesn't appear to be used here so I'm trying to adapt. As a multiplier, 1 would equal unity in the formula, (does not change it) which also uses the injector flow rate and other factors. For my engine to require at least a 3x multiplier in order to idle suggests that there is an unbalance elsewhere in the formula. Not the VE table, not the injector flow rate. Some other factor has to be off by and be dividing the formula by approximately 3 or more. I need to find out and fix that.

Does anyone have the actual fuel calculation formula?

Jim

[brian617]

My injector flow rate/kPa chart only varies by .22 gm/s but I have changed it to 4.4 across all values. I can't see that making much difference, certainly not enough to fix my lean condition but that's done.

My FPR has a hose that connects to the intake plenum so that it maintains 40psi in reference to the intake regardless of vacuum or boost conditions. There is no other modifier to the fuel pressure. This is consistent with most return-type fueling systems and is completely independent of the engine controller.



Jim

With vacuum ref the "Flow rate vs KPA" should be a fixed number across the table which reflects the flow rate of your injector @ xx psi. I believe that's 45psi in the LS world? Some correct me if wrong.

Example, a 01 Silverado with a 5.3 engine has a value of 25.1 across the table.

[kur4o]

I suggest you run fixed fuel pressure at 60psi. It is also very likely that the fuel flow constant is higher than actual. The injectors might flow less than you thought.

Start lowering the values at the fuel flow table by 10% and see how it goes. When you remove the map reference the table should take that into account.

Your best bet is to start with some ls1 stock injectors with good stock data, and make initial tuning with them. That way a big unknown is removed from the equation.

The % of ve is derived by the ve table values and cylinder volume. If anyone knows the formula I can add it to the xdf. I will play with efi live to see if I can rip the conversion they used.


Ls1 intake is with equal length runners and there is no need for individual cyl corrections. It could be some of the cylinders running leaner than others.


The xdf tables looks weird with 170kpa tables. I wouldn`t trust them too much. You have 1 bar map that can`t read 170kpa or there is some scaling with 2 bar map used.

A complete write down of your setup might be needed with the origin of the xdf and bin.

[Jim Blackwood]

The injector flow rate of the stock '01 Silverado is 3.18 gm/sec. I flowed my injectors at 35lb/hr (4.4gm/sec) @ 40psi. Actual was 37.7 and below saturation flow values were as expected. I could jack up the fuel pressure but that seems like an artificial solution, especially when it was running fine with that fuel pressure and comparable settings on the other controller. (MS2)

If anyone can find out the actual formula it should be possible to pinpoint the problem using that. Until then I guess I can fudge the thing by lowering the injector flow value to offset whatever is causing the problem. It's funny in a way, that was my first impulse. Going full circle here so maybe that's a sign.

About that MAP value, yes I am using a 2 bar sensor. I changed the multiplier to account for it and rescaled most of the tables that use MAP. May have missed one here or there but I don't see why that shouldn't work just fine.

Logic would seem to suggest that if the VE has to be tripled to get enough fuel, then the Injector flow number should be divided by 3 to achieve a comparable result. That seems like a lot but it won't hurt me to try it.

Jim

[brian617]

According to this page https://www.psiconversion.com/tech/l..._injector.html LM7 injectors basic 5.3 injectors flow 24.8 lbs per hour @ 58 psi. That's why the "Flow rate vs KPA" table has a value of 25.1 across the board. Again we don't know what injector you are using but an LM7 injector flowing 24.8lbs/hr would be 3.12grams/sec. That table has to reflect the flow rate of your injector at 58psi.

So your injector 35lbs/hr @ 40psi would be 42lbs/hr @ 58psi and the table value should be 42

However if you aren't running 58psi with the vacuum hose off the reg, then your'e going to have a lot of injector data to fudge.

[kur4o]

That 2bar map and scaling complicates things further. There is tons of map referenced tables that are not defined in the xdf.

