Initial fuel prime question

[Rich Z]

The description of the relevant tables (B3201, B3202, B3203) mentions that a prime pulse of fuel will be delivered when the ignition key is first turned to the "on" position. Does that mean when the starter is actually engaged or just when the key is turned to the position to where power is applied to the PCM? The reason I ask is because I am trying to firm up the start-up and I tend to turn the key on and allow it to set for a couple of seconds to get the fuel pressure up before actually starting the car. Cold start up is kind of weak, so I'm trying to figure out how to get a more solid sounding start up.

And along the same lines, at what point is the engine actually considered by the PCM to be "started" and the "after-start" tables kick in?

[joecar]

Hi Rich,

B3201 description:


When the ignition key is first turned to the on position the PCM will deliver a prime pulse of fuel.
This table defines the amount of fuel (in Grams) to be injected.

See also:
- {B3202} Initial Fuel Prime Delay


B3202 description:


Before the {B3201} "Initial Fuel Prime" pulse is delivered, the PCM will delay the prime pulse for this long after key on.

[joecar]

See B0106...

I don't know how else the PCM determines it's now in running mode.

[Rich Z]

See B0106...

I don't know how else the PCM determines it's now in running mode.

Hmm, in my tune, there is a value of "8" in that location. Since there are 24 reference pulses per crankcase revolution, my PCM switches from "cranking" to "running" after the crankshaft has only made a third of a revolution? That doesn't sound right to me.

[Rich Z]

Hi Rich,

B3201 description:


B3202 description:

Yeah, I read those descriptions, but I'm unclear about what they are telling me. Sorry, I like to understand the "why" behind things.

I guess I'm comparing this function to the accelerator pumps in the olden days with carburetors. You used to pump the gas pedal a few times before starting the engine to squirt fuel into the intake manifold. Obviously this would provide fuel for all 8 intake runners going to each cylinder. So in this case of EFI, do ALL the injectors provide a prime pulse of fuel merely by turning the key to the "on" position before the engine is even cranking over? Or does the engine actually need to start cranking before each cylinder gets that priming pulse in sequence?

B3201 appears to control the amount of fuel delivered in the prime pulse.
B3202 appears to control the delay between the ignition key "on" and when the pulse actually takes place.
B3203 isn't as clear to me, but seems to be an anti-flooding control to keep prime pulses from being delivered if the car doesn't start on the first and subsequent tries without allowing some time to elapse between those pulses.

Does increasing the amount of fuel substantially for the fuel prime pulse (B3201) help significantly with getting the engine to start with the minimum amount of cranking over? Anyone ever played with this?

And, about B3202. Is there any benefit to putting in any sort of delay here? I'm a bit confused about how this works. I'm presuming that once the engine is cranking, the injectors are providing fuel to each cylinder in sequence, so if the delay is longer than the time I move the key to engage the starter, does that cancel the priming function? As I mentioned, I normally turn the ignition key to "on" and then wait a few seconds for the fuel pressure to stabilize. Yeah, I know I probably don't need to, but it's gotten to be a habit to check the "vital signs" before engaging the starter. So would, then, a delay in this priming pulse be better to put the priming pulse closer to when I actually engage the starter? Or is this completely insignificant?

[vcls1]

hey the way i understand it to work is when the ignition is switched to on, prime takes place but that is only to pressurise the fuel rail, injector pulse does not take place until crank rotation hits the determined value of crank trigger sinusoidal pulse signals which you had said was set at 8 being 1/3 of a crank rotation.
As far as your startup is concerned i would be looking into IAC and commanded fuel when cranking to try and get it to burst into life a little better.

[hog]

There is alot going on during this time.

I have found tables for:

1st crank pulse fuel mass vs. coolant temp,
2nd crank pulse vs coolant temp
2nd crank pulse delay time vs. coolant temp (measured in crank pulses) 6 pulses was stock
Prime Pulse Delay after powerup vs. Coolant Temp (seconds)=0.3-0.1seconds depending on temp
Prime Pulse Enable Minimum Runtime vs Coolant Temp (seconds) 10 seconds
Crank Fuel Barometer Gain Multiplier vs Baro (multipliers 0.650-1.000)
Prime Pulse Fuel Mass vs. Coolant Temp. (0.300-0.018 grams)
Normal Crank Fuel/Air Equivalent Ratio vs. Coolant Temp. vs. Time
Extended Crank Fuel/Air Equiv. Ratio vs. Coolant Temp. vs Time
End of Injection Timing Trim Pulse vs .Coolant temp (Low Resolution Periods)
End Of Injection Timing vs. Coolant Temp (Low Resolution Periods)
Prime Fuel Disable Engine RPM


