Random Tuning thoughts, Cam Timing and Closed Loop…

[Brendan]

So, I’ve learned some things over the course of the last year or so, and have some things I’m still thinking on, but figured I might was well throw them out here for discussion or comment, or just plain bench racing… Some of these are somewhat anecdotal and what I’ve gathered from conversations or reading stuff here and there.. Please correct me if I’m wrong on anything below…

On cam timing: Last year I had retarded my intake cam and advanced my exhaust cam to gain more P-V clearance while running Schrick’s with stock pistons. An unintended consequence: This reduces overlap, which increases cylinder pressure, which can make the engine more susceptible to detonation… Not sure how much of a factor this is on its own.

On running closed loop: I really want to run closed loop, I have a good sensor, have the ability to set an authority limit, but I haven’t done it yet. I had the discussion with my tuner on this, his opinion: The sensors are accurate enough when new and calibrated, but don’t do it (At WOT). You’re betting your motor on one sensor. When WBO2 sensors fail, they usually start reading rich, and want to lean out the engine. If you had a 10% authority limit, and your car was running at 12.5 AFR at WOT, and the sensor started leaning things out, it could go 13.75 AFR (if I’m understanding how authority limits work) which could be enough to blow up an engine… Lots of people are running closed loop successfully, but this is food for thought. Probably a good idea to calibrate and replace sensors regularly. Or, if your ECU allows it, allow it a 0% authority to lean things out, and a 10-15% authority to richen the mixture… And, check your corrections regularly to make sure something hasn’t changed unexpectedly. I dunno, maybe I'm worrying unnecessarily.

On timing, race fuels, and detonation: The more you advance timing, the cooler an engine will run (Per this tuner’s experience, and some of my reading, but my guess is that this has limits). The reason for this is that during normal combustion that finishes completely within the cylinder, there is a layer of boundary gasses that insulate the block from the flame front from combustion. As you retard timing, more and more of the combustion cycle is finishing as the mixture is expelled through the exhaust valves and into the exhaust, and you are getting a less efficient burn. As this happens, you lose the insulating effect of these boundary gasses, and start transferring more and more heat to the exhaust valves, exhaust ports, and exhaust, which will translate into higher overall engine temperatures starting on the exhaust side of the block. His opinion was that retarding timing to avoid knock isn’t a great solution because of this heat issue (Knock, retard timing, more heat, more knock), and if you run more advanced timing with a higher octane fuel, you’ll be better off and have a cooler running engine…

Ok, that’s probably enough for one morning, and I’m running low on coffee… Discuss…

[Ironhead]

Closed loop: I have been running it for years, and I don't think it could damage the engine if the sensor failed. To begin with, you still tune the car to get the base maps as accurate as possible. The idea is for the WBO2 to do as little "correction" as possible. Also, at least with Motec, the ECU will only allow the WBO2 to do relatively minor corrections....maybe 5%, something like that. If the AFR from the base map gets more than 5% off what the Lambda map says the system is shooting for, you get an error, the closed loop shuts off, and the car runs from the base fuel map.

So the closed loop can only provide "fine tuning" of the fueling....it cannot subtract enough fuel to endanger the engine. Otherwise, you could theoretically run an engine without even tuning the fuel maps. Just plug in closed loop WBO2 and drive the car. But none of the systems allow this.

Since you do rely on the WBO2 sensor to do initial tuning and establish the base fuel maps, I guess you could get into trouble if you tuned with a bad O2 sensor. But this is not unique to running a closed loop system...tuning is always done using WBO2 and a bad sensor on a dyno or wherever would cause the same problems.

[Brendan]

Closed loop: I have been running it for years, and I don't think it could damage the engine if the sensor failed. To begin with, you still tune the car to get the base maps as accurate as possible. The idea is for the WBO2 to do as little "correction" as possible. Also, at least with Motec, the ECU will only allow the WBO2 to do relatively minor corrections....maybe 5%, something like that. If the AFR from the base map gets more than 5% off what the Lambda map says the system is shooting for, you get an error, the closed loop shuts off, and the car runs from the base fuel map.

