|Username||Post: EFI Idle Tuning Notes|
11-26-10 03:17 PM - Post#2009820
I recently spent several days reading (books) about EFI idle tuning and improving my engine's idle quality. I finally achieved, what I consider, an excellent idling engine...but it wasn't easy. In fact, I believe the Holley Hp/Dominator ECU is more sensitive to idle tuning (idle algorithm), than the Commander 950 is; meaning it's easier to achieve a good idle with the C950. I tried to disregard what I thought I knew about idle tuning and start with a fresh mindset...I'm glad I did. Every engine will idle slightly different and the engine combination (components) will affect the idling characteristics, especially the camshaft. These ECU parameters influence the idle quality:
Idle Speed RPM
Increasing the idle RPM provides a smoother idle. This doesn't mean your engine has to idle at 1000 RPM, however, don't expect a race engine with a radical camshaft to idle at 700 RPM. Generally, the bigger the camshaft the more idle RPM it needs. Most mild performance engines should idle well at 700-800 RPM and most potent street performance engines should idle well at 800-900 RPM. Serious race engines with radical camshafts will need to idle at 900-1000+ RPM, due to the camshaft's tight LSA which causes a high overlap period (not efficient at idle & low RPMs).
Target Air/Fuel Ratio
Most high performance engines will idle well with a target air/fuel ratio between 13.0:1 and 13.5:1. This surprised me and I wasn't comfortable with it at first. However, when I tried it, my engine responded so favorably, especially when I shifted it in gear. This became one of the three most significant aspects of my idle quality improvement (the other two being timing advance and Fuel Table tuning). This isn't as detrimental to fuel economy as one might think, because at idle, there are less injection events (engine cycles) over time, than at higher RPMs. Remember to also check the idle in gear and allow the engine idle for at least a minute with each change.
RPM & kPa Axis Configuration
When an engine is idling, there's a 'target idle cell' on the Fuel & Spark Tables. It's important to identify this single cell, in order to properly configure the RPM (X) axis and kPa (Y) axis. Ideally, the target idle cell should be in the center of the RPM & kPa range the engine idles in. Some engines exhibit better idle stability than others, as seen on the Fuel & Spark Tables. The goal is to configure the idle area so the idle has equal range above & below the target value. Fuel & Spark Tables should have three RPM columns and three kPa rows; one column/row above & below the target column/row. The values inputed above & below the target, depend on how the engine idles and the resolution of the Fuel & Spark Tables (e.g. 16x16 vs. 31x31). Typically, the RPM axis (idle area) has increments of 100-150 RPM and the kPa axis (idle area) has increments of 2-8 kPa. Remember to check the idle in gear, to verify the kPa axis configuration above idle, is acceptable.
Base Fuel Table
Most people I help, don't know how to tune the idle area of the Fuel Table (C950 or HP/Dominator ECU). It has a big impact on the idle stability. The Fuel Table can never be flat, meaning you can't have the same numbers in the idle area. Unlike the Spark Table, which should be flat. Here's a sample Fuel Table idle tuning scenario. The bold numbers indicate idling with no load (13) & idling in gear and/or accessories on (16). The surrounding cells direct the engine idle back to the target idle cell (center). Study the diagonal pattern of the idle numbers; that's the secret to good Fuel Table idle tuning. Notice how the values increase with higher load & less RPM and decrease with lower load & more RPM:
The Fuel Table example above is a good one, however, it illustrates the idle segment of a C950 Fuel Table with large injectors (Avenger/HP/Dominator Fuel Tables use "lbs/hr" unit of measure). On C950 Fuel Tables, smaller numbers mean larger injectors and larger numbers mean smaller injectors. Therefore, if your idle values (numbers) are larger than what's illustrated above (typical injector sizing), then the number pattern should change in increments of two or maybe even three. Example:
Since the new Avenger/HP/Dominator ECUs self-tune, why does the idle area of the Base Fuel Table need to be adjusted? Because the Learn function can only self-tune to the Target Air/Fuel Ratio map and it does so very well, in a floating 3x3 cell (9 block) pattern. When the Learn table self-tunes, it's adjusting the entire idle area the same amount; just like it's supposed to. In other words, the Base Fuel Table is responsible for idle quality and the self-tuning is responsible for O2 compensation. So the idle area of the Base Fuel Table still needs to be "configured" to promote a smooth idle, for the same aforementioned reasons. When Holley creates the base maps, they can't anticipate the exact location where every engine will idle. The Avenger/HP/Dominator Fuel Tables use "lbs/hr" unit of measure, so the number incrementation will not "look" to be spread out as far as the C950. However, the idle tuning scenario is the same. I've found that increments of .25-.50 lbs/hr, work well in the idle area of Holley Avenger/HP/Dominator Fuel Tables. Example:
Closed Loop Compensation Â±
ECUs without learning capability will have a closed loop compensation limit of around Â±25% and must be tuned so the ECU is subtracting about 5% from the Fuel Table. You'll never notice the ECU subtracting fuel but you can notice it adding fuel. For initial tuning purposes, allow the maximum value in the idle area, due to coolant & air temperature modifiers. ECUs with learning capability will have a virtually infinite closed loop compensation limit; as much as Â±999%. This is possible because the Learn Table modifies the Base Fuel Table by adding its learned amount to the base fuel values, so the ECU doesn't need to spend time repeating the entire modification (O2 compensation) to the Base Fuel Table, every time it returns to idle (or any other area for that matter). In some stubborn cases, a better idle may be attained by limiting the amount of closed loop compensation %, added or subtracted from the idle area.
