Self-Tuning to ease the boredom.

[dyn085]

Something strange I'm reading... Referring to "pro-tune" vs "ATR" tune...
It's the same thing. But, you are referring to an e-tune if you didn't have the tuner in your car. A protune would be when the car is tuned onsite.

One other thing you mentioned lean burn cruise. That's impossible in this car, you would have to develop an entirely different cylinder head.

A 'pro-tune' is a tune by a 'professional'-typically someone that gets paid for creating it. An 'e-tune' is generally a 'pro-tune' done via the internet versus your local dyno. Basically, you can get a pro-tune directly from the pro or via the internet, and you can get an e-tune from a pro or someone that's not a professional.

I've seen the discussion from both sides and either way it goes it's a bit pedantic.

I can concede that it's not 'lean burn' cruise in the way that most communities discuss it, but at cruise I am burning less fuel for combustion than the stoich rate of 14.7:1.

 

[blurred]

A 'pro-tune' is a tune by a 'professional'-typically someone that gets paid for creating it. An 'e-tune' is generally a 'pro-tune' done via the internet versus your local dyno. Basically, you can get a pro-tune directly from the pro or via the internet, and you can get an e-tune from a pro or someone that's not a professional.

I've seen the discussion from both sides and either way it goes it's a bit pedantic.

I can concede that it's not 'lean burn' cruise in the way that most communities discuss it, but at cruise I am burning less fuel for combustion than the stoich rate of 14.7:1.

You aren't getting better fuel economy. By all means try it, but it's a debunked theory.

Pedantic maybe, but unless the pro is in the car you're not getting his full expertise. His feel and ears are equally if not more important than logs. So yes there is a definite differentiation.

Yes my personality type is perfectionist followed by precisionist

 

[dyn085]

No worries, discussions are fun. There are times that people just have to agree to disagree.

 

[dyn085]

Knock, knock.

Who's there?

87E10.

Get it? Ugh, wtf was I thinking? My entire day has been wasted and I'm seriously contemplating just starting my car and letting it run until all the fuel is gone. Everything was a wash today and I've re-loaded the OTS map with a couple changes for global timing. I seriously can't believe that there are people out there that use this junk.

It's really quiet though...

 

[koozy]

so... you forgot you had 87 in the tank? I think I recall you putting that in to test out the effects of poor gas quality on OAR?

 

[dyn085]

so... you forgot you had 87 in the tank? I think I recall you putting that in to test out the effects of poor gas quality on OAR?

Lol, no-it's a joke about how much knock I'm getting with it in the FoST. The 2l EB is ridiculously sensitive compared to the 1.6l imho. Not only can I not get past -.22 OAR, but I've already done a -2 degree global reduction and I'm about to load up another tune with some switchable maps of varying global timing changes.

If I can get to the point of no more negative corrections I'll be happy enough because it's only a drop-dread resort in fuel choice, but I have been having a rough go at it so far. I guess I'm about to learn the 3D timing timing changes in the switchable maps.

 

[dyn085]

So I've been piling and compiling the data from yesterday looking for trends/similarities, issues, etc., while working with this crappy fuel and I realized a couple of things-

A. My pen sucks.
B. I'm doing lots of simple math manually, and that takes a lot of time and thought.

I don't know why I'm only just now realizing this, but because I'm following a certain process and relying on certain data within an Excel spreadsheet, I can probably reverse-engineer it into a simple workbook that would allow me to copy certain rows of data and let the program do the calculations. Theoretically, that would allow me the ability to copy whatever specific rows I was intending on using manually, paste them into the workbook, and let the math plot corrections into multiple tables at once. Obviously this means that datalogging would have to be pretty precise for the correct input (trash in equals trash out), but it works in my mind.

Maybe I've over-simplified it to the point that it actually becomes more difficult, but if it works then I would be able to share it and it could simplify the process for others. I'm only talking about ignition as a whole and this has nothing to do with any other facet of tuning, but there is already an Excel sheet (I haven't actually used it yet) for VE and this would ultimately really be no different...

Just thinking out loud. Feel free to weigh in your thoughts.

 

[blurred]

Try the airboy spreadsheet I used to use for my Subaru when I was tuning it. May or may not work, but might be a nice starting point

 

[Siestarider]

I would gladly help if I were fast enough with Excel to be useful, unfortunately I am just fast enough to be really painful.

 

[dyn085]

I'm pretty good with Excel, but I think that what I'm trying to do with the workbook is more complicated than I originally thought. Seems how I need to get these multiple baseline tunes done rather quickly it's been put on the backburner for now, but I'll revisit it shortly.

