Jump to content

Issue with oscillating/bouncing idle when at operating temps


jigga009

Recommended Posts

For quite a while I've had an issue with an oscillating idle especially during the winter months with a built Subaru application. We first came across it on my V88 ViPec ECU, but could not do too much about it given that I was out of tables on the ECU to bring online certain features such as idle ignition control. Since moving to the G4+, I have been able to activate the Idle Ignition table, which seems to have helped things out especially on warmup from cold, but I'm finding that when up to temperature, the oscillation seems to return, and the ECU appears to be bouncing into parts of the ignition map (i.e. at 2500rpm) that it shouldn't really be doing, given that target idle is more like 1200rpm, which is contributing to the oscillation.  As is now, when sitting at traffic lights, other motorists could easily be forgiven for thinking that I was trying to instigate a race, given how it sounds as if I am constantly blipping the throttle.

This isn't much of an issue at all while the engine is warming up, but tends to become a lot more prevalent once at operating temps.

I have taken a log and map for your perusal. If any of our resident experts knows what I need to be adjusting, I would very much be grateful. 

Map 1.1.pclr

Map 1.1 Idle log - Log 2018-01-20 9;54;42 pm.llg

Link to comment
Share on other sites

The ecu is not entering idle control or idle ignition control because the base position table is commanding the valve open too far so the engine settles above the RPM lockout.

Change all the settings in red below to match this example.  It will likely still not work correctly but hopefully will be getting closer and we may get some more useful data from a log.  Do another log for us with these settings.

BCOV7k2.png

Link to comment
Share on other sites

Adamw - I ran the car with the exact changes you mentioned, and it is MUCH better than before. Problem I find is that when I let the car idle for more than a minute or two at longer traffic lights, it stalls. I have a couple of logs where you can see this occur. 

- In Log 1, I let the car idle, and tried a few things such as turning the AC on to see what effect that has on the idle.

- The stalling issues can be seen in Log 2.

Glad to see we are on the right path though!

I have included the map again with the changes you suggested.

LINKECU FORUM TestMAP - 1.0.pclr

LinkECU forum testmap idle log with and without AC log - TestLog 2018-01-21 12;51;36 pm.llg

LinkECU forum testmap idle log with and without AC log 2 - TestLog 2018-01-21 12;51;36 pm.llg

Link to comment
Share on other sites

You are not getting into idle ignition control as your have it set for below 700rpm, change this to something like 1500 for now.

I would, turn off or unplug you ISCV and then set the idle screw on your throttle so it can idle.

Then set your idle ignition up, currently you have it set to enable below 700rpm, I would raise this to 1500rpm  then tweak the table, it is not uncommon to have zero degrees in the centre (0)  and then negative degrees towards the left hand side, but all cars are different so you will need to tweak this, once you can rev the engine and it falls back quickly to the centre (0) 

Then you can turn on your ISCV back on and set that up, but even with just idle ignition control it should be fairly stable.

Link to comment
Share on other sites

Hi Clint - According to the Help section of the G4+, it mentions that the Idle RPM Lockout is in addition to the idle target rpm, at which point the closed loop idle control would be disabled. If I'm reading it right, it means that on my map, the 700rpm lockout would mean that closed loop idle control would start kicking in as the RPM drop below about 2000rpm.

 

 

 

 

Link to comment
Share on other sites

1 hour ago, jigga009 said:

Hi Clint - According to the Help section of the G4+, it mentions that the Idle RPM Lockout is in addition to the idle target rpm, at which point the closed loop idle control would be disabled. If I'm reading it right, it means that on my map, the 700rpm lockout would mean that closed loop idle control would start kicking in as the RPM drop below about 2000rpm.

 

 

 

 

Look at Ignition - Idle Ignition Control its the lockout there we are talking about.

You can add Idle ignition Control to your logging view to watch when this becomes active.

Link to comment
Share on other sites

1 hour ago, ClintBHP said:

Look at Ignition - Idle Ignition Control its the lockout there we are talking about.

You can add Idle ignition Control to your logging view to watch when this becomes active.

Ahh ! I see what you mean; my mistake! I'll get that taken care of :)

Now that the idle ignition table will be kicking in, I guess I will now be able to tweak the table properly.