So the closed loop can only provide "fine tuning" of the fueling....it cannot subtract enough fuel to endanger the engine. Otherwise, you could theoretically run an engine without even tuning the fuel maps. Just plug in closed loop WBO2 and drive the car. But none of the systems allow this.

Since you do rely on the WBO2 sensor to do initial tuning and establish the base fuel maps, I guess you could get into trouble if you tuned with a bad O2 sensor. But this is not unique to running a closed loop system...tuning is always done using WBO2 and a bad sensor on a dyno or wherever would cause the same problems.

So here's a scenario. You tune your car on a good, solid fuel map to run at 12.7 AFR at WOT and then set a 5% authority (adjustment) limit on the WBO2 sensor (Usually, you can configure that authority limit). The sensor starts to fail, and starts reading rich, but not enough to throw an error code (If that's even an option). At WOT, the ecu leans the car out to 13.3 AFR (Just less than a 5% difference). Could that cause an issue? If your car is getting close to the detonation threshold, then maybe it could...

Basically, I agree with you that it can be done succesfully, and I'm probably being a little alarmist, but it could cause a problem under the right circumstances. Like you said: probably a good way to do it is to set the Authority limit very low, and then throw a warning light, or fall back to the map as soon as you go outside the limit... Coupled with that, make sure that your WBO2 sensor is a regular maintenance and calibration item, and that you are regularly looking at your O2 corrections to see if it looks like something is going bad...

I think the reason I brought it up is there are plenty of people out here who are just going to fire up their MAXX, turn on closed loop, and think everything will be fine because they haven't ever thought about it... And, this was also a recommendation by someone who makes his living doing tuning day in and day out, everything from MAXX and Megasquirt, through Vipec, Motec and Pectel...

(Also, aren't you running 100 Octane? I just switched myself which gives you much more of a safety margin)

[qwerte]

I think the reason I brought it up is there are plenty of people out here who are just going to fire up their MAXX, turn on closed loop, and think everything will be fine because they haven't ever thought about it... And, this was also a recommendation by someone who makes his living doing tuning day in and day out, everything from MAXX and Megasquirt, through Vipec, Motec and Pectel...

I have been watching this thread to get some more feedback as I am one of those people running MAXX who turned closed loop on and started driving and tuning. My intention is to leave closed loop on, but to also do significant amounts of logging and fuel adjustments so that I can reduce the allowed fuel adjustments to as minimal as possible. As others had mentioned, the MAXX setup with revert to the open loop control once limits are hit. I may be naive, but I hope I have a large enough margin of safety with regards to my timing to prevent detonation if at WOT my AFR goes from 12.8 to 13.5. Admittedly, my assumption is not well founded and Brendan, you have successfully planted the seed of doubt about whether I am approaching this properly.

I am all new to this stuff, so I am just watching others discuss the issue so that I can further learn.

[Brendan]

If you're going to run closed loop, just make sure you understand how everything works together and what the limitations are. I would:

Get good fuel maps to start
Calibrate the sensor frequently (If needed)
Replace sensor yearly or every other year
Limit the Authority, and hopefully have it fall back to open loop if an error occurs or if it goes outside a certain range
Keep an eye on your corrections. If you are only correcting a couple %, you're probably fine. If corrections suddenly change, you want to understand why.

How much of a difference this all makes depends on how mild or wild your motor is. If you're close to stock with a conservative spark map, you probably don't have to worry about it. My car is a track car that gets the crap flogged out of it on a racetrack with ambient temps that can reach 110 degrees in the summer time, and it is pretty substantially upgraded for a 2.3L (To the tune of approximately 70whp more than stock...)

And admittedly, I'm a worrier right now because I spun a bearing due to detonation last year (Even though I don't think it was due to running lean).

(It'd be great if someone like Neel @ Apex could chime in on this, and put my amateur hour ramblings to rest )

[xworks]

Good discussion. Do modern production engine management systems run full closed loop at wide open throttle?

[///Mflossin]

I love the closed loop MAXX. My car runs great and I would recommend it to anybody that wants to stick with the motronic or can't afford a full up stand-alone.