I am sure your inj flow tables is off the chart. Vacuum is derived from map and baro, so it is unclear what values are you getting. It is too bad that it might end with tail chasing tuning. The ve table also is capped well above 100%. Try increasing substantially cylinder volume constant and play with inj flow rate. You might find the sweet spot at the end.

With boost 40psi FP is well too low. I would run constant 60psi if the injectors can handle it.

[kur4o]

Someone good at math can figure the conversion of ve to % here are some raw values in hex converted to % according to efi program.

$a8e7 ve with $5b23 cyl volume constant equals to 343.54% ve
$7a67 ve with $5b23 cyl volume =266.24%

[Jim Blackwood]

Probably the biggest clue to what is going on here is that this is no small discrepancy. That was demonstrated by today's testing. Unlike possible or at least theoretical limitations in the VE table values that may have prevented efforts to find the range, there is no limitation in reducing the given injector flow rate. Of necessity the flow rate has to be a divisor. The larger the number, the shorter the resulting injector pulse. We know this Empirically because a larger injector will deliver the same volume of fuel with a shorter pulse. Therefore as the number approaches zero the pulse width approaches infinity.

To get to the point, there was one entire order of magnitude (x10) between the flow setting required to create an injector pulse long enough to run stoich at idle, and the beginning injector flow rate in grams per second. The ending test rate was 0.46 g/s and the starting point was 4.4 g/s.

What this means is that for some reason there is a 10x multiplier that has dropped out of the fuel formula. I'm honestly having a bit of difficulty figuring out what that could be. Not fuel pressure, I'd have to be running 400 psi to make that much difference. (Testable by checking the injector PW with a scope but just the magnitude says that isn't it, to say nothing of the fact that this engine ran fine on those injectors with that pressure before the controllers were changed.) Differences in MAP values would never be on that size of a scale either so pointing at the 2 bar MAP sensor isn't productive either. The "some here some there" theory MIGHT have some promise, but for 2 errors we'd have to see at the very least a threefold (x3) error IN EACH ONE to even have a chance to make that theory work. Fuel pressure x MAP still isn't anywhere close to enough. Adders? Nope, we have to have a multiplier here to explain it. There is no adder that brings in that much extra fuel. Not for a street driven engine.

All I can think of that's left is cylinder volume, maybe I'll play with that tomorrow. But it seems like there has to be another factor somewhere that is missing. A small table that's zero'd out or something. That seems like the most probable cause.

Jim

[kur4o]

I am almost sure the issue is within the 2bar map and the scaling. There is too many variables that are not accounted for. Not a single small table.
Something must be capped in the equation and than a default built in value is used.
Anyway you will never get good enough fuel flow data since vacuum is converted by ??? and the result is used at the table lookup. I am not sure but there might be a switch what to use for the table lookup map or vacuum.

Start increasing the cylinder volume with correct inj data and see how it goes.

1000% in fuel flow equals to huge airflow error. You are missing 1 decimal point in the airflow. And that airflow is used everywhere for running the engine including spark.

[Jim Blackwood]

OK, I ordered a new 1 bar sensor. When it comes in I will try it like that. As long as I stay out of boost it should work just like if it wasn't supercharged. Then if I get that working right will be soon enough to worry about the 2 bar. Anyway at this point I really hope you are right. As likely to be that as anything else.

Jim

[Jim Blackwood]

You know, I might have an idea what it could be. There is a scalar for the MAP sensor. (C6301) When I modded the tune I changed the scalar. I doubled the value. But if I went the wrong way that could easily be a fourfold difference in the resulting value. It doesn't seem like enough, but it sure could be a big part of it. Then too, the Camaro tune I got from Scott had a value of 94 where the Silverado had 73 which seems like a significant difference. I used a value of 188. I think I'll set it to 35 and see what happens. I also changed the sensor offset from 10 to 5. I'm not sure what effect that will have but if doubling the two values is bad maybe halving them will be good.

Jim