The Timing Table must be flat at idle, meaning the same value used in the entire idle area (especially if using Idle Spark Control). Most street performance engines will idle well with 15Â°-25Â° of timing advance. The exact amount depends heavily on the camshaft specifications. Generally, tighter LSA - lobe separation angles (overlap) and larger lobe duration figures, require more timing advance at idle. 106Â°-108Â° is considered a tight LSA (idle quality suffers with less idle vacuum), 108Â°-110Â° is moderate, 110Â°-112Â° is moderately wide and 112Â°-114Â° is wide (idle quality improves with more idle vacuum). One must pay close attention to how much timing advance is used at idle. Resist the urge to use too much; I've made this mistake in the past. Advancing the timing, offers better fuel efficiency and raises the idle speed, however, excessive timing creates an unstable idle speed due to the engine having too much torque at idle (idle spark control becomes ineffective). Retarded timing lowers the idle speed, decreases engine torque and increases the coolant temperature. Excessively retarded timing also causes the exhaust headers to glow red hot. Remember to check the idle in gear, to verify the amount used is also acceptable.
Idle Spark Control Tuning
The Idle Spark Control basically fine tunes the idle speed by manipulating the timing (quickly increasing & decreasing in accordance to RPM). The idle quality must be well tuned before adjusting these two parameters. It may help to datalog various P&D combinations (name the datalog by the two numbers to decipher them) and look for the straightest RPM line. This is because you can't watch the idle RPMs on the Data Monitor, since they change too fast and the tachometer usually isn't an accurate enough indicator of idle stability. I've found values of 20 (P term) and 40 (D term) are a good start on Holley Avenger/HP/Dominator ECUs.
P&D Definitions - Excerpt from Holley EFI manual:
â€¢ P Term â€“ The speed at which the ignition timing is varied to maintain target idle.
â€¢ D Term â€“ Eliminates the overshoot.
IAC Control Tuning
The PID terms must be fine tuned to each particular engine. Start with all three PID parameters at 10. Tune the Proportional first, the Derivative second, the Integral last, then fine tune the three together. Temporarily disable the Idle Spark Control while tuning the idle PID parameters and ensure the Fuel Map is well tuned in the idle area. It may help to datalog various PID combinations (name the datalog by the three numbers to decipher them) and look for the straightest RPM line. This is because you can't watch the idle RPMs on the Data Monitor, since they change too fast and the tachometer usually isn't an accurate enough indicator of idle stability. Of course, you don't have to datalog the number combinations that obviously make it idle worse. Typically, slower IAC control promotes better idle stability.
PID Definitions - Excerpt from the Holley EFI manual:
â€¢ P Term â€“ (Proportional) Speed/gain of the system when there is a large deviation in the target idle speed. Raising this value increases the speed at which the IAC moves in order to remove target idle speed error. If this value is too high for a specific application, the IAC position will oscillate (be out of control) and cause the engine speed to surge up and down. If the value is too low for a specific application, the IAC will be slow to react to quick changes in idle speed deviation. However, it is much better for this term to be conservatively slow, rather than too fast.
â€¢ I Term â€“ (Integral) speed of the system when the engine speed is near the target idle speed. Raising this value increases the speed at which the IAC moves when the engine speed is close to target idle speed. If this value is too high for a specific application, there may be â€œtoo much IAC activityâ€ around the target idle speed. If the value is too low for a specific application, the IAC will be slow to react to changes near target idle speed. However, it is much better for this term to be conservatively slow, rather than too fast.
â€¢ D Term â€“ (Derivative) Higher derivative terms reduce the tendency of idle speed overshoot. Smaller derivative terms slow down the IAC movement as target idle speed is approached.
