Txdragon

I have tried interpolation after interpolation and I still end up with a map that looks like this... What is the best way to smooth it out? I finally nailed a great street tune but my map is embarrassing.. Ugh.. Help!

Txdragon

Gonna give this the ol' "bumparoo"

pogeeboy27

iTrader: (3)

Go cell by cell if interpolation doesn't work. PITA but gotta do what you gotta do.

Txdragon

I get a good table, it looks like this, then I interpolate and bleh.. Almost back to square 1. I suppose I can tackle cell by cell if I must. Although, that gives me an opportunity to test out the method I found. This should be interesting, if not fun..

pogeeboy27

iTrader: (3)

I've never found the interpolate method to work flawlessly. When I do use it I still have to edit cells in between. Hopefully you saved your last good map.

Txdragon

I did. It was this one lol. Runs great, just shitty looking map. I was under the impression if a map looked like this, the tune wasn't good. Doesn't seem the case though **shrug**

motoxxxman

yes, it is still the case.

that is not a quality tune. a halfway decent tune will look like it has already been interpolated correctly.

no offense, but that map is absolutely terrible. and I think your idea of it running well may be compared to a 1920's ford lol

Txdragon

then I suppose you have insight on how it can be fixed? Everything I have done has left my map looking like this. Either breaking it to its core by tuning cell by cell individually, was not good enough, or the software is junk?

motoxxxman

what does the ignition map look like? because timing accuracy and smoothness has a HUGE impact on afr readouts. a lot of people tune just the fuel maps to a desired afr, and never even look at the timing maps. and incorrect timing can cause some pretty drastically false afr readings, which can result in a fuel map like yours above

motoxxxman

also remember, a tune can only be done as accurately as the method of datalogging. the ecu is always interpolating from 4 different cells at all times. so to get each cell as accurate as possible, you need to narrow down your method to track as closely to each cell as possible. this can really only be done in one of two ways: 1) on a dyno, locking it to a steady rpm, and tuning each pressure cell individually by manually holding specific manifold pressure (expensive, time consuming, and lots of heat with incorrect load on the engine. or 2) on the street under real conditions, by staying within a single column at a time. ie starting at a very low rpm and maintaining manifold pressure as steadily as possible and datalogging all rpms for that specific manifold pressure. You will have to gradually increase throttle as rpm climbs in order to maintain the same manifold pressure, but it has to be gradual and smooth. takes some practice to get used to doing, especially as you work your way through the different manifold pressures and rates of acceleration. this requires long stretches of smooth road with varying grades up and down, little to no traffic, and extremely smooth throttle operation. any twitch of the throttle will alter the manifold pressure as well as trigger tip-in settings which will heavily throw off your readings.
It should be fairly obvious that lines should never cross. if they do, something is definitely wrong. each line should hold a very similar shape to the ones immediately above and below it.

motoxxxman

as an example, here is my fuel map before doing any smoothing. I just did this 2 weeks ago after making some minor changes to the engine setup. You can see a couple points need minor touchups for smoothness, but it's very very minimal. I did this on the street using the method I mentioned above. notice how the lines are very evenly spaced, shapes of surrounding lines are near identical, and of course nothing is crossing. again, zero smoothing to this map yet, all just target afr adjustments according to datalogs.
I also never use the interpolate feature while tuning. the only time I ever use interpolation is when creating a custom basemap, before doing any actual tuning.

Txdragon

What you're explaining for street tuning, I am certainly giving my best efforts to stay on any given location on the map and as steady as humanly possible. I try to hold a spot for 20 seconds or longer.. My ignition maps are stock except for decel and WOT. Decel I tuned for some crackles, ok, stop laughing... WOT I tuned at 29 deg. I took the stock values and just adjusted using the constant RPM variable to equal 29 deg at WOT in the RPM/Load range of 3rd-5th gears. All else I have not touched.

One thing I noticed today is that I let my car reach as close to 175 deg ECT as possible before driving. I started it and let it idle for about 15 min then connected. I was at 160 lol.. Anywho, I started logging at that point and noticed that I am MUCH closer to my actual set target values than when I just start and warm up a couple minutes then go.. I drive about 15 miles to work and when I got to work, half my map is +/- 1% in the majority of the cells with a few at +2 and a few lean spots (3-5%) tossed around here and there. Really rich around idle with some as high as 18%..
I have noticed that when I am making adjustments to a cell or a couple cells, at times, others will change as well. This gets frustrating because it throws the whole column off again and I can't seem to dial in anything perfect.. Sigh..

I will be trying out crome before too long, I got the free version and went through all the dances to see if I could load it to my demon unit and it was a battle at first but I got it to load. Switched back and decided to research a bit more on that particular program. If I am having ANY of the same problems, I would then look closer at what I am doing wrong. If not, I would look closer at deserting my Neptune license..

motoxxxman

when tuning on the street, don't focus on specific cells. just focus on keeping the manifold pressure as spot on as possible to the column you're working on. like if you're doing the column for 14.5"/hg of vacuum, just start at super low rpm, hold throttle at 14.5"/hg, and watch your guage as closely as possible, smoothly operating the throttle to keep manifold pressure at 14.5"/hg through the entire rpm range. don't worry about how long it dwells on each cell. as long as manifold pressure stays at the target pressure, it will gather enough data points for each rpm point (each cell). do it 1 column at a time. I usually set it up so it only logs in 3rd gear. that way I can drive it however I need to in any other gear without recording false data, so I can avoid traffic, bumps, cops, etc etc.

also, ignore the first cell and last cell that it records for each column. they will usually record really lean or really rich due to tip-in settings. that might be why you have super large spikes in the low rpm area. it's probably lean spiking on tip-in, giving false readings for the actual fuel map settings.