Not sure if its possible on the Vettes, but on the trucks if you have the key in the run position, and you get under the truck and grab the crank pulley and slowly turn the engine over(slow enough to let the compression bleed off so you arent fighting it) eventually all of a sudden the fuel pump will cycle. It will cycle for 3 seconds stock IIRC(there are PCM tables that control this pump prime length) and it will cycle twice for every complete rotation of the crank, so 4 pulses in my case and 24 pulses in your case.

It would be easy enought to see when the injectors allow a prime volume of fuel by monitoring their power supply. If they pulse when you turn the key to run, then that is the prime pulse volume. If you dont detect any injector activity upon Key On, then you know that the Prime pulse volume must occur when you continue from the "Run" to the "Crank" position.

Rich, B3203 simply holds back the Secondary Prime event until that timer has been met, provided there is a continued no start. Once the timer calibration has been met, the PCM allow the secondary prime pulse. If this table wasnt present, the PCM would tell the injectors to continually give the engine its secondary prime pulse volume, every 6 pulsecounts(or whatever the 2nd crank pulse delay time-measured in pulsecounts is entered as a value) until either the engine started or the engine flooded out.
So primarily you have the primary injecor prime pulses, if that doesnt work, after a certain amount of crank pulses you'll get the secondary injector prime volume, then if it hasnt started, the actual TIME counter kicks in and wont allow more fuel until the TIME counter has expired. These are interwoven and overlapped, all in an attempt to prevent flooding.

Then if it does flood, you hold the accelerator on the floor(which kills the injector pulsewidths by forcing the ECM/PCM into "CLear Flooded Engine Mode". Once you crank the engine over for a few seconds without any fuel being added, you can release the accelerator whgich allows the crank/startup tables to be followed again thus allowing fueling for startup.

peace
Hog

[Rich Z]

Well, OK, that all makes sense to me.

I've actually had pretty good success fiddling around with the tune, and the car starts with the engine just barely cranking, which is the way I want it. And it's not just some feeble start, neither, like one cylinder starts firing and the others decide they might as well too. I turn the ignition key to ON, and let it set for a couple of seconds, then engage the starter and it starts up firmly in about a second.

I took baby steps in the changes I made, so it took me a few days to get it where I want it. So I'm satisfied with what I've got now and can move on to some other minor problems.

The most annoying of which is that when I start the car, I usually let it sit for a while, as I am logging with the scanner and like to just see how things are looking while the engine temps start warming. It idles real well, but when I back out of the garage and take my foot off of the gas pedal to drift into the driveway, the engine has a very pronounced surge to it. When I stop, there is no surge, and when I start moving forward, even taking my foot off of the gas to drift at the same 2mph, there is no surging. Only seems to happen when going in reverse out of the garage. Does the PCM really even know that the car is going in reverse? If so, does it use different tables for that condition?

Along the same lines, does the PCM know when the clutch is depressed? Since there are tables indicating that certain tables and values are used depending on whether the drivetrain is "in gear" or at "idle/park", I guess there must be some way for the PCM to know this. Honestly I haven't yet looked for a PID that indicates this and I've never looked for any sort of sensor actually on the clutch engagement system itself. So how does the PCM know whether you are "in gear" or not?

[IJ.]

There will be a Clutch switch for Cruise control and start inhibit, I'm assuming the PCM can see this?

[joecar]

Yes, the PCM see the the CPP (clutch pedal position) switch, it uses this to choose between the P/N tables or the IG (in-gear) tables.

[Rich Z]

Yes, the PCM see the the CPP (clutch pedal position) switch, it uses this to choose between the P/N tables or the IG (in-gear) tables.

Is that switch, or in any other way the status of "in-gear", monitored via a PID?

[joecar]

The is a pid, either STATExx or TSTATExx.

[Rich Z]

OK, I think I am making headway here...