New books for reading-

I finally got through that tank of 87, and the resulting tune will do what I need but not much more. I tried making ignition changes on the OTS 87 tune, and it seemed that no matter what I did I couldn't get it to play well with my finicky cyl 4. I ended up having to scrap the work I put into it and load up the stage 0 tune and go from there. I only ever messed with the ignition timing and hit the 'turbo off' button. All I really wanted the tune to do is not knock, and it will do that (mostly), so it will serve it's function for now. Maybe I'll revisit it in the future once I have a better idea of what I'm doing.

My cylinder 4 is really frustrating me because it acts up a lot compared to the others, so I'm just tuning for it and letting the other cylinders do their own thing, though I took the KS advance down to 2 degrees to not let them completely walk away.

Either way, I started working on my 89 tune yesterday (I only just found out that I don't have access to 91, lulz) and it's been much more pleasant to deal with than the 87. I started with the OTS stage 3 91 octane tune and have been making minor ignition adjustments to stay out of knock. I think it's as close as I need it to be for now, so I'm going to try and do some VE and HDFX work/adjustments today. Super-baby steps as I'm still learning how everything functions individually as well as together.

I really like the Banish books.

 

[blurred]

If you're having an issue with one cylinder then you might want to first rule out mechanical issues. Most important thing to learn is how to read plugs. Also to make sure you have a good baseline make sure all the plugs are gapped properly, you could be tuning on top of a problem

 

[RAAMaudio]

I have had issues with cyl 4 as well and sometimes cyl 1 but mostly 4. It got better when I swapped and indexed the plugs but then found they were the wrong plugs so back to stock and made sure the best indexing was on 1 and 4 and though still not perfect it was much better. Going to start with Russ on the EFR tuning hopefully today, if we see issues on any cyl I will order more stock plugs and really dial in the indexing to see if it helps.

 

[dyn085]

If you're having an issue with one cylinder then you might want to first rule out mechanical issues. Most important thing to learn is how to read plugs. Also to make sure you have a good baseline make sure all the plugs are gapped properly, you could be tuning on top of a problem

Post #18. Reading plugs is good, but the only truly accurate way to read them is to put them in, do a WOT pull, and shut the car off as soon as possible after putting in the clutch-and because I'm not shooting for maximum power right now that would be a waste. Luckily, the ECU gives me enough data to work with without doing that, though it would simpler on a dyno.

I have had issues with cyl 4 as well and sometimes cyl 1 but mostly 4. It got better when I swapped and indexed the plugs but then found they were the wrong plugs so back to stock and made sure the best indexing was on 1 and 4 and though still not perfect it was much better. Going to start with Russ on the EFR tuning hopefully today, if we see issues on any cyl I will order more stock plugs and really dial in the indexing to see if it helps.

Unfortunately, the FoST is known for having cyl 4 issues and that is typically the one that has ringland failures. I've addressed the timing issue associated with it on the 93 tune, and soon it should be fine on the 89 tune as well. There is a larger issue that we may not fully understand yet, but people much smarter than I are working it.

 

[RAAMaudio]

It sounds like an oiling or water flow issue or compounded issue caused by one or both effecting each other which could be a tough fix, opening up the oil gallery or bringing in water from both sides of the head, which I have seen both done.

Sounds like you are on the right path, tune the worst the best you can and live with the others not fully optimized.

I got the EFR FiST running today, have to sort out one item then can start making tuning runs. I am going to keep a good eye on #1 and #4 and see how it goes.

I am not doing the tuning but will go over every log carefully to ensure all is well.

Since I do not drive this car much my aim is not ultimate fuel mileage it is instead making the power I want on track, good power on the street and both a safe tune on 91 Octane so I can avoid the hassle of mixing fuel, WMI, etc...been there, done that, not interested.

I have been spending more time looking into natural gas and propane injection for my duramax, not for power, it has plenty, just better fuel mileage. So far I am not that impressed with the cost/benefit ratio. It has DEF and a particulate filter to regen, that stuff sucks but I never smell a thing out the exhaust and that is important to me as I do care about this little planet we live on.

 

[dyn085]

Definitely ensure the logs are consistent. I've actually been surprised at just how much variance I've seen between logs because I'm road tuning on a road that I normally wouldn't use due to its location and infrequent traffic. It doesn't help that I'm using gas that I would normally never use, but that pain will be done with shortly and I'll go back to 93 soon.

 

[RAAMaudio]

I have a road I can use but it is a bit bumpy at speed where I can get away with it and could be longer, the better section is to easily seen from the freeway were I cannot see traffic well enough and could get a nasty speeding ticket. I have another road that has no traffic but not long enough to be safe and even bumpier. I use the freeway close to me as I have a section where I can see everything, get to speed and slow down quick enough, just have to time the traffic to not pass anyone or hold anyone up, usually no cars or trucks on it but sometimes I have to abort a run or go back around and try again. There is no place for the law to hide, wide, smooth, but it is slightly uphill so I make a run each way and compare the data which does not very by much, it is the best I can do as living in the mountains is great fun, just not many roads for tuning safely.

