Big Cams Have the Potential to Idle Like Stock

[cmitchell17]

I believe there is something in the GEN V control system that is sabotaging us. Something that is completely baffling me is how I can run my LT1 with a 228/236 cam under idle type loading conditions at 500 rpms and it run smooth as stock?

I used the dynamic testing tools to lock the TCC and lock it in 4th gear and basically idle up a hill at 500 rpms with a OLSD tune. I regularly do this to tune the high load/low rpm cells in the VVE table. However I logged this and you can see how I am at almost the exact same load on both MAF and MAP and around the same or lower rpms, only difference here is one is moving and the other is at 0 mph.

I am not sure I can even believe this, how is this possible with the huge cam (228/236 .635/.635 114-LSA) I have it is running this smooth, compare the left side of the graph where I am moving up a hill at 10 mph to the right where I am at 0 mph:
Capture.JPG

Look at the left side with its smooth RPM error and its constant flat line of spark advance along with the constant EQ error and wideband reading, then compare that to the right side with the timing bouncing all over the place and the jagged idle RPM error. How can this even be explained? Basically once the torque sources go from "axel" to "idle" the car starts running like crap.

I mean I expect to have diminished idle (and low rpm operation) quality with a cam, but how much to expect I am not sure. But according to what I see here and feel while driving you apparently don't have to sacrifice any driverability with a cam? But unfortunately the stupid GEN V control system with all the stuff going on we can't see and all the stuff that makes sense that we don't understand is apparently sabotaging our driveability and idle quality by throwing timing all over the place. As far as I know there is no difference in the 2 different loading scenarios on the engine, in fact it seems the load when moving would be more unsteady.

I really hope someone can explain this, or maybe in fact the DI and GEN V combustion system are just so good it can run with a cam like stock?

Log 97 Comparison of low rpms at idle vs under load while moving in high gear with TC locked.hpl

[Hondaeater]

Look at your log. What RPM and air mass is the trouble area?

[Hondaeater]

Look at your Virtual Torque table and remember the values in the same trouble area. Torque management>>Peak torque table should be about 70-100 more than vt.

[tjd1991]

Have you made any changes to the idle torque figures? If the low rpms aren’t an issue until your predicted torque source goes to idle then that may be why. It might run well at idle rpms in gear when predicted torque source is axle because you have tuned your vve and maf like that. I have read that the idle torque settings shouldn’t be adjusted until your are fairly dialed in airflow wise and still have idle issues. I would imagine a big cam makes less torque at idle so the ecu is going to try to command more spark to try to raise your idle torque but then the cam may make more torque than the idle tables allow. Probably more to it then that that I don’t understand yet.

[cmitchell17]

So I have since been fighting my idle trying everything I can to make it command decent timing. About the only thing that kind of works is to lower the VT at the idle cells in the negative and low spark tables by about 75-50 ft-lbs then steadily take out less as you get higher in spark. I am assuming this lets the ECM see that when it wants to idle with negative and really low spark its going to make a significant amount less in torque according to the VT tables we just decreased, so then it figures it can raise timing to reference the higher timing tables in the VT tables since we did'nt or didn't at all decrease torque in the higher timing VT tables.

That being said what I was trying to explain above is that there is something I am missing here or not getting the whole picture to. You have 2 different scenarios where from what I can tell the major output sensors of the engine are the same (MAP, MAF, rpms, EQ ratio, etc.) but under the 0 mph idle the timing goes all crazy and when I'm at a moving idle the timing stays constant. So my proposal is we theoretically should be able to get a better idle that we have. I too thought that with a big cam its just a fact of life and the idle is going to be rough and vibrate and surge and buck, but apparently not. Is there something I am missing about this? I can even notice when I have the TC locked and forced into high gear (4th gear) and roll along everything will be extremely smooth, but If I let off the gas a little too much I think the torque source switches back to idle and the driveshaft and driveline start bouncing around and the pinion backlash in the differential starts clunking, really fast in the probably 20hz range, when the timing goes all crazy.

Would it be possible to maybe make a custom OS that would bypass idle torque control? I have also tried to extend the cranking and startup idle to try and use those permanently but it doesn't seem to work.

[mbray01]

With that cam, you shouldn't have to touch the torque tables, if you do, it would be very minimal. Your issue likely lies in the airflow models not being properly calibrated. Specifically the vve. Your torque is directly derived from the vve model. If your maf only, its not gonna be happy. Getting your airflow model 100% correct is the only correct way to get these things to function like they are supposed to.

[cmitchell17]

I think I may have to get an additional wideband. I have spent months trying to my VVE better, it seems to get to around +/-2-3% at idle, then the next day I will check it again when the weather is different and it will go to 10%. I probably have over 100 logs trying to correct VVE.

I think I also am going to have to find a way to override timing so I can get proper combustion to get accurate wideband feedback. That also makes me think if proper combustion is just not possible with the cam so I would never get accurate feedback? meaning I would always be showing a false lean condition because of excess unburnt oxygen in the exhaust? My idle misfire counts are around 10-20 range on different cylinder histories at idle. I have found that as a workaround I can lock the car into high gear (4th) with the TCC locked and idle it up a hill while moving, it kicks it out of "idle" torque source into "axel" torque source and gives a consistent and flat 30+ degrees of timing and the engine works extremely smooth like stock.

I have also tried tuning with LTFT, which I find way more consistent than wideband feedback, but its still isn't enough, I can get it to under +/-5%, but then out of nowhere the next day its back to 10%? Which I guess I should question which LTFTs I am using? the right or left bank? or is which PID is the average one?