[Ironhead]

So here's a scenario. You tune your car on a good, solid fuel map to run at 12.7 AFR at WOT and then set a 5% authority (adjustment) limit on the WBO2 sensor (Usually, you can configure that authority limit). The sensor starts to fail, and starts reading rich, but not enough to throw an error code (If that's even an option). At WOT, the ecu leans the car out to 13.3 AFR (Just less than a 5% difference). Could that cause an issue? If your car is getting close to the detonation threshold, then maybe it could...

I guess if your car is on the absolute ragged edge of detonation, a 5% leaning of the fueling could cause a problem at WOT....maybe. But how likely is this scenario compared to the myriad of other things that might kill an engine? A failed or failing fuel pump or FPR that causes a sudden drop in fuel pressure is in my mind 100X more likely to kill an engine, and how many people run a fuel pressure sensor with a (highly visible) warning light that goes on if fuel pressure drops? I'm not saying your concerns are wrong or invalid...I'm just saying they need to be kept in perspective. In my case, if my engine dies it will almost certainly be because I suck as a driver and missing a downshift.....

If you recently lost an engine to detonation, I 100% understand the "paranoia" (hope you don't mind that description).....but again, perspective is important. If you have not done so, I would install BRIGHT highly visible warning lights for oil pressure (30 PSI threshold), fuel pressure (35 PSI), and temperature (mine goes on at 210 degrees). Unless you are positive what circumstances killed your engine, assumptions are very dangerous... You might focus on the wrong issue.

Again, you have to keep in mind what closed loop control is intended to do. It is not a replacement for thorough tuning, it just gives the EMS some capability to adapt to variations in running conditions, wear on components, etc. I always turn off the closed loop when "road tuning" the car, and I cannot even detect any difference in how the car runs with it off. In fact, the AFR graphs are much smoother with closed loop turned off, because the system is not constantly tweeking the fueling and making minor corrections.

And yes, I always run pure 100 octane on the track. To me the insurance is well worth the extra cost per gallon.

[Ironhead]

Good discussion. Do modern production engine management systems run full closed loop at wide open throttle?

I think they do...in fact I am almost positive they do. One of the reasons modern cars get such good gas mileage is that they run very lean everywhere they can get away with it, and incorporate sophisticated knock sensors to keep things running right on the "edge".

I too would like to hear Neel's opinion on this topic.

[Brendan]

Good discussion. Do modern production engine management systems run full closed loop at wide open throttle?

I know they never used to in the past, but I wouldn't be surprised if they do now. OEM's control strategies and fail safes are much better than we have at our disposal though. Knock sensors with ability to dump extra fuel and retard timing if knock occurs, knock sensors that are specifically calibrated to the knock frequency for the engine in question through destructive testing, dual lambda sensors, limp home modes, etc, etc. Ion sensing is even starting to be used instead of audio knock detection, and cylinder pressure sensors are probably on the horizon if not here already...

If you recently lost an engine to detonation, I 100% understand the "paranoia" (hope you don't mind that description).....but again, perspective is important. If you have not done so, I would install BRIGHT highly visible warning lights for oil pressure (30 PSI threshold), fuel pressure (35 PSI), and temperature (mine goes on at 210 degrees). Unless you are positive what circumstances killed your engine, assumptions are very dangerous... You might focus on the wrong issue.

It is definitely paranoia To address my blown engine for those who are curious: I have more warning lights than I know what to do with (I have a full programmable logger dash in my car). I had oil pressure, oil temp, water temp, and fuel pressure all data logged to review against RPM and GPS location on the track after the fact. I was missing electrodes on two spark plugs which points to detonation. In my case, I think it was a combination of fuel quality (I accidentally added a couple gallons of 91), the wrong spark plugs (X5DC instead of colder), 105 degree ambient temp, high oil temperatures due to oil cooler ducting falling out of the way, cam timing as indicated in my first post in this thread, and probably overall power levels were approaching inappropriate for street fuel (219 whp out of a 2.3 L, running revs up to 8K). I was fine for 3 full sessions throughout the day, I suspect my problem started due to heat soak in the mid afternoon after a day of tracking the car, and a bunch of small things added up. Because I don't think it was a Rich/Lean issue, I don't think closed / open loop would have made a damn bit of difference in my case.... (This year, I'm running better cam timing, XR2CS plugs, 100 Octane, and am trying to get a knock monitor up, running, and tested...)