The EFR will not take that many runs I am sure since Russ has so much experience with a very similar system on Rod's car, EFR and now GTX, he knows just what to look for and should be able to dial in without to many runs.

 

[Hypntick]

Subbed, very interested to see where you get with this.

 

[dyn085]

My latest cut/paste-

Well, I finally made the switch to metric gauges. It sure seemed to simplify things.

I'll post a pic of my BL table 7 (because it seems to get the most work) in a minute from my phone, but my timing seems excessively low for some reason. I don't expect it to be through the roof because it's only 89E10, but I was hoping that someone more experienced could weigh in with their thoughts on it. I've basically tuned out most all knock and set the KS max advance to 1 everywhere but above 5k+ at 1.2L+, and if I go much further at all my cyl 4 gets knock-y and pulls it's timing while the others walk away. Part of me thinks that it would be ok to let them walk and try to be all that they can be, but the other part of me thinks that it makes more sense to have them all running closely (even if not optimally) with cyl 4 not knocking.
Edit:

Basically, I've followed two separate tactics along with the occasional WOT-

1.) Set rpm to 2k, 2.5k, or 3k while on cruise and in a certain gear for a hill, allowing the load to change while isolating a certain column.
2.) Managing load via the go-pedal in as close of proximity to a certain row, and trying to keep it there through as much or all of the rpm range. Needless to say there was no .4 load done at 6k and the 1.4 load went significantly faster than the others.

Usually at the end of the day I would log 5-minute increments of normal driving back to the house for some overall evaluation and they seem to be looking pretty good for the most part, though fuel mileage doesn't seem to be quite tops. It's pretty hard to evaluate that with the abuse the car is going through currently, though. What I have noticed, as retarded as this will seem, is that I see more knock on cyl 4 in left turns than right. I know that it has something to do with the 2-2.5k, .6-.8l cells of BL 6&7 (they're very similar), I just haven't figured out why yet.

I've been doing a lot of tail-chasing, and I sometimes wonder if it's from compensations. I have been looking towards the OAR table and was going to drop the advance to zero, but both the advance and retard tables state that the tables aren't used in the tuning tips. I don't know whether to believe that or the tuners that state that they use OAR for tuning because both are more experienced than I, but if I can't use those tables then it seems like it would be a better idea to push the tune until OAR read 0, or as close to zero (preferably on the negative side) as possible?

Using the 91 stage 3 OTS as the base-

I have another post that I will be putting up in here in a few minutes (hours?) that should be a bit more applicable in the 'general' sense for those of you with ATR to confirm/deny. I've been having a lot of fun with the overall system so I think that I may be ordering it for my FiST as well...but Idk yet.

 

[dyn085]

Ok, and now some additional thoughts beyond what I've posted in the FoST community-

So far I've really only been messing with timing and evaluating ways to make it easier for me to log and evaluate, and then adjust for. Because the FoST community is already pretty established in their process, I figured I would post my method here and others with ATR can confirm how similar it is/isn't to the FoST. It may/may not be redundant and if it is I apologize; trying to share this data in multiple communities and follow the individual discussion-paths causes me to be more repetitive than I usually am.

First off, my overall process of maintaining data on my laptop. Everyone is going to have their own specific way of organization, but I figured that I would post this as what I'm doing simply because of this one single reason-you're going to collect more data than you would possibly imagine. Especially for those that consider making tunes for multiple fuels, it's very easy to get overwhelmed with the amount of data you're collecting and it becomes even harder to try and remember what was the change in-between this 'datalog1' and that 'datalog1'. If you have tons of time then it may be worthwhile to simply save each log with specific info, but I've literally amassed over 200 datalogs since starting this and that went out the window on Day 1.

First off, here is an image of my desktop. You'll notice that there are three separate sections-programs, data, and supporting data.

Programs and supporting data is pretty self explanatory-one opens up/adjusts/evaluates data while the other is background info for what you're working with. I'm going to focus on the data portion because that's where I created a huge mess initially.

You'll notice that there's the three different fuels plus another folder called Maps. Maps is strictly completed tunes that are ready for upload to the AP3. This usually contains one base (OTS v312) map that I can upload at any moment if I've frustrated myself and want to start from scratch. Everything else is a completed file saved directly from ATR. As of this typing, this is what it looks like inside-

At some point in time I get tired of having a ton of maps in there and distribute them to the related octane folder, which is another system I'll describe....now. The octane folder is where I save data directly from the AP3. Every time I stop to evaluate data and make changes, the AP Manager gets opened up, the datalog files are highlighted to be saved and are placed into a file titled after the tune from which it was evaluated. So this is what it looks like when I open an octane folder-

So you can see that every time I've made a change to a tune, I simply add one to the tune number and save it accordingly. I'll address that shortly because we're kinda working backwards in showing how I store data, which is the opposite of how it's collected. This folder usually only has the datalogs in it until I move the associated tune file into it, but here is one that's complete-

So now I know at any future point in time that if I want to backtrack or re-evaluate some data, I can go to a specific folder and open it to see the tune that was evaluated and the resulting data from that particular tune. It's not often that I do this, but it can mean the difference of going back one or two files as opposed to starting from scratch on the second day. This happens occasionally when you overshoot a target or inadvertently make an adjustment in the wrong direction, and a lot of times I will delete the entire folder in order to avoid picking back up with it on accident.