The cold start tuning has been coming along pretty well, so I made up a video showing the progression along the way. I still want the initial startup to be stronger than it is, however. But at least I got the startup to be QUICKER after turning the key for the starter.

https://www.youtube.com/watch?v=edI3F-yjnGA

https://www.youtube.com/watch?v=edI3F-yjnGA

I was feeling pretty good about the progress I made when I started up the car on Saturday morning to go along on a cruise up to Thomasville, GA. In the past when I would start the car up, I would just leave my foot off of the gas pedal completely the entire time till I let the car warm up a bit, but this time I goosed it a little bit and the engine just fell right on it's face when the rpms came back down to what should have been idle speed. I cancelled the trip, not knowing what the heck was going on. Later on in the day when the car had cooled down sufficiently, I loaded in one of the old tunes from my tuner and tried the same gas pedal blipping thing, and it did the same thing. Apparently it isn't anything I inadvertently introduced with my own fumbling tuning efforts. It's just something that hasn't been addressed yet in this tuning effort.

<continued below as apparently only one video is allowed per post here>

[Rich Z]

So yesterday I set up the laptop for logging and also had video running when I did another cold start to check out this problem. I then took video off of the laptop and merged that video with the audio of the car actually running so I could watch the log data while listening to what the engine is actually doing.

https://www.youtube.com/watch?v=eogTdJ-l9aQ

https://www.youtube.com/watch?v=eogTdJ-l9aQ

Obviously I had to cut and paste sections of the video of the log file to fit the audio, as it appears that one or both are not actually in "real" time compared to the other. But it is close enough, I think.

So apparently I still have more work to do in this department concerning the cold start process. The engine is running in open loop at this time, so I need to figure out what tables would be active when the engine is coming down from an abrupt blip to the gas pedal and try to compensate for this transition that allows the rpms to drop down into a stall. It is only an abrupt transition that causes this problem, as a slow pressure and release seems to work OK.

Ah well, more digging...

[Rich Z]

Well, the most obvious event I can see when the engine stalls after the throttle blips is that the throttle blade opening goes from the normal 17.3 percent down to 14.1 percent, so apparently the engine is starving for air flow, which chokes the engine at the lower ECT levels. Once the temps reach a higher level, then that throttle blade choke doesn't affect the engine as drastically.

Anyone got any ideas on what table(s) I need to address to get that throttle blade to come back down and STAY at 17.3 percent after the jabs at the gas pedal?

[ScarabEpic22]

Have you adjusted your idle airflow tables at all? I know for the F body (cable TB) you need to do RAFIG and RAFIP/N, but I dont know about the Vettes. I had to add a bunch of air to my TBSS when I went to my longtubes and 4" intake, but the tables wont be the same (LS1 vs E67).

[Wheelz]

I had that exact problem 2 months ago when I got my 05 5.3 silverado (i have no idea why.. its bone stock). I bumped B2706 max etc idle contribution from 4.69 to 6. Mine would also stumble slowing to a stop occasionally so I bumped B4309 throttle cracker in gear 1000 rpms at 0 mph from 0 to .05. I also idle at 625 rpms, but that didnt seem to have any effect on the stumble after rev

You might also play with B4307 desired airflow to fine tune added airflow at each engine temp, which should open the blade up a little more

Edit: not sure if my table numbers will match (ls1b). I assumed the basic ones were the same

[Rich Z]

Thanks for the tips. This cold start tuning is a problem because I really only get one shot per day to try out any changes. So I'm changing several tables and will see if they help. Not the optimum way to do this, but with the temps warming up quickly during the day, I only have the morning to start the car when it is relatively cool.

[hog]

Excellent video's Rich. I have never seen video datalogging overlayed with engine audio, very nice thanks for sharing it.
I notice that right after the throttle stab the timing advance drops off down all the way to 0º and follows the rpm down as the engine stalls.
Once you are in Closed loop when the ECT reaches 115ºF your issues clears up. Even as the rpms come down after a throttle jab, the timing advance will NOT come down under 16º while engine revs never drop under 700rpm.

I just noticed that there is 2 tables for:
Base Spark Advance vs. Load vs. RPM(Closed Throttle, In Drive)
Base Spark Advance vs. Load vs. RPM(Closed Throttle, in Park/Neutral)

Does it make a difference if you push in the clutch pedal while in closed loop?