Back to the issue at hand - I may be paranoid, but I was ready to enable closed loop across the board when someone who tunes professionally advised against it for the reasons stated above. To me, that carries more weight than from weekend warriors like me... That's why having opinions of people with a lot of professional experience would be great (Any other tuners out there???).

Ultimately, I think closed loop should work, if you know what it's limitations are....

[Jake]

In my case, I think it was a combination of fuel quality (I accidentally added a couple gallons of 91), the wrong spark plugs (X5DC instead of colder), 105 degree ambient temp, high oil temperatures due to oil cooler ducting falling out of the way, cam timing as indicated in my first post in this thread, and probably overall power levels were approaching inappropriate for street fuel (219 whp out of a 2.3 L, running revs up to 8K). I was fine for 3 full sessions throughout the day, I suspect my problem started due to heat soak in the mid afternoon after a day of tracking the car, and a bunch of small things added up. Because I don't think it was a Rich/Lean issue, I don't think closed / open loop would have made a damn bit of difference in my case.... (This year, I'm running better cam timing, XR2CS plugs, 100 Octane, and am trying to get a knock monitor up, running, and tested...)

I think a lot of people can learn a thing or two from your excellent post Brendan. It seems most folks go digging for that ONE thing that caused their problem when in reality its several items that decided to go family style on your ass. There are a TON of variables. The chances of an issue only being caused by one thing is most unlikely.

Jake

[Brendan]

t seems most folks go digging for that ONE thing that caused their problem when in reality its several items that decided to go family style on your ass.

Ha. I laughed out loud when I read that. I think some of my coworkers think I'm nuts sitting at my desk... But yes, very true. While in some scenarios you might be able to pin it to one thing (For example, a failing fuel pump like Ironhead mentioned above), most of the time it's a combination of things. In my case, there's no way to be sure what caused it, but there were several smaller factors that add up to form a pretty good picture...

[stevesingo]

RE Closed Loop: I would run it, and I intend to run it when I get on the road. But, ensure your base map is good, the error limits are set pretty tight on the rich side and if it goes out of limits, it should shut down. On a track car, it should be easy to rig up an error light on the dash for the ecu. You can also keep logs after every run you you can monitor what the ECU is doing to Inj PW and correction to Lambda.

RE Cam Timing: Retarding the inlet cam will not necessarily increase the dynamic compression throughout the engines operating range. It will change cylinder filling but by different amounts both positive and negative throughout the rev range.

Low engine speed 1000-2500rpm: With a less overlap and a later inlet closing event; the reduction in overlap will prevent the descending piston drawing from both the inlet valve and the exhaust (there is little inertia in the exhaust at these engine speeds), which is a plus, but the later inlet closing will reduce cylinder pressures as the piston will be further up the cylinder before the valve is closed, which will negate the reduced overlap advantage as the rising piston will try to force air out of the cylinder as soon as it starts to rise from BDC. If you increase static compression, this compression loss can be completely negated.

Mid engine speed 2500-5500rpm: The reduction in overlap will reduce scavenging (where the inertia of the mass of exhaust travelling out of the exhaust port creates a depression in the combustion chamber around TDC when both inlet and exhaust are open. This aids the air entering the chamber as the pressure differential is greater than it would be in a static condition). Again the later inlet closing will reduce the amount of compression this incoming air will receive, as still the rising piston will try to push the air out.

High engine speed 5500+rpm: This is where the later closing of the inlet comes into play. At higer speeds the incoming air has a lot of inertia, which is attained as the piston falls, creating a depression into which the air flows, gathering inertia as it does so. The inertia gained by the incoming air, will be greater that the force exerted by the piston in it's attempts at furcing air out of the cylinder as the piston rises from BDC. During the rise of the piston from BDC, cylinder filling will still take place.