What is typical from your experience as far as timing jumps as you go from 0 outwards? As is now, I have jumps of 2 degrees... What would be considered excessive?

 

Link to comment
Share on other sites

I can’t remember what it looked like and I am away from my computer.

Certainly it will be easier to tune when you get it stabilised, I am sure as I can be that  it’s just a tuning issue and not a problem with the engine if that’s what you are looking for.

It’s all about watching them logs and watch what it’s doing when it stalls, does it go too lean is there not enough timing, the logs will show.

Link to comment
Share on other sites

20 minutes ago, ClintBHP said:

I can’t remember what it looked like and I am away from my computer.

Certainly it will be easier to tune when you get it stabilised, I am sure as I can be that  it’s just a tuning issue and not a problem with the engine if that’s what you are looking for.

It’s all about watching them logs and watch what it’s doing when it stalls, does it go too lean is there not enough timing, the logs will show.

Yes I don't think it is an engine problem either, as it is a very fresh rebuild still in the break in process. I had the exact same idle issue before the engine refresh, but was unable to do anything about it at the time because the ViPec V88 that ran the engine had no room on its processor to enable the idle ignition table. With the new engine and the new G4+, I am only now enabling the idle ignition table for the first time and trying to get the engine to finally idle properly.

My question as to typical jumps between cells was more because I'm no tuner. I can read help guides and make minor changes to the existing map, but I rely on the instruction of those that know what they are doing when it comes to making those specific changes. I just like to have a guide as to what is "normal" (e.g. using negative timing figures if needed...something I did not know until you mentioned it) and what is not normal (e.g. I am guessing that I don't want to be increasing and decreasing timing in increments of 15 degrees as I go from 0 outwards on both sides of the idle ignition table?).

I will report back with more logs based on the changes you suggested making.

Link to comment
Share on other sites

I personally would leave the idle ignition disabled for now until you get the idle valve working reasonably well - otherwise it can mask problems with the idle valve tuning.  That is why I didnt ask you to change the RPM lockout originally.

The stalling is caused when fan #1 turns on, this is because your Fan#2 settings are wrong so that the ECU has already applied the "fan step" at 80°C, long before Fan 1 turns on.  (I also note you have no fan set to turn on with AC but we will leave that for now).

Try changing all red values again as per the screenshot below.  Leave idle ign off for now and do another log.  

7dmETXv.png

Link to comment
Share on other sites

40 minutes ago, Adamw said:

I personally would leave the idle ignition disabled for now until you get the idle valve working reasonably well - otherwise it can mask problems with the idle valve tuning.  That is why I didnt ask you to change the RPM lockout originally.

The stalling is caused when fan #1 turns on, this is because your Fan#2 settings are wrong so that the ECU has already applied the "fan step" at 80°C, long before Fan 1 turns on.  (I also note you have no fan set to turn on with AC but we will leave that for now).

Try changing all red values again as per the screenshot below.  Leave idle ign off for now and do another log.  

7dmETXv.png

Thanks Adam, will do. I will set the idle ignition lockout back to the previous value which kept it off. I know the idle is not perfect yet, but compared to where it was before, it's a night and day difference so far. The map was ported over from my Vipec V88 where it was controlling the fans as they should.  It seems there are a few settings that get lost in the translation when making the upgrade. I will report back with logs and the updated map soon.

Thanks everyone for your contributions so far. I'm extremely grateful!

Link to comment
Share on other sites

Ok, carried out the changes as suggested above and gave things a try. Getting better, although it still nearly stalled once or twice with the AC on, but never completely dying like it did last time. I have included 3 logs. On log 2, I experimented breifly with turning on the rear window defrost, and it didn't seem to like that much at all. Also played with the AC as well - better than last time, but still drags things down. I'll let the logs do the talking, so as not cause any bias.

Linkforum ECU map log for rev 1.1.llg

Linkforum ECU map log 3 for rev 1.1.llg

Linkforum ECU map log 2 for rev 1.1.llg

LINKECU FORUM TestMAP - 1.1.pclr

Link to comment
Share on other sites

Ok, idle valve operation is looking reasonable now.  A little more AC step is needed.  We will also bump proportional, and add anti stall.  Proportional can possibly go higher than I show below to speed up response but try these settings first.  You can also try the idle ignition settings below.

yR66lWk.png

Link to comment
Share on other sites

Ok, added the tweaks, and things are on the up and up, but still the engine wants to stall when I start turning things on inside the cabin, especially the rear defroster when the HVAC fan is turned up and windshield dehumidifier setting turned on. Things get a bit worse when the seat heaters come into the equation. These can be seen in the logs.