Your cranking times have come down nicely in your post #13.

peace
Hog

[Rich Z]

Thanks. Just seems to me that being able to actually HEAR what the engine is doing while viewing the log is beneficial. Sure wish EFILive could do something like this natively.....

Anyway, I'm just running out of ideas on this stalling problem. I made a lot of changes in the startup and idle sections and nothing made any difference at all on the engine falling on it's face when ECTs are cold and I tickle the gas pedal. I'm thinking the problem is with the throttle blade closing too far after I blip the accelerator and the RPMs come down. As for timing, well, maybe it's a chicken and the egg thing, but I'm thinking the timing is trying to recover from the engine going into a stall by pulling back on the timing. What exactly does the "stall saver" do that is mentioned in some of the tables? I've tried to figure out how to get that to kick in, but so far nothing I have done has had any effect. I looked at all of the enabling settings and tried to make sure they weren't inhibiting stall saver from kicking in, but either they don't do what I want them to do, or I'm still missing something.

So, reducing it down to as small a step as possible, HOW do I keep the throttle blade from closing up to 14.1 percent? How can I tell it to stay opened up at 17.3 percent to give the engine more airflow?

There are an awful lot of trees in this forest, so I'm sure it's just a matter of me not barking up the right one.

Oh, I forgot to try pushing in the clutch pedal during the stall area. I have the tranny in neutral, of course, so does the clutch pedal signal the PCM in some manner to command something different with the engine management routines?

I've attached the actual log file if anyone wants to see it natively.

Thanks....

[hog]

I think you may be right about the timing and the "Chicken and egg" thing. I wonder what would happen if you "locked" in 15º of timing advance"? I do know that typically you can give a cammed engine more timing to help stabilize the idle.

I just did some digging about the "Stall Saver" routine and apparently when rpm falls below a certain level, the PCM enters the "stall saver" mode. This allows the derivative RPM idle control system into action at a faster rate, and will disengage the A/C clutch in order to prevent a stall.

As Joecar stated, the clutch depressed would signal P/N tables, but with the VSS reading ZERO speed, and the clutch pedal released, does the PCM use an " P/N" table or "In gear" table?

peace
Hog

[Rich Z]

The is a pid, either STATExx or TSTATExx.

OK, I see some PIDs that are STATE00 thru STATE07. These are bit mapped pids, from what it is telling me. In STATE02, bit 6 indicates the status of the clutch pedal switch as being either released (status=0) or applied (status=1). There is no TSTATExx that I can see in my PID list.

So what is this telling me? Is the status of whether the car is in gear or in park/neutral solely determined by the status of the clutch pedal position with a manual transmission car?

[joecar]

OK, I see some PIDs that are STATE00 thru STATE07. These are bit mapped pids, from what it is telling me. In STATE02, bit 6 indicates the status of the clutch pedal switch as being either released (status=0) or applied (status=1). There is no TSTATExx that I can see in my PID list.

So what is this telling me? Is the status of whether the car is in gear or in park/neutral solely determined by the status of the clutch pedal position with a manual transmission car?Yes, correct, P/N is clutch pedal in, IG is clutch pedal out.

[Rich Z]

Yes, correct, P/N is clutch pedal in, IG is clutch pedal out.

Yeah, the PID I used when I scanned shows the indicator changing when I push in the clutch.

So does the PCM also know whether the manual transmission is actually IN gear? Since there is that skip shift PIA, I'm thinking there most be some information of that nature passed to the PCM. And I see torque control values for individual gears as well.

Anyway, I'm just not having any luck at all figuring out how to get rid of the engine falling on it's face when I blip the gas pedal when cold started. Once the ECT gets over 100 degrees, the throttle blade being set at 14.1% doesn't bother it any longer. But I just haven't found the magic incantation to keep the throttle blade from closing down to 14.1 from 17.3 after I blip the pedal. All of the air control tables I've fiddled with don't make a darn bit of difference to that setting. There must be a way to do that, but darn if I've been able to find it yet. I've looked over older threads on this site for similar problems, but so far none of the things those people tried has helped my problem.

Anyone want to toss me a bone? Or is this just an unresolvable problem using this EFILive package? In all other aspects the throttle body seems to be working properly, so if it's a mechanical problem, I will be at a loss to figure out what that might be as well. It appears the PCM is commanding the throttle blade to go to 14.1% so how the heck do I tell it to NOT do that until the temps reach 110 degrees?

