Page 1 of 4 1234 LastLast
Results 1 to 20 of 77

Thread: Cold start seems a bit...off SD

  1. #1

    Cold start seems a bit...off SD

    I seem to be having a bit of trouble getting my cold start and warmup setting sorted out. Included below is the log file, tune file, and the layout file for my scanner. It uses the egr port for the aem wideband. This is a speed density tune, there is no MAF.

    Does anything jump out at you about it and why it would act the way it does? Thanks. I realize it looks a bit rich but it was running pretty good before it got cold and I was getting close to 20mpg.

    Jim
    Attached Files Attached Files

  2. #2
    Tuning Addict 5FDP's Avatar
    Join Date
    May 2012
    Location
    Rogers, MN
    Posts
    13,533
    How come you have the idle timing so low for park/gear?

    You also have a ECT spark correction table removing a ton of timing when the engine is cold. I'd probably zero out that ECT table for the colder coolant temps and adjust the idle timing tables again. Have you tried adding more base running airflow at idle?

    The injectors are kinda going nuts too, huge spikes in the pulse width when first started. Hard to see what the MAP is doing without it being logged.
    2016 Silverado CCSB 5.3/6L80e, not as slow but still heavy.

    If you don't post your tune and logs when you have questions you aren't helping yourself.

  3. #3
    Quote Originally Posted by 5FDP View Post
    How come you have the idle timing so low for park/gear?

    You also have a ECT spark correction table removing a ton of timing when the engine is cold. I'd probably zero out that ECT table for the colder coolant temps and adjust the idle timing tables again. Have you tried adding more base running airflow at idle?

    The injectors are kinda going nuts too, huge spikes in the pulse width when first started. Hard to see what the MAP is doing without it being logged.
    Thanks for the suggestions. The MAP is logged and it pretty much tracks the cylinder loading.

    In a word, clueless? If you look at the spark tables it's clear that the controller is pulling out a considerable amount of advance and I'm still learning what causes it to do that. I know the adaptive spark control is involved but I was unaware of the ECT table. I'll zero that out and see what it does. Not sure if the adaptive will offset it anyway, but I figure the fewer cooks in the kitchen maybe the less commotion. I find the use of the adaptive to be very good at controlling oscillation but suspect that at these colder temperatures it's range may be exceeded.

    Yes the injectors are going nuts. That is the major symptom perhaps but I don't think it is the cause since it runs fine after warmup. Something must be causing that. I've been mostly looking at the enrichments, and in particular the Park/Drive Enrichment Initial Adders vs ECT (12347) but I'm not getting the results I expect in terms of leaning out the area where the injector spikes are. Clearly I'm missing something.

    Jim

  4. #4
    I think that was a positive move, despite the shorter run time. Datalog attached. The advance came up some and the injector spikes eased off a good bit. Could you suggest a next move? THX.

    Jim
    Attached Files Attached Files

  5. #5
    Tuning Addict blindsquirrel's Avatar
    Join Date
    Apr 2017
    Location
    Meridian MS
    Posts
    7,451
    You are logging MAP (volts) but not MAP (pressure) and in a custom setup like yours with so much changed, knowing that the linear/offset are producing the correct pressure reading is what's important, since that's what all the tables in the tune use. None of the tables have a 'volts' axis.

    And just in general, why so few channels? Easier to look past a channel you recorded but didn't need than it is to find out what was happening with a channel that wasn't logged at all. (insert bad metaphor here something something dark room blindfold while wearing mittens etc.)

  6. #6
    Yep, that's me, mining for acorns using a feather to dig. I sorely wish to have had the benefit of your experience in setting it all up, but needs must you know. I can only plead ignorance. Had to get a reading somehow. Many might consider me unteachable I'm sure.

    As for the channels, I do have vast experience with lag and sought only to reduce that any way I can. Got the fairly hot laptop, check. Bought the new usbC cables, check. Trimmed the list to what looked like the essentials, check. Got more or less thoroughly confused by online advisers, check. Added and removed channels willy nilly to arrive at this destination, check. It's an ongoing project.

    Jim

  7. #7
    Tuning Addict blindsquirrel's Avatar
    Join Date
    Apr 2017
    Location
    Meridian MS
    Posts
    7,451
    I am routinely logging 50 PIDs with a dinky-assed 'slim' Toshiba 1.6GHz dual core and don't have any glitching or sync issues. It is Win7 which is a little lighter resource-wise, but then again has all the pretty bits still turned on so it's probably a wash compared to a less out of date OS.

    I didn't mean to sound critical, just that some day you'll have a need to refer back to an old log to compare something and the thing you need isn't going to be there.