I actually wouldn't have a problem if the PCM just has to idle at 5-10 degrees advance, I am just trying to make it consistently idle at one place. The timing swings 10-15 degrees at idle (that happen within fractions of a second) is what is messing up the drivability, causing excessive vibration, and causing the surging and bucking feeling when slowly rolling forward. You would think you could just go into the proportional gain constants for the immediate torque and reduce those at lower rpm errors and maybe increase the predicted constants to make up for it and it would smooth out the timing command. Unfortunately it dosen't work.

With the Camaro OS we are missing a lot of the parameters that the Corvette OS has too, I am not sure this would make a difference though, but I really would like to increase the immediate control deadband:
Capture.JPG

[Jonblarc7]

You still removing you bank 2 position 1 O2 sensor and putting your wideband in there to tune VVE?

If so have you thought about welding it into your Y pipe and still retaining both your O2 senors

[cmitchell17]

You still removing you bank 2 position 1 O2 sensor and putting your wideband in there to tune VVE?

If so have you thought about welding it into your Y pipe and still retaining both your O2 senors

I am trying to find someone to weld in a bung. It seems like it may just be better to put the narrowband back in there and try to tune the VVE again with LTFTs, it seemed way more consistent with trims and you wouldn't get random swings lean or rich in the histogram when tuning with LTFTs.

[cmitchell17]

I made some adjustments to idle torque reserve, external load, and MIN spark as Hondaeater suggested and it helped get the average timing back up into the 15-20 degree range using stock virtual torque tables. However, the jaggedy rough spark is still there.

I really want to put in more work to try to get VVE better, but I just really feel like I am chasing my tail, especially at idle. I use the scan tool to reset the LTFT's each time I log the histogram to correct VVE. I also notice that sometimes right when I start it the fuel trims will start off way lean or way rich and I know something is off, so I have to stop scanning and restart scanning and it will magically go to a more reasonable fuel trim value. I do think I am getting some convergence, meaning if I see say around 10% lean one log, I copy and paste % by half, then I see about 6-8% lean the next log. However, idle still seems to do weird things and I believe with a cam we just can't trust the combustion to be proper enough at idle to infer airmass by reading the oxygen exhaust content and inferring the AFR from that. It makes it even worse when the engine wants to pull timing at idle and the rpms and timing are spiking all over the place and won't stabilize.

As an example, this is a log from when I had the stock cam, this is what the timing signature I think needs to look like in order to have good idle quality and driveability:
Capture.JPG

But this is what I am getting with my aftermarket cam when torque source is idle:
Capture.JPG

But when I do the lock the TCC and lock in 4th gear and let the car roll at idle I get smooth timing again and no shaking or vibration or surging or bucking:
Capture.JPG

On the last picture where I was doing the rolling around at idle RPM speeds with the TCC locked in 4th gear, you can see it is having some clear fuel trim feedback issues:
Capture.JPG
So you can see my normal 0mph idle cells circled in blue, it looks like I am getting pretty close (I am using LTFT+STFT equation for feedback on histogram), but then circled in red is the areas where I am doing the rolling around in 4th gear with the TCC locked and its showing extremely lean.

I am in OLSD, this is my VVE table, it is linear in this region at low rpms and high MAP:
Capture.JPG

So I don't think I can even trust the PCM fuel trim feedbacks to filter all the transients and noise to give me correct airflow data based on O2 feedback. The fuel trims do seem a lot more stable than the wideband, which can be all over the place especially when you get low cell counts. I just think at low rpms with big cams we get maybe a false lean condition?

B Tune 2 (my os).hptLog 100 OLSD more adjustements External Load and Idle Reserve and Min Spark 4th Gear TCC Locked .hpl

[Hondaeater]

The 650 idle seems very low for a 228/236 .635/.635 114-LSA cam. It should idle happier at 750-775.

[cmitchell17]

The 650 idle seems very low for a 228/236 .635/.635 114-LSA cam. It should idle happier at 750-775.

So I've tried a idle that high, and even higher. Unfortunately it still does the same behavior. The highest I have tried taking it is 950 until it started to get smoother and less choppy bucking and surging. However once you get past 750 your car then sounds like it's broke worse.

[cmitchell17]

It's looking like I may just have to give up on this one. I am almost positive the potential is there, maybe we will get a custom os patch or maybe some more parameters that can break it so the pcm no longer switches the torque control to idle source. I also am going to put in a ticket with HP Tuners and try to request the same parameters for the Camaro LT1 as the Corvette and Trucks have. I wish I could try flashing in their os but you would have to register that file with HP Tuners.

The potential is there and that's evidenced by the fact it's smooth when crawling at low idle rpm with the TC locked in 4th gear and another thing is the engine is also smooth during initial start up before it's warmed up and within about a min or 2 starts jumping timing around then you start getting vibration and surging.

[kingtal0n]

did you try using the scanner to lock the timing or command richer air fuel ratios manually during these situations while simultaneously locking the airflow in the scanner,

in other words, if you control air, fuel, timing, all the be a set or steady pace, it might narrow down which of these is related more than others

[cmitchell17]

did you try using the scanner to lock the timing or command richer air fuel ratios manually during these situations while simultaneously locking the airflow in the scanner,

in other words, if you control air, fuel, timing, all the be a set or steady pace, it might narrow down which of these is related more than others

I really wish I could use the scanner tools, but the only thing that works on them is the idle speed rpms control. The timing override does not work. I don't think GEN V stuff has has AFR override, I am sure this is becauase the EPA said no you can't have that because people could use it to bypass and create harmful emissions that are guaranteed to kill millions.

I have heard if the airflow model (MAF and VVE), and thus torque model is off, the PCM will not allow timing override, but I honestly can't get VVE better than it is.