This is more a heads up than anything. As far as I can recall, the most in depth guidance I've encountered regarding tuning a fuel injector size change across the forums has been limited to setting a base injector pulse width arrived at by calculation or trial and error for an acceptable idle for a start.
Not much is said about injector offset vs. pump or battery voltage. In my case running 8F with 60 lb/hr injectors about 12 psi above their rating, I believe I have discovered the cause of my persistent rich idle or at least part of the cause.
I also setup a history table in Tunerpro5 to follow pump voltage and have recorded voltage variation in excess of 1 volt and nearly as much as 2.
After looking at the offset table in my tune which for the most part shows values for the stock 23ish lb/hr injectors, I got to searching until I found the spec sheet for my DEKA injectors, noted the differences and realized that this over sight contributes to the difficulty in establishing a stable idle after such a change.
Whenever an accessory is turned on causing a voltage change it might affect the fuel delivery. When the engine/alternator warms up the voltage changes also.
I'll be making one important change for a more stable fuel pump voltage to go along with intended value changes, bypassing the stock pump wiring in favor of a "Hot wire" arrangement with direct battery feed to the relay and then to the pump for better current and voltage delivery.
At the moment I can hear my fuel pump oscillate with the turn signal on. I'll be looking at an alternator replacement also as it seems to drop off on occasion.
I found this link on the forum addressing EGR and fuel delivery that includes the formula for calculating the base pulse constant;
I put the theory to the test this morning and sure enough there was an immediate effect on idle AFR after adjusting the pump voltage vs injector offset. I also adjusted the base pulse constant (BPC) down one point.
I adjusted the idle VE table but the response to the changes were poor until an excessive value change was made and even with that it did not reflect as well in the AFR reading but moreso in the idle quality.
I stumbled across another table while confirming that certain injector related tables; lbs/hr, gal/hr, single inject, etc were not effective and therefore inactive in my tune; fuel injector offset vs base pulse width.
When I lowered the value corresponding to my idle rpm using the cell follower, the engine idle started swinging giving a rich/lean reading on the AFR display eventhough the chip is set for openloop operation and no correction should be made though I suspect engine stall prevention logic must have been the cause (When I adjusted the voltage table to low the engine just stalled out). The surge related rich/lean operation was characteristic of what is experienced and complained about by many when running closedloop operation after injector size is increased, while cruising along in closed loop you can feel the rich/lean crosses as well as see them on the AFR gauge.
The problem is figuring out which table to change to correct the problem which no one seems to have done definitively but looking at the values in my table and given the response to my temporary change it appears the BPW offset table is the possible solution.
When you consider the code mask many of us are running was calibrated for injectors that are as much as 20 yrs old and some of us are using newer more efficient injectors, chances are pretty good these tables are even more significant to a stable tune than first thought and should probably be the first tables adjusted before any serious VE correction is made except where it's essential.
That to me seems to be a valid theory , as you have explained . There are a few guys on other forums experiencing so of the same after injector changes or after brining the care out after storage or weather getting colder and using accessories more IE: heater / defroster Separating the fuel pump circuit and installing a larger output Alt and upgrading you wiring and grounds . Should make a larger difference in voltage drop and more consistant pulsewidth over a greater range od temps and load. One other thing that not one talks about which may help is shielding all of the injector wiring , making sure there are separte twisted pairs and not allowed to touch any metal at all (i know its extreme but we are talking milivolts here) to keep any other signal out all together . Great write up Joe ,Although i know nothing about OBD1 tuning others here maybe able help alot more.
Last edited by unchained01; 12-20-2011 at 10:06 AM.
This is common in OBD2 and aftermarket computers. Doesn't work the same at all for OBD1 unfortunately.
Agreed, but every little bit helps. The trip home today was much more like what I hoped to see in the datalog, the average AFR was 14.xxx all over the history table instead of 13-16 and that's in openloop so the fuel map is much closer than what it has been or what I was able to achieve consistently before the changes made this morning. There seems to be a point where the VE table just doesn't have any effect or seems that way with changes beyond a certain value level.