  8. #8
    Senior Tuner kingtal0n's Avatar
    Join Date
    Feb 2017
    Location
    miami
    Posts
    1,799
    I don't think the number of PID really matters as much as the logging rate. You can right click to adjust the log speed, if you choose say 1ms is means 1000 data logging points per second for that one PID, which can significantly tax a computer hardware. I recommend going through your entire list as a sort of pre-tuning checklist to ensure each logging rate is exactly as low as it could possibly be, in order to free up processing power and reduce the chances of having a PID "lag" or "incorrectly" capture data which I have seen happen.

    Similarly for PID which are very important such as RPM input and wideband inputs, make sure the rate is fast enough. For example you can't really use wideband data that is being logged at 500ms , it would be awful.


    Another bit of advice just sort of as an aside. I notice you are trying to learn and frequent posting thoughtful ideas of how to learn.
    Something I Do personally when first picking up a new stand-alone software (or ECU) is go through the entire software, click and read everything. Literally just go button by button, page by page, and make notes as you go, connect the dots between maps and charts so you try not to miss anything, You may find many hidden maps that can influence your timing and fueling. Try to make use of everything and test everything if possible, because some of the maps won't work the way its advertised. For example in one of the PE tables you can enter positive or negative values, but in my 411 when I try a negative PE timing subtractor it sends 100% injector pulse instantly. So there are some bugs and things that won't do what it says.
    Last edited by kingtal0n; 01-20-2022 at 11:39 PM.

  9. #9
    Ah, I think I see the problem here. Because this engine is normally fitted with a blower I'm running a 2bar tune. Right now the 'mule' is in the car which is N/A but within the year the boosted engine goes back in. Since they are more alike than different it makes sense to tune using the 2bar setup, which will make the fine tuning much simpler after the swap. However, that means running a 2bar MAP sensor, and that throws a curve since it is not a generic parameter. I'm not real sure how to deal with this and that's why it's set up the way it is and if anyone can suggest a better way I'll try it. I was not getting any usable data the way I tried it before. I'll take a look at the logging rates, that sounds like something that could really help. But in the meantime any suggestions about the warmup settings are the main focus right now.

    For instance, I've been fighting with the enrichments. My prime pulse mass [12891] is zeroed out because my fueling system bleeds off so there is no rail pressure at initial power-on. My first pulse Mass [12892] and 2nd pulse [12893] were presently set to exponential curves but I've dropped the 1st to zero at 50 and below as a test. Seemed to help a little. Any insights here could go a long way.

    In that last test the spiking of the injectors dropped off dramatically so it's clearly closer. Ended up not being enough fuel to keep it running but that's OK, if I can get the trace fairly flat I can deal with the level. It's clearly a balancing act.

    Jim

  10. #10
    Thanks for the advice guys but can anybody suggest the next step? I mean it's pretty clear that after zeroing out that timing table it's not going to be easy to take that approach any farther. And it did make a pretty dramatic improvement. That's about the best result anyone can ever hope for from online advice I realize, maybe I'm just being greedy asking for more... but... Is there any more?

    Jim

  11. #11
    Senior Tuner kingtal0n's Avatar
    Join Date
    Feb 2017
    Location
    miami
    Posts
    1,799
    for what, cold start?

    My advice is different than usual. I like my engine to run kind of lean all the time, whether cold or hot. My hot idle cruise is 14.8 to 15.5 usually. My cold starts as well.

    It doesn't hurt the engine, its just going to run rougher. If you have automatic trans it doesn't hamper the ability of the car to drive normal.
    A manual trans I would probably smooth the engine out a bit using richer a/f

    I prefer lean a/f because that is keeping the engine more free from carbon buildup, plugs stay cleaner, last longer, CC will less coating with carbon. Pistons stay cleaner.

    None of that matters for the initial fire-up event, however. I am not sure whether you are having trouble getting it to fire up or getting it to run good when cold.
    Because you mention the prime and crank tables.

    For cold start (fire up) I found the best table that works for me is the FA Multiplier. No matter what values I tried in the first and second pulse mass I could never get a very responsive fire but when i tried a large number (like 8.00, from the original 1.25 or whatever on the first line) it gets the engine going rather immediately when cold in the FA multiplier.

    From there, the table for stoich target is useful, dial it to get some extra fuel when cold but like I said I don't run it rich. Still 1.12 to 1.05's in this table will land me in lean territory so it is useful to help smooth the engine even while lean and cold

    Once you get 130*F the 'cold' stuff is basically over with. Are you having trouble getting a fire up initial?