The log files seem to be getting a bit large to attach to my posts, so I have uploaded them to dropbox in order to keep the map and logs together:

LinkECU Testmap 1.2 with logs

Link to comment
Share on other sites

You are quite rich there, have you tried tuning the table a little, I am assuming your on petrol/gasoline as your have 14.7 as Stoich, your dropping close to the 11's I would be surprised if your plugs are not black and misfiring.

I also see that the main change in the map is injector deadtime, this looks very wrong as the idea is that as the battery voltage drops the speed of the injector changes and requires more opening time to keep the same AFR, however as soon as your deadtime changes you get dumped into 0.7 Lambda.

I would change the deadtime so 12, 13, 14 & 15 volts are the same at say 0.7ms (doesnt matter for now if this is correct) then I would tune the map so you are targeting nearer 0.95 Lambda, then as the voltage drops trim the deadtime map so Lambda stays the same as when you tuned it.

 

Screen Shot 2018-01-23 at 21.20.08.png

Link to comment
Share on other sites

49 minutes ago, ClintBHP said:

You are quite rich there, have you tried tuning the table a little, I am assuming your on petrol/gasoline as your have 14.7 as Stoich, your dropping close to the 11's I would be surprised if your plugs are not black and misfiring.

I also see that the main change in the map is injector deadtime, this looks very wrong as the idea is that as the battery voltage drops the speed of the injector changes and requires more opening time to keep the same AFR, however as soon as your deadtime changes you get dumped into 0.7 Lambda.

I would change the deadtime so 12, 13, 14 & 15 volts are the same at say 0.7ms (doesnt matter for now if this is correct) then I would tune the map so you are targeting nearer 0.95 Lambda, then as the voltage drops trim the deadtime map so Lambda stays the same as when you tuned it.

 

Screen Shot 2018-01-23 at 21.20.08.png

Hey Clint! Thanks for chiming in. Yes, the car runs exclusively on pump fuel. The engine has some aggressive headwork and relatively radical cams, and seems to need a richer AFR in order to maintain a decent idle when you start really turning on the electrical loads which put more load on the engine...At least that is my understanding of it when the car was tuned. I know that it has traditionally exhibited this kind of behaviour, but I never knew enough about tuning to know why this was the case. It seems to jive with other articles I've read on tuning engines with hot cams, so never gave too much thought to it. If it were possible to maintain the same idle AFR just as my stock 2.5RS does when the electrical loads come on, that would be great. Thought it was due to the lumpy cams. 

 

Link to comment
Share on other sites

On 1/23/2018 at 1:18 AM, Adamw said:

Ok, idle valve operation is looking reasonable now.  A little more AC step is needed.  We will also bump proportional, and add anti stall.  Proportional can possibly go higher than I show below to speed up response but try these settings first.  You can also try the idle ignition settings below.

yR66lWk.png

Any thoughts on the most recent logs based on these suggested changes? Thanks so much for your help so far! :D 

Logs for Testmap 1.2 as well as Copy of Testmap 1.2

Link to comment
Share on other sites

19 hours ago, ClintBHP said:

You are quite rich there, have you tried tuning the table a little, I am assuming your on petrol/gasoline as your have 14.7 as Stoich, your dropping close to the 11's I would be surprised if your plugs are not black and misfiring.

I also see that the main change in the map is injector deadtime, this looks very wrong as the idea is that as the battery voltage drops the speed of the injector changes and requires more opening time to keep the same AFR, however as soon as your deadtime changes you get dumped into 0.7 Lambda.

I would change the deadtime so 12, 13, 14 & 15 volts are the same at say 0.7ms (doesnt matter for now if this is correct) then I would tune the map so you are targeting nearer 0.95 Lambda, then as the voltage drops trim the deadtime map so Lambda stays the same as when you tuned it.

 

Screen Shot 2018-01-23 at 21.20.08.png

Clint - a few questions, if you wouldn't mind:

Should the expectation be for the engine to be able to maintain the same AFR at idle when electrical loads are off and when they are also on (e.g. rear defroster and HVAC fans)?

