EFI Idle tuning and reversion
 

Hi-
Trying to sort out brand new build and hoping to get some advice on possible causes for what seems to be excessive reversion. I am unable to keep an O2 sensor alive (gets wet at anything below 1100 RPM) and in the process of spending time dockside working on tuning/adjustments, I have found that excessive moisture in the oil can accumulate.
The basic question is whether poor idle/tuning can be the main contributor or is the cam selection or something else about the set-up susceptible to reversion. The idle is not very smooth with what feels like a fair bit of misfiring. I have replaced the plugs to eliminate fouled plugs as a potential source but Idle is still quite rough. I am trying to tune the idle to below 900 RPM for drive life (IMCO SC) but so far have not been able to achieve something I would consider "smooth" and consistently firing on all 8 cylinders.
It is a Holley HP EFI MPI set-up running DIS (crank trigger), cam sensor for fully sequential and CNP. The engine is a 555 Dart Big M with Dart aluminum heads with CMI Sport Tube Headers. She made 710HP and 680Tq without really having the time to fully dial things in so it seems to run pretty strong. This is also my first time with fully programmable EFI so I am a bit of a Newb. The engine was dyno tuned but with dry headers and admittedly the dyno work was more for run-in and to dial in at higher power ranges. (very little work done to optimize idle).
I am hoping that if I can get the idle smoothed and better dialed in, I will be able to reduce reversion and have a better chance at using the O2 sensor and the closed loop learn functions of the software under actual use conditions. The O2 bung is located very close to the front of the tail pipe so about 5-6 inches behind the header collection point and approximately 12 inches ahead of the water entry point at the back of the tail pipe. I am also using Gibson clamp-on mufflers if that matters. I have thought about removing them for tuning purposes as something else to try.
It seems to me that I should be able to get the idle smoothed much better but with my limited experience, so-far, not so good. Oh yeah, Cam spec.s are as follows:
Gross Lift .650/.650, Duration @.05 247/253, LSA 112, Duration @ .006, 383/308, Timing Events (OP/CL) @.05 Intake 15.5/51.5, Exh 62.5/10.5.

I know there is a LOT to getting these things right but I am looking for any experience from more experienced tuners that might help me determine if there is something inherent with my set-up I need to focus on. I am getting to be fairly proficient in navigating the Holley software and have fiddled with the idle settings but so far have not been able to get to a stable and smooth idle below 950-1000RPM.
Sorry for the long story but season is getting short up here in Maine! Any help or advice would be greatly appreciated.
Thanks!
Paul

Sounds like one of your settings is wrong and or something is not in sinc.
I have HOLLEY EFI on my 598 and it idles in gear at 700 RPM hot.
And my cam is a LOT bigger than yours.
You have something wrong.
Does your cam have a different firing order than stock?
Any chance you have injector or coils wired wrong?

Tinkerer, Thanks for the quick response. Never thought to ask the builder about the cam firing order being different than stock. Is there a way to tell from the spec.s?
I have the EFI system set up for stock firing order - 18436572. I am using the Holley Coil driver packs for CNP which I did have to wire individually myself. I double and triple checked that before putting the motor in the boat but am not opposed to re-checking. I sort of figured things had to be correct for it to run as strong on the Dyno? Is it possible to get those results with something like firing/fueling order wrong?
I agree though, something seems to be wrong. I have been concerned that the spark may not be all that stable at lower RPM? I am using all Holley components (except for individual coils) including their DIS crank sensor and injector harness so that should be fine.
With the wiring pretty much all enclosed at this point, is there an easy/reliable way to check that the coil firing sequence matches the programmed firing order?
The engine starts right up and runs well over 1500RPM but I am at the point of suspecting just about anything at this point.
I am happy to hear you are having success. Are you running closed loop with the O2 sensor active?
Thanks

Yes I am running the O2 sensor.
My cam builder did the 4/7 firing order swap but I didn't see it on the cam card until I looked for the answer to my problem. Ran badly.
Changed the firing order and found a computer setting was wrong. Idles perfect now

Is your o2 senor mounted near the 12 o clock position on your tail pipe? Are you trying to run closed loop at idle?