    There is a table for airflow with the engine off key on, and one for engine off key off. Two different values I think, one called "park position" and the other is a IAT vs Cranking airflow type of table. I raised the Cranking airflow vs IAT table for my hot starts but not so much cold starts. I also make sure park position is open more than cranking so if the iacv is moving during cranking or key on it is always closing which helps prevent flooding or fuel suspension/vaporization issues while cranking because it was open too far at first when you got the first shot of fuel.
    Last edited by kingtal0n; 01-22-2022 at 02:33 PM.

  12. #12
    Thanks Sire, some really good stuff there. My brain is on hold today, what with the playoffs, cold weather and all, but I'm going to give very close attention to everything you've said. May take me a few days to digest it all and try various remedies.

    Yes, cold start is the primary issue. As it stands, regular warm start generally requires double bumping the starter and cold it is worse. More cranking required than you would want to be sure. Then once it lights off I have that erratica in the injectors to deal with. Warm it doesn't do that.

    Plenty of time later to address the running mixture. I'd like eventually to lean it out but of course to do that properly the transitions have to be handled well and I'm nowhere even close to having that sorted yet and don't expect to be before mid-summer if then.

    Your help on these preliminaries are wonderfully helpful and greatly appreciated.

    Jim

    (Beg your pardon on the wording, I've been reading Stevenson)

  13. #13
    OK, kingtal0n when looking at my air fuel multiplier (AFx) tables in the cold start regions my values start high and go low. At 32 degrees I start out at 6.65, and 5.12@50 dropping to 1.33 and 1.02 respectively by the bottom of AFx2. So I am already much closer to your preferred values I think. Would you recommend bumping them up by 25% or so? I aim to try that tomorrow absent advice to the contrary and I'll just see what happens that way.

    Then immediately after starting I have that injector craziness where I think something is out of range. I wouldn't expect the AFx to change that, but I'll wait to see.

    Jim

  14. #14
    Senior Tuner kingtal0n's Avatar
    Join Date
    Feb 2017
    Location
    miami
    Posts
    1,799
    I Cant make sense of what you say try a screen shot of the table. x-axis is temp and y-axis is in seconds I think? It doesn't say iirc

    Also post a shot of the stoich target table (temperature vs multiplier)

    And the airflow vs IAT
    and airflow PARK

    Raise the airflow vs IAT so the engine can breath better when cranking until you get the fuel quantity sorted out (it will band-aid the poor fueling for crnaking tables for now)

  15. #15
    Attachments:
    latest datalog with 25% increase in warmup area of AFx tables. No significant change that I can see, startup was unaffected.
    screenshot of AFx1 and AFx2
    Initial adders vs IAT & ECT + Park/Drive enrichment vs ECT + prime pulses + delay decay.


    I'm not finding the Stoich Target table. I do have a Stoich AFR (14.63) Could it have a different name or be in a later ECM? This is a '0411.
    Also not finding Airflow vs IAT, same with airflow Park.

    I suspect I just don't know the common abbreviations. It would be EXTREMELY helpful to me if you could give the path to these tables that you'd like to see, even if only the first letter of each step in the path. That way I can get exactly the table you want to see and don't have to guess if it is right or not. I realize this is not usually done, those of you experienced with HPT know where they are and eventually I hope to also but I'm a slow learner where strict memory is needed like here. I have trouble seeing the right tree for the forest. Plus, if it isn't there because of a version change or something I can confirm that. Please have pity on a poor ignorant student of the art.

    Many thanks,

    Jim
    Attached Images Attached Images
    Attached Files Attached Files
    Last edited by Jim Blackwood; 01-25-2022 at 01:00 PM.

  16. #16
    Senior Tuner kingtal0n's Avatar
    Join Date
    Feb 2017
    Location
    miami
    Posts
    1,799
    I have a 411 but mine doesn't have the second FA Multiply table. I've never seen that nor do I Know why there is a second table for the same thing.

    In any case here is what I notice

    Your first pulse mass table values seem oddly high to me. IIrc my hot start is like .068g and for colder 80*F it is maybe .098g? Your values seem like they might be flooding the engine when you first crank which could delay the startup. You can either reduce these values or you can hold the throttle kind of open (50-75% open) when cranking to see if that helps with starting. I would just reduce them and see what happens (its safe)

    Next the FA Multiplier table, I dont know how cold you are there. My "cold" here in FLorida is like 60*F or 80*F outside. At say 70*F I am using like 6.5 in the top row for the engine to fire up quickly. Yours says 1.8 and to the left of that is a 2.4. I would try doubling those (or work your way up to a double over time) to see if that helps.