Also, if I adjust some of the dead time settings, will that throw off my main fuel map in any way? In other words, would I need to get the car retuned again from scratch once I start playing with the fuel injector deadtimes as I turn on different electrical loads? 

Since the car idles near stoich at idle already without any of the electronic doodads on inside the car, would one just be adjusting the appropriate deadtime voltage box when the electronic accessories is turned on?

Should I be looking to ensure that idle RPM remains rock solid with all of the accessories and electrical loads coming on and off before looking to adjust the injector deadtimes as you suggested?

Thanks so much for your input!

 

Link to comment
Share on other sites

Short answer:

All injectors take an amount of time to open (and close) that needs to be known, this is commonly somewhere between 0.7 and 1ms at 14v. 

 

So so from a fast injector to a slow injector that’s 0.4ms difference.

At idle when you may only have a 2ms plulse for ideal idle that 0.4v is 20% but higher up the Rev range it’s far less so will affect ARF much less.

you should be able to maintain the same AFR no matter what the battery voltage, and that’s the reason your idle is bad your AFR is dropping into plug fouling territory.

 

 

Link to comment
Share on other sites

12 minutes ago, ClintBHP said:

Short answer:

All injectors take an amount of time to open (and close) that needs to be known, this is commonly somewhere between 0.7 and 1ms at 14v. 

 

So so from a fast injector to a slow injector that’s 0.4ms difference.

At idle when you may only have a 2ms plulse for ideal idle that 0.4v is 20% but higher up the Rev range it’s far less so will affect ARF much less.

you should be able to maintain the same AFR no matter what the battery voltage, and that’s the reason your idle is bad your AFR is dropping into plug fouling territory.

 

 

ClintBHP - Thanks for the response. Since adjusting these settings will not have an effect on the way the car drives (e.g. in boost), I will adjust the latencies until I can get the idle AFR to stay constant as possible as I turn on the electronics in the car, and report back with a log. Again, I'm grateful for your input. Since the car already idles pretty close to stoich when everything is off, I think it will be boxes that the ECU dips into once I start turning electricals on inside the car that I will focus on.

Link to comment
Share on other sites

22 hours ago, ClintBHP said:

Short answer:

All injectors take an amount of time to open (and close) that needs to be known, this is commonly somewhere between 0.7 and 1ms at 14v. 

 

So so from a fast injector to a slow injector that’s 0.4ms difference.

At idle when you may only have a 2ms plulse for ideal idle that 0.4v is 20% but higher up the Rev range it’s far less so will affect ARF much less.

you should be able to maintain the same AFR no matter what the battery voltage, and that’s the reason your idle is bad your AFR is dropping into plug fouling territory.

 

 

Clint, made some adjustments to the deadtime values:

Logs and the map that produced the logs:

https://www.dropbox.com/sh/4uoztacmhkahc6o/AABt-Dgv-T9UGTlQJgb6EnTVa?dl=0

I did notice after using these settings that the engine had a tendency to stall when AC compressor kicked in. I did not try the rear defroster with these settings as I was sure that the engine would stall. 

I also noticed that the car was running noticeably leaner when I tried to drive away from the parking spot where I was making he adjustments. So much so that I had to revert to the previous map in order to make it home, as I did not want to damage anything on this fresh engine.

What has me perplexed about this is that the previous deadtime settings were from Link, included in the basemap for the Ver 7 WRX, and I would think should have been on the money for Denso Subaru OEM stock-type top feed injectors (a derivative of such I am running on the car as primary injectors). Is it that Link's settings were completely off on the basemap they include?

Link to comment
Share on other sites

I said above  "I would change the deadtime so 12, 13, 14 & 15 volts are the same at say 0.7ms"

Your deadtimes are now:  0.741    0.591    0.496    0.49    0.489

Change them so they are all the same and you wont get the massive spikes in AFR you are seeing in your log, can you see where the deadtime changes the AFR drops massively.

You then trim these settings so the AFR stays the same as your target as the voltage changes. You can see what box its taking the reading from as it highlights it in pink.

By the way this isn't all of your issue but its not helping.


 

 

Screen Shot 2018-01-25 at 18.46.36.png

Screen Shot 2018-01-25 at 18.41.02.png

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...