IAC Counts or Position (%)
The throttle blades must be set so the engine is breathing sufficient air on its own (without excessive IAC intervention), for idling and starting. The IAC motor/valve is supposed to regulate the idle airflow to help maintain the idle speed. IAC motors have an IAC position range of either 0-255 counts or 0%-100% (each 1% is the equivalent of 2.55 counts). The throttle blades (idle speed screw) is often set to position the IAC motor at 10-20 counts or 10%-15% at hot idle. I've found this to be too much IAC control when it's closing (toward zero). I prefer to further open the throttle blades to achieve a IAC position of 5 counts or 2% at hot idle. This only allows a minute amount of airflow reduction, which always seems to promote a more steady idle speed and improve starting.
Injector End Angle
I also spent some time tuning the injector end angle timing. I like the idea of injecting fuel a little earlier during the intake stroke. ECUs have different end angle software parameters, so there's an excerpt of how Holley does it below. The end angle has a diminishing effect as the injector duty cycle reaches 90%, however, it's nice to tune this for optimum air/fuel ratio balance at idle and in the lower RPM ranges. On my engine, -10Â° injector end angle equalized the left & right bank AFR. I believe most mild performance street engines (with Holley EFI), will benefit from -20Â° to 0Â° injector end angle. Torque, economy, emissions and idle quality are all effected by the injector timing. The optimum value depends on the engine RPM and load. At high duty cycles (RPMs), the injector timing will have minimal influence, since the injectors are on for most of the engine cycle.
TIP: I found that watching the left & right AFR, by staring at the data monitor was difficult to track (once I got close to the optimum value). Instead, I narrowed it down to three possibilities, then datalogged each change (named each datalog "AFR Test -xxÂ°"), so I could "see" the definitive AFR variance. I hope this provides an unambiguous method to help someone tune their injector end angle. Unfortunately, this requires dual wideband O2 sensors or one "ECU integrated" WBO2 sensor and one "stand-alone" aftermarket WBO2 sensor.
Injection End Angle - Excerpt from the Holley EFI manual:
This is the crankshaft angle in degrees when the injectors will finish their injection event. A value of 0Â° will end the injection event at BDC (Bottom Dead Center) before the compression stroke. A negative value moves the event before BDC and a positive value moves the event after BDC. There could be some improvements in cylinder to cylinder fuel distribution by tuning this parameter. If modifying, it is best to move this value in a negative direction to start.
It's important to use the proper size fuel injectors. Injectors that are too large for your application, won't idle well due to low pulse-width operation. Also, don't decrease fuel pressure in order to increase pulse width times at idle. Decreasing fuel pressure results in less than optimum injector spray patterns. Holley suggests the pulse-width at idle shouldn't be lower than 1.7 msec (which won't be an issue when using the appropriate size injectors) and that it shouldn't fluctuate more than .3 msec.
Injector Voltage Offset
All injectors have a voltage off-time/voltage curve and this needs to be correctly entered into the ECU's software. Sometimes this data can be difficult to find and contacting the manufacturer is the only way to acquire it. The injector off-time values don't have to be exactly correct. In fact, some aftermarket ECUs just provide generic values. Adjusting the fuel pressure will skew the results somewhat because increasing the fuel pressure does increase the injector off-time values. If the data sheet you find or receive is a short list, you can mathematically find the value in between any two consecutive data points, by averaging.
Example - The data sheet gives you the values for 8 volts and 10 volts but you want to find the value for 9 volts: 1.669 (8v) + 1.057 (10v) Ã· 2 = 1.363 (9v) Just make an educated guess for the values beyond 15 volts because you'll never be there anyway.
Another aspect of proper voltage is the alternator. Ensure the alternator can maintain the battery charge at idle. This is a big problem with older alternator designs (and they output a lot of electrical noise). However, modern design alternators are capable of producing most of their power (amps) at idle speeds. These modern designs simply have larger stators, more windings, better rectifiers, etc for greater idle speed output. Ford owners can retrofit a Ford 3G alternator. GM owners can retrofit a CS-144 alternator. I believe Mopar owners can retrofit a certain model Denso alternator.
Alpha-N @ Idle
Alpha-N @ idle is a fuel injection strategy, based on TPS position and RPM. It's for serious race engines with radical camshafts that can't maintain a steady kPa value because it fluctuates too much. Alpha-N @ idle still has the benefits of O2 compensation, however, it can't sense load because it doesn't use the MAP sensor. In Alpha-N mode, most ECUs use the MAP sensor as a Barometric sensor. Alpha-N @ idle should only be used as a last resort. Adding idle resolution to the kPa (Y) axis usually resolves issues with a radical camshaft. Of course, forced-induction applications must use speed-density mode (MAP sensor), since they need to measure load points in psi.