Obviously the engine speeds I mention are broadly generic, and a lot of where in the rev range these effects take place is related to the spec of the engine.

In general, for a given cam duration, more overlap=more mig range, less overlap=bottom eng that will tolerate more coam duration, and a better top end.


I don't recall what the compression of Brendens motor was, but on something moderate (11:1 ish) with the 284 inlet cam being used, I don't think this is the root of the expiration.

RE Temps/Ign Adv/Det: an engine running retarded ign advance will have excessive EGTs for the obvious reason that the combustion is occuring in the exhaust. As we advnace towards optimal this will reduce in a fairly linear fashion, but as we start to get knock, the EGTs will fall sharply as the energy released is not forceing the piston down (it is still on it's way up) so the energy has to go somewhere. This energy is absorbed by the water jacket, but mostly by the pistons, rings, head, spark, plugs, big ends. This is why it destroys engines.

Tuning to EGTs is not a good plan though. There are too many other things that can effect EGTs to make it reliable.

On the subject of Det and big ends; If we have an engine that has relatively low compression (10.5:1), on good fuel (98RON), and we start mapping the ignition:

Method 1: we advance until we here pinging the back off a few degrees for "safety".

Method 2: we advance until we see no further gains on torque fro the given load/rpm.

The problems with method 1 is that due to the moderate compression ratio, the engine is very tolerant of over advance (ie over advancing does not generate enough heat to start self igniting the air/fuel mixture). SO great, loads of advance ond no knocking! Not quite!!! The combustion process could be occurring too early, as the piston is still rising. This causes a great deal of stress to the BIG ENDs, wearing excessivly and resulting in spun bearings.

I would like to make a data base of the s14 engines that have spun bearings including , stock/modified VE, stock/modified ign, MAXX?, Fuel. I think it would show a trend towards modified engines with standard ign maps, or stock engines with aftermarket chips.

Steve

[///schwartzman]

If you recently lost an engine to detonation, I 100% understand the "paranoia" (hope you don't mind that description).....but again, perspective is important. If you have not done so, I would install BRIGHT highly visible warning lights for oil pressure (30 PSI threshold), fuel pressure (35 PSI), and temperature (mine goes on at 210 degrees). Unless you are positive what circumstances killed your engine, assumptions are very dangerous... You might focus on the wrong issue.

Again, you have to keep in mind what closed loop control is intended to do. It is not a replacement for thorough tuning, it just gives the EMS some capability to adapt to variations in running conditions, wear on components, etc. I always turn off the closed loop when "road tuning" the car, and I cannot even detect any difference in how the car runs with it off. In fact, the AFR graphs are much smoother with closed loop turned off, because the system is not constantly tweeking the fueling and making minor corrections.

And yes, I always run pure 100 octane on the track. To me the insurance is well worth the extra cost per gallon.

I love these threads as it always adds to the knowledge base of what the possibilities could be...

I know we've addressed the oil pressure issue before, but I had my warning lights set to 30psi and after driving my car for 20 min. it was going off @ idle constantly (@80*+ ambient). It is set to 20 psi and it's rare that it goes off now but I worry, as I think we all do about oil starvation.

I think a knock table plotting the most common trouble areas would be nice, but there are so many engine variations...

[Ironhead]

I love these threads as it always adds to the knowledge base of what the possibilities could be...

I know we've addressed the oil pressure issue before, but I had my warning lights set to 30psi and after driving my car for 20 min. it was going off @ idle constantly (@80*+ ambient). It is set to 20 psi and it's rare that it goes off now but I worry, as I think we all do about oil starvation.

I just try to set the warning light to go on slightly below the lowest oil pressure the car ever has at idle....on my car that is 30 PSI. That is one of the things I love about the Spa Design gauges....they all have programable outputs for warning lights....so you can easily adjust it where you need it. The gauges are not the easiest to read at quick glances, but I think it is better to rely on warning lights as attention getters anyway.