The two often ignored tables made a tremendous difference. I'm not spending much time trying to perfect it now because I have a small exhaust leak ahead of the narrow band O2 sensor. My wideband is about 10" away from the turbo and that's what I used to make adjustments.
At some point I'll sit down and attempt to enter the spec sheet values and proper base constant value after I install the adjustable fuel pressure regulator and reduce fuel pressure to the spec rating. My alternator is definately part of the problem, the ride home was in warmer weather and during a hot idle log battery voltage dropped to about 12.3 with the coolant fan, park lights and heater on, otherwise it hoovers just over 13 volts.
I would not lower the pressure to 39psi on a turbo car and I would not consider changing voltage tables until you have a real stable voltage.
You wouldn't lower the pressure to spec for the injectors even with a 1:1 boost sensing fuel pressure regulator? My preference is higher static fuel pressure than the current ~50 psi for good atomization but I'm not sure the relatively higher pressure is not contributing to idle AFR inconsistency. I don't favor the idea but I want a proper idle among other things and it seems the best approach is being maticulus with the numbers.
The stock injectors for the 3900 are about 28 lb/hr and the DEKA 60s about 62 lb/hr at 42 psi so I'm already above that with current static pressure readings and uncertainty as to how well the 730 ecm is managing them at low pulse width.
My current base pulse width at idle is about .119
I understand the concern about holding off on injector offset changes until the voltage is more stable but I don't expect it to cause much of a problem since my adjustments have been restricted to the low end reading where idle sits to clean up the AFR and eliminate the rich fuel smell.
Is there something I've forgotten to factor in?
Last edited by Joseph Upson; 12-20-2011 at 05:57 PM.
Im going to use 2 "Peak-n-hold" driver cards and some 70lb Low-z injectors. They can be used on any ecm that controls the injector by switching negative. http://www.jbperf.com/p&h_board/v2_0.html
It should also help clean up the RPM signal since injector coil "flyback" voltage is kept out of the Megasquirt altogether. Im hoping (more like praying) that I can get way with 14-15 afr's while cruising.
95" camaro with turbo/ intercooled 99" Grand-am full motor swap & custom mounts w/ TKO600 5-speed. "update!" Soon to be "bullet proof" 3500 LX9 swap
Im going to use 2 "Peak-n-hold" driver cards and some 70lb Low-z injectors. They can be used on any ecm that controls the injector by switching negative. http://www.jbperf.com/p&h_board/v2_0.html
It should also help clean up the RPM signal since injector coil "flyback" voltage is kept out of the Megasquirt altogether. Im hoping (more like praying) that I can get way with 14-15 afr's while cruising.
It's much wiser to use no more fuel injector than what's needed especially in a daily driver. The side effects/ trouble with tuning higher capacity injectors is not worth it until all the necessary parameters needing adjustment can be identified. I didn't have an idle problem until I installed the 60s I noticed from looking at video of the car with slightly larger injectors installed.
It's much wiser to use no more fuel injector than what's needed especially in a daily driver. The side effects/ trouble with tuning higher capacity injectors is not worth it until all the necessary parameters needing adjustment can be identified. I didn't have an idle problem until I installed the 60s I noticed from looking at video of the car with slightly larger injectors installed.
My turbo flows a little over 72lbs of air/ min. A set of 70lb injectors will support a little over 630-650hp. I need the extra fuel volume, so smaller fuel injectors aren't a possibility. Low-Z injectors allow for faster opening and closing rates than a standard saturated injector. (about .5ms faster)
95" camaro with turbo/ intercooled 99" Grand-am full motor swap & custom mounts w/ TKO600 5-speed. "update!" Soon to be "bullet proof" 3500 LX9 swap
Originally Posted by SappySE107
Posting Permissions
Reply With Quote