Less timing at idle will help with reversion. I run my Holley o2 sensors full
time and use the innovative o2 heat sink to help with water contamination

O2 sensor is at the 9:00 position. And I run closed loop full time.
You can check the firing order by pulling all of the plugs and turning the engine over with a wrench and feeling for compression at each plug hole following the firing order. IF the cam has a different firing order you wont have compression at two plugs following the stock order.
Any chance the rocker arms are too tight causing valves to not close properly?
I had one of the initial computer settings off and lost spark in cylinder #5.
If I pulled the coil wire you could see it spark 7 times not 8. dead spot.
Go back over all of your settings.
Talk to someone that does this for a living.

My O2 sensor is mounted in the collector on my Lightning headers.

Why the heat sink for water? I looked it up and seemed more for heat, how does it help?

Where are you guys mixing the water at the tips or running dry?

Any issues with 12 o'clock? That's where mine is and the biggest problem that I've had so far is it's hard to hide the ugly wires.

Raises it up and protects it some. When sensor is installed in it, it's element will be located in the larger top section of the heat sink.

Take a look.
Innovate Motorsports 3729 Heat-Sink Bung Extender HBX-1
by Innovate Motorsports
Link: http://a.co/dtaAnSI

http://cdn7.frsport.com/images/detai...d88f606212.jpg

http://cdn5.frsport.com/images/detai...d88f5d9bcb.jpg
https://images-na.ssl-images-amazon....L._SL1500_.jpg

I run mine at the 12 o clock position and haven't had any problems with moisture. I run the NTK sensors and I think they hold up better than the Bosch ones.
I run closed loop with a NTK sensor on each bank, but I don't enter closed loop until 1800 rpm. At idle with the mufflers open, I revert outside air and the computers would try and compensate by going too rich.
Also if you have a plug out you will lean on that bank due to the unburned mixture pumped into the exhaust stream and the computer will try to compensate again.

Sb is correct it raises it up and protects it. The hole has an indexing mark on it so you locate the hole pointing forward and the exit is on the bottom. Some people have argued that it changes the reading. I have done back to back dyno pulls and only very slightly slows the reading down. The ntk sensor is a far superior sensor but find myself still using the Bosch and just replacing them alittle more often just to be on the safe side. Like stated earlier the sensor at the 12 o'clock position is critical. No matter how mild the cam is the sensor won't live if it's down towards the bottom of the pipe.

This is also a good point. Don't have the auto learn on at idle.

I have not used it, so I may be incorrect ....but I have used like products (actually spark plug foulers lol) to raise rear oxygen sensors in auto's which slows down the response of the 02. For the car guys- If it does not read as fast as the primary 02 sensor, sometimes the ECU will stop throwing cat converter codes. :) If it follows the primary 02 sensor's response, it will throw the codes.

Sorry for diversion, but it has a little something to do with the thread. :)

Yea have done that before on the anti fouler lol. If you call innovative they will tell you the magic is in the hole placement, size, chamfer etc. When I say slow down it really acts like the smoothing factor on a dyno. Makes the sensor act less erratic. You would see more delay in the sensor/afr response moving it from the primary to
the collector in a car type header.

I use the Innovate heat sink SB posted. I will echo virtually every comment that Indysupra has stated regarding using this product. AFRs are unchanged (I too did back to back).

For a marine application this is basically used to prevent water from coming in contact with the O2 sensor. Heat is not the issue.

I too use the Innovative extender. Can't remember the last time I replaced an O2 sensor. Same for me with back-to-back testing. It's a nice design - the chamfer creates a low pressure point at the bottom hole and helps draw the exhaust gasses through the front hole and into the chamber where the O2 resides.