So basically, the process goes like this, starting with the OTS V312 map-

1. Load V312 into the AP3.
2. Load V312 into the ECU.
3. Datalog V312.
4. Open AP Manager, highlight datalogs and save them in new-folder V312, located in E89 folder.
5. Open ATR, load V312 map.
6. Evaluate V312 datalogs and make appropriate changes in ATR.
7. Save new map as V313 in Maps folder, proceed back to step 1.

At some point in time I get frustrated with the number of maps in the Maps folder and make a bulk transition from there to the appropriate octane folder, and later the individual map files can be moved and stored into the appropriate folder containing the datalogs.

I also keep a notepad with chicken-scratch that briefly outlines what each datalog was for (i.e.-WOT, hills, load, rpm, 5-minute drive) and brief notes on what changes I made in ATR associated with the tune before saving. That may or may not help you if you have to reflect back on something, and a lot of times I feel like it's wasted time because I can pretty clearly see what's being evaluated between logs. In all fairness, sometimes I add an oddball change that may not be related (KS adjustments, etc) or immediately noticeable in a log, so I suppose it could help in the long run.

Anyways, this is just what works for me and where I'm at through trial-and-error, and something different may work for someone else. All I can say is that I ended up having to trash a lot of data because I wasn't this organized initially, so it's worth considering if you're thinking about before trying to learn tuning.

I'll post some more soon with examples of logs, how I collect them, and my process of evaluation.

 

[dyn085]

Ok, I'm going to get a little more in-depth now, though this is still generalized discussion that's not model-specific.

First off, lets discuss using metric gauges. It's difficult at first, but will make things significantly easier for those that are considering going the route of self-tuning. As an example, your wideband O2 could care less what kind of fuel you're using. You may see 14.7 AFR on your gauges, but what your O2 sees is that you're at 1 lambda. That stoichiometric 1 lambda is no different from E0 fuel (14.7:1) than it is from E85 (9.87:1)-it's showing that you've made stoichiometric balance with the fuel you're using, regardless of how little or much fuel it took to get there. All of your changes in ATR are metric, so you may as well get used to seeing it now.

Secondly, get in the habit of datalogging parameters that are only related to one general aspect of engine operation-air, fuel, spark, or turbo. I've spent a lot of time using and looking at datalogs with the default parameters plus all four ignition corrections, but as much as that has helped me learn and get comfortable with the overall datalogging/evaluation process it just doesn't give you as much detail as you want when you're making adjustments of parameters in multiple areas of your tune. The FiST may/may not be similar/different, but this is an example of what I log for making ignition corrections (spark) on the FoST-

Engine RPM
All 15 HDFX Weighted Percentages
All four ignition timing
Load actual
Spark Limit Source
Spark Source

Occasionally there are a few that I might throw in for curiosity of other areas or to ease the evaluation process (ignition corrections), but the more detailed information that you can get about one specific area then the more accurate your decisions can be. Unfortunately, in order to get really accurate information in one area you'll need to forget about the others and move through them all accordingly. And yes, changes in one area can/will affect the others, so it's a continual process as you work each area into a more 'optimal' operating range. Obviously many of those monitors are useless to VE (air), but many are directly applicable and would be logged when evaluating that area as well. Find what is directly applicable/necessary and forget the rest.

Lastly, know and understand from the onset that the most accurate information that you can get would be on a load-bearing dyno. This doesn't mean that road tuning is impossible or that dyno tuning is the only way, but it's much easier to make changes based in a world where environmental conditions are essentially the same as opposed to logging/adjusting information at 100 degrees and then working through throttle-closures when ambient is back at 70. It's a crude example, but an example nonetheless. It's also significantly easier to set the dyno to maintain a certain load or rpm than it is to try and do manually while on different roads, elevations, etc. I intend on using ATR and learning/establishing my processes on the road, scheduling some dyno time when I feel comfortable that I won't be wasting my money with it, and then further fine-tuning back on the road if necessary.

In a while I'll post an example of a datalog used for evaluation, how I personally evaluated it, etc., along with different shots of the log as I moved through one-single process. I should probably eat first, and I think I'm out of beer too so that will need to be rectified first.