    Everything in those tables is for grams of fuel, which is calibrated to your injector flow rate table. So if the injector flow table says 60lb/hr injectors and you have actually got 60lb/hr injectors it should be correct.

    The IAT vs IACV table is in there somewhere. Not sure what its called but it clearly marked IAT and Airflow (should be in mass of air). When I get my laptop back from school I will check but it could be a while.

  17. #17
    Senior Tuner kingtal0n's Avatar
    Join Date
    Feb 2017
    Location
    miami
    Posts
    1,799
    Here is a link to my tune file
    https://forum.hptuners.com/showthrea...l=1#post648378

    That isnt the final version but its close. Its been almost a year since I've opened a tune file to do anything.

    You can try compare view, and try my fuel pulse prime, FA cranking, etc... values for your setup if its a gasoline like mine,

    Try to find the table with IAT vs airflow, factory is like 12g/sec and I use roughly 20g/sec (almost double) to get the engine to fire up immediately.

    For hot refires it seems like the .058g in the first pulse mass is the key with the 20g/sec airflow for 100*F~ iat (underhood temps & Heat soaking)

    For cold starts it seems like the larger 6.25 to 7.25 top row values in the FA multiplier table were the key

    My engine fires up very fast, cold or hot, Almost immediately and I don't even have to hold the key, I just turn the key for a moment then let go, the engine kind of slows down for a split second almost like its going to stop and then VROOOOM!!! roars to life, thats how I Like it
    Last edited by kingtal0n; 01-25-2022 at 05:00 PM.

  18. #18
    Evidently the 2nd table is just an extension of the 1st according to the notes, and progression is down the 1st table and then to the 2nd and down it. The entries follow that pattern anyway. I'm guessing that the vertical axis is in engine cycles but does this mean cycles out of the 4 cycle engine sequence, camshaft rotations, crankshaft rotations or what? Nobody seems to be saying. The top is clearly temp. Cold start in my location is from about 50 degrees down to about 10 below but if I can get the area around 45 or so to work well I think I can get the rest.

    Presently my startup above 60 degrees is acceptable but not quite there yet so I will work on the next range below that (50, 32, 14) and with success will move upwards. I will just have to extrapolate downwards for now.

    My prime pulse masses (1st and 2nd) are copies and I had to raise them to get the engine to light off. But it sounds like maybe the balance is off between that and the FAx table with more fuel needed in FAx and less in PP. I will shift those and see what develops. For this test I am doubling FAx and cutting both prime pulses to 1/3. I'll let you know what it does.

    The engine seems oddly oversensitive to enrichment on cranking. Hopefully this will help.

    That should not account for the problems after initial light off though as it is Cranking FAx. I need help with the after-start enrichment and initial warm-up. But I can see where I might need to get the cranking fueling set first perhaps. Question: Assuming that there are basically 3 segments to the process, those being Cranking, After Start and Warmup,(and it looks like there could be 4 based on my datalog responses to tuning) how do they overlap and influence each other? No doubt the fully warmed up settings do influence pretty much everything else (VE table for instance) Does that hold true as you go back towards cranking? Does warmup influence Afterstart which then influences Cranking as well? Or are some such as Cranking completely independent? It would help to know that.

    I will try to do a compare with your file tomorrow and work up some revisions based on what I see there.

    I have found: [ECM] 12148 - Idle Startup Airflow Initial vs. ECT: Initial value for Friction Airflow, decay starts from this value.
    Upon further investigation and study what I'm seeing is that the idle air control seems to be based almost entirely on ECT. So that's what I've been looking at. Apparently the startup airflow and friction airflow settings can cause the after-start oscillations that I am seeing and others have had issues with this. So at least there is hope.

    I'd love to be able to approach this in an organized, logical manner. Unfortunately I don't have a guide that lays out how it all works, what everything does, and in what sequence. For a former engineer this is maddening of course but I see no real remedy except to muddle ahead.

    Jim

  19. #19
    Excellent results from that test, light off was immediate and brisk. Don't think I could ask for better. Thanks immensely for the help on that, it's been a problem from the first. Hopefully the result will be consistent and reliable. I will work to fine tune it, any suggestions are more than welcome.

    Next up, the oscillations on after-start, datalog attached. As I mentioned it seems others have solved this by adjusting the startup airflow vs ect and the decay rate to stretch out the decay. Sounds like maybe the decay is fast enough to drop off a little too quick letting the engine want to stall, then recovering, which sets up an oscillation cycle.

    So first thing I'm going to try is to just decrease the startup airflow decay (2172). Current value is .8 so I'm going to halve that to .4. If it's going to make a change that should do it.