Wow, Lots of good info here....
I am just back in town from travel with work.
I did order the "heat Sink" adapter from innovative but unfortunately won't have it this weekend! My bung orientation is what I would consider 9:00 (port side protruding horizontally to port). After reading this. I am wondering if I can simply rotate my tail pipe to put the sensor at 12:00 and make up a longer jumper hose to accommodate the connection between the header outlet and the tail-pipe inlet? Either way, for my current situation, getting the O2 sensor a bit out of the direct line of exhaust flow should help longevity. The last one I had literally died after about 20 minutes of idling out the channel I need to navigate before getting to running space.
Tinkerer, I discussed the cam timing with the engine builder and he was quite sure it was standard firing order but the process you described to verify makes sense to me and I will confirm using that.
Kvogt - The fouled plug issue is for real in my case. I have replaced my plugs and have yet another set. The fresh set definitely improved things but the idle is still very rough and if I pull the O2 sensor at Idle, there is lots of water spritzing out of the bung so I know it is getting very wet.
All- I agree on keeping closed loop off below higher RPM's but while this will help reduce erroneous fuelling adjustments, I don't believe if keeps the sensor from getting fouled by water/reversion?? Essentially, the sensor is powered (heater on) and providing signal output, just not feeding back to the ECU to engage the fueling adjustment algorithms, no?

I am going to try different Ignition settings in the software tomorrow and do some more timing light tests. Are any of you guys using Coil-Near-Plug (CNP) set-ups and the Holley Coil driver packs? There are choices in the pull-down menus for Ignition types. My current ICF includes the Holley 60-2 DIS but it seems to me the "custom" option should be selected as my Crank Sensor is set up 11 Teeth past TDC for Cyl #1. This was established through the mounting position and the instructions which recommended that the Sensor be located some number of teeth ATDC #1. This value can't be entered unless you are in the Custom configuration. I am also suspicious about some of the settings in the custom Ignition fields such as Dwell time. I have found some info about base settings for the various coil configurations. Any experience on Coil types/configurations?
I feel like maybe I should have purchased the Holley waste-fire smart coils but my prior EFI set up had individual 2-wire "Dumb Coils". Holley offers the Driver packs for this configuration but there is not a lot in the way of the details for set-up information.
I am also running fully sequential with cam Sync trigger. I fitted the engine with an EFI Connection magnetic cam phase sensor distributor housing (digital falling) in accordance with the Holley installation instructions.
Lastly, I am using the Holley Magnetic Crank trigger. Seems there are some software setting considerations for this as well.

Anyway, back into the bilge tomorrow to try some things. I will verify cam timing, and see if setting up Custom Ignition parameters (Dwell timing etc) help to improve Idle/reversion. I am also going to see how the timing light tracks expected firing events at cranking RPM to see if the different Ignition configuration settings have an effect.

Thanks to everyone for your contributions to this thread!
I will post with progress (hopefully).
GT2750

When I pull the bung out of my pipe it shoots water everywhere enough for me to ask the question here, and it turns out that any air leak will cause reversion. I have the heatsink extender and when I pull it out it's covered in dry soot with zero moisture. That was a real good demonstration on how a motor could hydrolock because of a bad header gasket.

If you have a fouled plug the engine will revert excessively because of the large delay between firing cylinders. Can you attach your program and do you have any logs to post.

I am running the Holley crank trigger setup with a cam sensor and CNP. The setup I use is Custom = 60-2. Inductive delay = 60 usec. Timing offset is used to correct the timing difference between the actual engine timing and the holley efi timing. TDC tooth = 11. Sensor type = digital falling.

I run a custom cam sensor that uses a MSD sensor. My set up is a single pulse type= digital falling.

Output setup = DIS coil on plug x RPM y MAP and I enable the dwell table.

Dumb question but what is your voltage at idle?

BR,
I have not ruled out air leaks but everything is fresh and tight - Gaskets etc... I am used to seeing water dripping at the header flange if things are leaking there and ave not seen anything like that.

kvogt,
I have not closely looked at idle voltage but the times I have glanced at the gage and the digital display on the laptop, it was reading 14V+/-.
Definitely with you on the plug fouling but I am not sure if mis-fires are more related to excessive water on the plugs.
No data logs yet but I will try and attempt to attach the current GF I am using. May have to upgrade my membership?