[joecar]

...

So does the PCM also know whether the manual transmission is actually IN gear? Since there is that skip shift PIA, I'm thinking there most be some information of that nature passed to the PCM. And I see torque control values for individual gears as well.

...It figures it out somehow from the input:output speed ratio which gear it is in.

[Rich Z]

It figures it out somehow from the input:output speed ratio which gear it is in.

Yeah, I guess for any table intersection of engine speed with vehicle speed, the transmission can ONLY be in one particular transmission gear.

That aside, any ideas on why I can't seem to make any headway in getting my throttle blade to stop closing up on me after those throttle blips in cold starts? Or am I just barking up the wrong tree completely and need to fiddle with timing instead? Am I chasing the chicken when I need to be chasing the egg?

[Rich Z]

Well finally! Something I did made a difference. I made changes to 13 tables that I thought might make a change to that stalling problem. Apparently one or more of them did the trick. Now I know this really isn't the recommended way to do tuning on a car, especially if you don't know what you are doing (like me) as most people will recommend only making a single change each time and then test the results. But this would take forever with the cold start issues. The iterative process is pretty drawn out because of the time necessary for the engine coolant to cool back down in between runs. So I can only get maybe two runs per day to check out the changes I make.

Anyway, here's where I am now...


http://www.youtube.com/watch?v=FglvcTzbAQE&feature=youtu.be

The stalling has ceased, probably because the throttle blade doesn't drop down below 16.9 percent. But there is still a loping of the idle after the gas pedal blip. And apparently idle is revving up quite high as the engine warms up right before going into closed loop mode.

Some of the difficulties of this stuff is trying to figure out whether the chicken or the egg came first when looking at the results seen in the log data. You can see the RESULTS, but what is the actual CAUSE of what you see? Is airflow, or the AFR mix causing the timing to change, or visa versa?

So I'm going to make changes to each of those 13 tables mentioned above, in order, one at a time, to try to get the idle to come back down to a more normal speed. The engine loping is still an issue, but it's not really a killer like the stalling issue was. I may or may not spend time trying to track that down and fix it immediately, as I want to move on to throttle response as soon as I can.

But at least now I feel that I am moving forward...

[joecar]

Well finally! Something I did made a difference. I made changes to 13 tables that I thought might make a change to that stalling problem. Apparently one or more of them did the trick. Now I know this really isn't the recommended way to do tuning on a car, especially if you don't know what you are doing (like me) as most people will recommend only making a single change each time and then test the results. But this would take forever with the cold start issues. The iterative process is pretty drawn out because of the time necessary for the engine coolant to cool back down in between runs. So I can only get maybe two runs per day to check out the changes I make.

...

So I'm going to make changes to each of those 13 tables mentioned above, in order, one at a time, to try to get the idle to come back down to a more normal speed. The engine loping is still an issue, but it's not really a killer like the stalling issue was. I may or may not spend time trying to track that down and fix it immediately, as I want to move on to throttle response as soon as I can.

But at least now I feel that I am moving forward...Hi Rich,

Sometimes it takes a "special" approach...

idle tuning is the most difficult for the reasons you noted.

[joecar]

...

Anyway, here's where I am now...


http://www.youtube.com/watch?v=FglvcTzbAQE&feature=youtu.be

The stalling has ceased, probably because the throttle blade doesn't drop down below 16.9 percent. But there is still a loping of the idle after the gas pedal blip. And apparently idle is revving up quite high as the engine warms up right before going into closed loop mode.

Some of the difficulties of this stuff is trying to figure out whether the chicken or the egg came first when looking at the results seen in the log data. You can see the RESULTS, but what is the actual CAUSE of what you see? Is airflow, or the AFR mix causing the timing to change, or visa versa?

...Can you post the log file from that video, and can you place a bookmark at the problem.

[Rich Z]

Well, I don't have my laptop running to put in a bookmark, and the log file is on my network file server. But the oscillation problem happens right after I stab the gas pedal and continues for 19 or so seconds afterwards.... It's pretty obvious on the log chart with the RPM trace looking like a slow sine wave.

BTW, I emailed you both the log file and the latest tune file.