motoxxxman

if you have a mostly stock setup, including stock injectors, it may help to just start over again with a stock map. just add 5% to the entire fuel map, then start right off the bat with wot tuning, and do each column 1 at a time, working from highest to lowest. and don't use any autotune feature, do everything manually with hotkeys for increase or decrease. go slow, a little at a time. if you get a lean reading, get out of the throttle immediately, add a good amount of fuel, and start again. every time you make a change, erase all recorded data and start again double checking. literally only do 1 column at a time. record the full column. pull over, make changes to that one column. also adjust the column to each side to match the same shape and spacing. then double check the same column again. if all is well, proceed to the next column.

remember, if say you're tuning the 14.5"/hg column, if the adjacent columns are say 18 and 11, if you can't keep mani pressure 100% exact (it's near impossible) and you drift to say even 14.0, it will also be reading from the 11"/hg column values. so if that isn't close, it will give terribly inaccurate readings. which is why you should always be adjusting adjacent columns to match the shape of the column you're focusing on.

it's a long slow process that takes a lot of practice to do efficiently, and even then is still very time consuming. it can take anywhere from 2-6 hours just to tweak the low cam fuel map properly. then you have to factor in all the compensations that can change it. and if the compensations aren't set properly, you'll get different readings every time you hit the street.

something else I forgot to mention. because IAT has such an impact on afr, you should never record anything in the first couple minutes after getting back on the street from being parked hot. it heatsoaks and gets false iat readings, which leans out the fueling a lot more than it should. drive for a coupe minutes after being parked, preferably heavier throttle, to cool the manifold back down and eliminate the heatsoak so it doesn't have false iat readings. once you see iat stabilize, then you can start logging again.

it takes a lot of time and patience to learn how to street tune well. if you're not up to the task, hire a professional. if you're determined to do it yourself, expect to take many many days practicing and tweaking everything multiple times. compensations alone can take frickin forever to get right sometimes.

Txdragon

Thanks for that, this will actually help me tons. I hadn't watched my map input more than I was the table. I have been more focused on keeping my traces in a specific spot on the table.. Ugh.. I have been avoiding the first and last though. Somebody gave me that info some time back as well. Guess with this new bit of info I shall wipe the slate and start fresh... For the 30th time lol. Still learning some cool **** along the way though. I'll report back here in a couple days with Hopefully, better results

motoxxxman

Good luck! I've been tuning for about 15 years now, and I still learn things with almost every car I tune. Don't feel bad. It's not something picked up in just a matter of days. It took me 3 or 4 years of frequently tuning my cars and friends' cars before I was comfortable starting to tune for paying customers. During that learning phase, I also bought a couple single slammers I could practice on and not care if I damaged, so I could learn just how much of an effect certain things had. It worked nicely, I actually had fun causing damage and learning in the process lol. It helped tremendously in learning what NOT to do

Txdragon

We all learn what not to do long before we learn it proper lol

motoxxxman

if you find it doesn't collect enough data points for each cell in higher throttle areas, try doing it on a slight uphill.
and vice versa, for low throttle (high vacuum) go downhill, so it can actually accelerate through the full rpm range at such low throttle. Or, get to high rpm, let off to the high vacuum column you want, and decel while maintaining that pressure. If you want to use the decel method but it doesn't decel fast enough, you can ride the brake a little.

also, look at your iat at cold start, compare it to ambient temp. hopefully it should be very close. After driving/cruising for 30 mins or so, look at the iat again and notice how much higher it is than ambient temp. That is your target iat for when logging. because now, park for a few mins, idling or shut off, then look at iat again. you'll see it has skyrocketed. now go driving again, staying above 30mph at all times, and watch iat and notice the different things that make it drop back down, how long it takes, etc. the goal is to know what iat it should drop back down to before recording again, to make sure iat compensations dont screw up your readings from heatsoak.
For example, on a 60 degree day, after cruising for a while my iat's are around 70-73 degrees at higher speed, and 75-85 in 25mph-30mph areas. After parking for 10 mins or so, my iat's climb to around 100-120. After driving again for a minute, temp drops pretty quick back down to 90-100, but can take several more minutes, up to 10 minutes, to drop back down to the typical temps.

motoxxxman

lol yup. hopefully without irreversible or unintentional damage

Txdragon

I haven't broke anything yet (knock on wood) except cutting my fuel economy by about 30%. Not too big a deal about that since it only takes 26 bucks worth of premium to fill it up. I had an 06 Durango before this that would take 29 dollars of regular to get me about 1/4 tank... Needless to say, it's still a win!

motoxxxman

I know the feeling lol. I'm on e85 now and have a heavy foot, and I still get better average mpg than my jeep XJ with a light foot lol

Txdragon

Think I FINALLY figured out my issue.. Will have more info on this later on or tomorrow..

motoxxxman

curious to hear

Txdragon

I decided to steer away from tuning by table and went straight to the map. I will check the table, of course, to see where I need to go, then make my adjustments to the map. I figure it like painting a picture lol I then go to individual cells only if necessary. The following map is my last 2 days worth of driving, logging, and adjustments. I am still a way off from what I want, but it is looking WAY better than any of my previous maps. Have I really noticed a difference in feel? I will be honest and say there is a noticeable difference in pretty map vs seismographic error map.. The top 2 rows I am never in, so I adjust them as necessary to keep it pretty.

And before it's mentioned, Yes, I fixed the skewed row at the 5k mark lol

motoxxxman

lookin good! it always helps to try different methods

"seismographic error map" lol, I like the terminology