Hardware wise it sounds as though we have very similar set-ups.
I feel like it is definitely in the settings. Reading the manuals (system ICF Configurations section), it indicated for GM 58x crank wheel and "factory Coils" to set the dwell at 3.5Ms. With the dwell table enabled, do you set any base value? They also mention magnetic filtering for the crank trigger sensor. I believe the Holley sensor is a Hall effect sensor. Is this anything you enabled? It would seem the Holley sensor ought to be plug-n-play but one more thing I will look at.

My next step will be to move to the Custom ignition type and follow your settings. I'll take my old-school timing light out to see if the spark events are making sense at cranking speeds.

One more question, did you use closed loop learning at all for Idle tuning? Where is your timing and TAFR below 1000RPM? I ask because I have considered disconnecting the water jumper to the Tail pipe (dump it overboard) for short periods at Idle to keep the O2 sensor from getting wet as well as a way to sort out the spark plug(s) getting wet causing mis-fires. Obviously very limited to keep things from getting too hot!

Really appreciate you-all sharing this info!

Where's Haxby now ?

Actually, Alex (Haxby) has been a great help on this project! I could not have got this far without him.
Assuming he is busy, away or tied up with something burning.

I have the dwell table activated and my dwell values range from 3ms to 4.5ms depending on boost and RPM. This will also be coil dependent and I run the ING1A coils.
The Holley crank trigger is hall effect. I use the custom 60-2 settings.

I manually tuned the idle and shoot for an AFR in the mid 13s for both in and out of gear. Even if you dump the water out of the exhaust you will probably get false lean readings from outside air getting sucked back into the exhaust. With my mufflers closed the reversion is much reduced.

A brief update:
I went through cam timing by pulling the valve covers and walking through the standard firing order - Confirmed 18436572
I have an ATI Damper with full 360 degree markings and verified spark timing is correct for all cylinders.
Set ignition configuration to Custom and enabled the Dwell table. Set all values to 3.5Ms. (I subsequently looked up spec.s on the coils and it looks like they would recommend 3Ms for N/A, non-high compression.)
Verified voltage to be 13.7V-14.0V at Idle
Adjusted throttle plates to get IAC to hover in and around 15%-20% (Ran TPS Auto-set with each adjustment)
Ran static idle tests and while the measured timing is within a couple degrees of command, for some reason the timing offsets entered did not seem to have 1 for 1 effect. Not sure if inductive delay has an impact but I varied that a bit and it seemed the base timing stayed closer to set timing at lower values. This was baffling to me and any thoughts on what would impact the timing offset from being a direct influence would be helpful. I left the inductive delay at 60 to match Kvogt' settings and verified steady set timing up through 3200 RPM.
I also threw another set of plugs in.
FYI have done many plug readings after running at higher loads and RPM and they look great although I am still running open loop,

I was able to get an improved and stable idle in the 800-900 RPM although still rather "lopey"
I ran a couple of tests to see if the O2 sensor was still getting wet (used the fouled O2 sensors I have) and while reversion seems to be improved, the sensor still got wet. I don't yet know if it is okay for sustained idling without moisture/water getting into the oil. Where I boat, there is a 1/4 mile of channel to get to open water. I should also mention that the water entry point is more like 10 inches aft of the O2 Sensor bung and I am running a Hardin Marine SST water pump which pumps a crap-load of water.

I had ordered the Innovative bung insulator but unfortunately it did not come until after working on the boat so I did not have the courage to load a new sensor (to run in closed loop learn) in without it. I am anxiously awaiting next weekend so I can run the same wet tests with the bung insulator (and old sensor) before risking a 3rd new sensor.
I am wanting to run the O2 sensor because it seems I need to have excessively high TAFR's (14.5-15.5) to keep it from burning excessively rich at speeds below 1000RPM. (loading up and smells super rich). I did one trial where I set the TAFR (in the idle zones for RPM and MAP) closer to 12.5 just to see if the poor idle could be due to a lack of fuel. Because we had trouble with the O2 sensor at the Dyno, I am thinking my base fuel table was never right down at Idle speed ranges.

Anyway, some progress especially with the feedback and experience. Much appreciate the help and will be back at it Saturday.

Cheers,
GT