Later on today I made some minor changes to B4307 that seems to have corrected the higher RPMs I noted earlier. I lowered cells 90 thru 133 20 percent as it appeared that the RPMs were climbing too much in step with the coolant temps rising.

Thanks.

[joecar]

ok, thanks.

[Rich Z]

Anyone willing to take a stab at letting me know what might be causing that engine speed oscillation after I jolt the gas pedal after a cold start?

[joecar]

My stab/guess: it looks like it may be trying to use spark timing to adjust engine speed, but maybe spark timing is too high.

( now, the same comment could probably be applied to idle air )

( spark timing is related to idle air in the same manner as chicken/egg )

[joecar]

Spark timing seems to overshoot to bring engine speed up;

airmass seems to undershoot to bring engine speed back down.

[Rich Z]

Spark timing seems to overshoot to bring engine speed up;

airmass seems to undershoot to bring engine speed back down.

So this issue appears to be timing related entirely? If I can stop the overshoot via timing, then the undershoot won't need to take place with airmass?

[Rich Z]

Speaking of which.....

Any ideas on why there is an apparent state change from idle when I blip the gas pedal? It acts like for 20 seconds it is accessing different tables after that event than it was using prior to the blip. So apparently the gas pedal blip is the trigger for that change.

Is there any way in EFIlive to actually show (as per something similar to a debug routine in most programming languages) which tables are being accessed by the PCM in real time?

[GMPX]

Is there any way in EFIlive to actually show (as per something similar to a debug routine in most programming languages) which tables are being accessed by the PCM in real time?
Unfortunately that is technically impossible to convey down a super slow VPW communications bus.

[Wheelz]

Unfortunately that is technically impossible to convey down a super slow VPW communications bus.

Is that a feature that could be worked into the scan tool viewer/runway highlighting... post processed? Assuming you selected the appropriate pids

[Rich Z]

Is that a feature that could be worked into the scan tool viewer/runway highlighting... post processed? Assuming you selected the appropriate pids

Yeah, I would settle for that without the real time requirement. There seems to be several alternate tables that COULD be used based on values from sensors and settings in enabler tables that are difficult to predict just looking at one table at a time. For instance, just looking at the spark advance tables in my tune, there are SEVERAL possible tables that could be being used at any given time. When I am looking at a log graph, at any particular point, WHICH table, or tables, is actually being used to provide that data point? The graph only shows me the results. Without sitting down an drawing out a flow chart (even assuming I will find every single enabler switch possible), it's pretty darn tough figuring out which table is actually TELLING the PCM what spark advance value to use.

Heck, just with B5916, which is the Base Spark Closed Throttle Threshold, the description says "If the throttle is below this value and vehicle speed is below threshold, then it is considered closed and the base spark tables will be used." Well, that's fine, but which spark table is being used when the throttle is ABOVE this value? Looks like it could be any of several based on ECT, IAT, knock sensors, fuel mixture correction, EGR spark correction, in gear or in park/neutral, etc., etc. And then we have multipliers to deal with as well.

Heck, has anyone actually created a flow chart of this process to be able to MANUALLY trace what table is being used based on values that ARE discernible?

[Rich Z]

Spark timing seems to overshoot to bring engine speed up;

airmass seems to undershoot to bring engine speed back down.

Actually I think you may be right on the money.

I've been playing around with B4307 (Desired Airflow) and I've made a lot of headway. It looks to me that the values I had in this table were providing too much airflow, and when I destabilized idle by punching the throttle, airflow and spark advance were playing a tug of war trying to get the idle to stabilize again, taking 19 to 20 seconds before they were able to work together. I was also having a problem with high idle hanging when coming to a dead stop. Fiddling with this table seems to be getting all of the problems (idle surge, stalling, high rolling idle) straightened out, by reducing the amount of airflow. I've got the destabilization down to just 2 seconds, rolling idle runs around 900 or so rpm, and it no longer stalls when I punch the gas pedal right after cold start. I just need to work with it a bit more to try to find the sweet spot to get that idle destabilization down to zero seconds without messing up the other facets.

Speaking of this B4307 table, can anyone tell me why the values in the Park/Neutral column are all zeroes? I checked a stock tune for a 2002 Z06 and that is what is there too. And there are other tables where there are only values in the In Gear column, but nothing but zeros in the Park/Neutral column. Is that the way it is supposed to be?