    If my understanding is correct, a decay that works at 60 degrees might not continue to work as the temperature drops and more fuel is needed for a longer period of time. There may be some other control for that variable but I'm going to try this.

    Jim

    OK test results good but qualified rapid light off is consistent, oscillation is gone, but the engine died (lean) within about 3 seconds of starting. Datalog below. 012722a
    Attached Images Attached Images
    Attached Files Attached Files
    Last edited by Jim Blackwood; 01-27-2022 at 02:37 PM.

  20. #20
    Senior Tuner kingtal0n's Avatar
    Join Date
    Feb 2017
    Location
    miami
    Posts
    1,799
    The trick with cold situations is having the engine fully tuned in a warm situation first. Then all you do is add some air and fuel (multiplier tables, various) and its good.

    The VE Map needs to be smooth and you need to predict where the engine may run if it tries to oscillate or if the vacuum dips or whatever. Look carefully at my VE table try to make it smooth like that.

    An engine in general should run cold fairly well even without any of the cold-fuel adders, or almost so. Think of a carb engine that starts without a choke- its fine right? If the initial shot of fuel for cranking is large enough it should run fine 'as if it were warmed up already a little rough' within reason (below 20*F probably not?)
    The real issue is controlling the extra air (With a carb its harder to 'program' extra cold start air over some duration)

    I spend time removing fuel from all the cold areas as much as possible because its just a waste of fuel and it carbons up the engine anyways. The OEM Likes to run cold rich because it smooths out the engine behavior (less complaints from customers) but we are tuning the engine and have determined the roughness is fine, preference rough = economical (unless you literally do not want it to run rough when cold for some reason). I Think alot of people get stuck on following 'rules' or OEM-like protocols for tuning such as narrowband closed loop and cold start richness without actually thinking about what they are doing, the reason for those things.


    The primary table afaik for influence air on cold start is the table showing temp vs enrich multiply for In-gear vs Neutral. That is where the majority of your cold-adder air should come from.

    The rest of those tables like delay adders and so forth, you can add to them, the issue I found is when adjusting multiple tables simultanously you kinda lost track of which table is doing what. So I recommend just changing the 'main' tables first to get acceptable startup behavior and then FINE tuning those other tables over time after that.

    THose tables you have a screen shot of there are not the one I am suggesting you modify. Look for In-Gear Airflow vs temp and In Neutral airflow vs temp sort of table and increase the cold-area numbers alot. Once you see the effect they have you can dial it back down gradually to something nice.

    You should be logging IACV all the time to see what the ECU thinks it is doing. At idle IACV should be steady, not moving much. Slow it down in the PID section of IACV control. Depends on the cam, look at my tables for mild cam, its perfect.
    The IACV should sit warmed up with very low counts open, like 50 out of 200 or something like that. Like 1/4 to 1/6 of the total possible IACV valve I guess. You want it open just a little bit. Do not drill anything to get this it can usually be made up for in PCV flow and other small details.

    And just another reminder, if the fully warmed up engine tables aren't tuned well, if the VE table isn't smooth, if there is some kind of pressure leak on the intake manifold (do a pressure test IMO), if the PCV system isn't setup properly (You should have OEM pcv valve on one side and a hose attaching crankcase to post-air filter tube on the other side, not a breather) all of those things influence cold behavior AND warmed up behavior.

    I think if you tuned a stand-alone computer first, all of this would make way more sense. A stand-alone has a VE type of table, and then maybe 1 or 2 other tables that adjust for cold situations. It is very simple and that is kind of the key to cold starts is keeping it very simple, just use good warm values multiplied up a bit for the cold and wham its done. Cranking fuel (during cranking) is the only thing thats really empirical in nature and varies greatly between engines and swaps because of the myriad small details which influence injector on-time during cranking. For example the enormous voltage drop during cranking (9volts?) will cause injectors to become very lazy, it may need some ridiculous on-time value just to get a normal amount of fuel from them, like 3ms just to get 1ms of fuel out of one, this varies by injector design and quality as well. Then, the power of the starter and battery and rate of the engine acceleration during initial cranking is a factor in the cylinder fill (more air into cylinder due to faster initial cranking usually). If we examine ten different swaps with the same engine but each has different battery voltage, wiring and injectors its possible that none of them will use the similar cranking fuel settings.

    Keep trying and if you are still kinda stuck when I get my laptop back ill examine your files but rn I cant open anything so take screen shots pictures of whatever you want me to see
    Last edited by kingtal0n; 01-27-2022 at 06:16 PM.