Forum » General Tuning Discussion » Realizing what a mess my tuner left for me. Need some guidance cleaning it up

Realizing what a mess my tuner left for me. Need some guidance cleaning it up

General Tuning Discussion

Forum Posts

Courses

Blog

Tech Articles

Discuss all things tuning in this section. News, products, problems and results. 

= Resolved threads

Page 1
Author
125 Views

Ever since I left my tuner (visited him twice for various issues) I haven't been able to shake the feeling that his work was not very thorough and something wasn't right. That's why I took all the HPA courses actually because I felt like if I didn't understand what I was looking at how could I scrutinize it? Well I've found some serious issues here but I think they are fixable and probably don't need a dyno to do it. Or maybe this is hopeless and I just need to go find a new tuner and start over. I'll try to keep this short as I can but it's a bit of a challenge.

First let me start by saying when driving the car on street in the normal operating ranges AFRs seem ok. Doing WOT pulls if anything they were too rich though I've toned that down a bit and have that I think pretty good.

However...

I kept noticing that on cold start and until the engine was warm to a certain point my wideband was reading pretty rich, but as the engine warmed up the tables quickly turned. On cold start afrs will immediately hit 10:1 and sit there for a second or 2. Then hang around 11:1 to 11.5:1 for maybe a minute then spend a lot of time 12:1 to 12.5:1, then warmer spend a lot of time at 13.5 and when it's finally fully warm it does sit at 14.7 until I drive it. After I drive it and then let it idle it always goes to 14.9 sometimes 15.0.

The problem goes deeper than idle though.

I've noticed that even though during road driving AFRs seem reasonable that if I hit the gas in neutral they are not. They are often quite lean. This explains why sometimes I swear I heard knock when I blip the throttle hard when I was burping my coolant some time ago.

Attached are 4 screen shots with some sample lambda from each test. I separated the tests based on ECT correction boundaries.

0. Cold start

1. ECT 128 Degrees F

2. ECT 174 Degrees F

3. ECT 178 Degrees F

These 4 screenshots are named accordingly. Each shows a snippet of the graphs showing a sample of the ECT and IAT during the test and the corrections along with the recorded lambdas in the fuel table. It should be noted that the corrections displayed in the graph are unfortunately not always accurate, or at least suspect. I say this because in the Hondata software I have it showing me the sensor reading live in the display window and in file #2 I am certain the display window showed ECT.c of ~7% however in the graph it says 0%. I believe the graph is wrong and it's a bug in the software. If you look at the other attached screenshot of the specified parameters for the corrections you'll see it should not be 0% for this test as well.

The attachment named "Temp Corrections" shows all the temp corrections. I reverse engineered to figure out that these are the default corrections for a stock Integra Type R (I'm running a b18c1 setup similar to a Type-R just with mildly more aggressive cams and the usual I/H/E so this makes sense at least). However I did notice that the tuner had changed Open Loop High Load @ 168 degrees to 10.2% whereas it should be 20.3%. And for 177 degrees he has a correction of 1.5% whereas the default is supposed to be 0%. I think this really sucks because if I put back the correct value for 168 degrees the whole fuel map > 50kpa will go 10% richer until the ECT gets to 177 degrees. Then at 177 he's added fuel when it shouldn't be so if I remove that it will make the whole fuel map that much leaner > 50kpa. I watched him while he tuned the car and though I didn't understand what he was doing at the time I saw he started with some basemap from some other customer with who knows what setup and went from there. I believe these 2 vales in the compensation tables are artifacts from some else's tune.

There are some other over-arching fuel compensations mixed in I question as well. Also attached is a file called "Fuel Trim" He's got 4% overall added to the whole map all the time. I think that's an artifact too. I know there are tuning processes where you do an overall trim to a base map and find a magic number but then you're supposed to take it out of there and apply it to all the cells instead. Anyway so on top of the temp corrections we've got another 4% there. That post start fuel adjustment is default 0% on stock basemap so I'm pretty sure that's an artifact too and why on cold start it takes me a while to get out of the 11's and 12's in AFRs.

So in summary the problem is pretty much the bottom left corner of the fuel map. It's all lean once the car is warmed up. I'm not sure how to approach fixing this. One thing I still don't understand is when adding fuel I notice people select usually 4x4 grouping and blend groupings together etc... rather than just do the exact lambda correction per cell from the feedback provided by the ecu. I think I could fill out most of the table just sitting in neutral and hitting those spots and could extrapolate for the nearby zones I couldn't hit.

But then I'm at a loss about how to make this play with the temp compensations. Regarding low load those compensation tables are all still default so should I just wait till it's warmed up and then do a process to fix the cells? This will have me adding fuel to all those cells, however I sense that while the car is still cold this will cause all of those cells to be quite rich. How rich is ok at the coldest and mid-range temps? How far into the fuel map should I be going regarding this neutral/not-in-gear situation? When I'm in gear and driving AFRs are ok.

But then the next problem the high load ECT compensations that are atypical. I can live with the 168 degree thing (though I'd rather not) but the 177 degree one to me seems like should be changed. That seems like a ball of wax though or could I just highlight the whole table after 50kpa and remove 1.6% so that once the car is warmed up it should even out?

To make matters worse it's VTEC so I've got 2 fuel tables.

I feel bad asking for so much advice, I need a process to follow here, or an official recommendation to just start over.

I think what the tuner did was start with some else's map as a basemap. I recall he said something about it being from a Turbo B20 setup (note I'm naturally aspirated). He then quickly went through and made some changes to the parameters but not thoroughly, played with the overall trim to find a happy place in idle and cruise. Then left that there and proceeded to do the ramp runs and tuned those sort of properly except they are a bit un-pure due to the correction artifacts. Maybe some of those artifacts make a lot more sense if I had a turbo.

One last note: I am still on open loop.

Attached Files

I think your expectation may be a bit off. Does the car run smoothly when warming up? If so, then perhaps there is nothing seriously wrong. What happens if you immediately start and drive away -- does it run smoothly?

To achieve that you often need it to be a bit rich, and as long as the Lambda is .85 or higher I would give the engine what it wants. So I wouldn't totally throw the tuner out unless you have running issues and he isn't fixing them (or at least willing to try).

If you really want perfection, then you will probably have to spend your own time and resources to achieve that. Find a friendly load-bearing dyno you can rent. Follow the HPA process, tune the fuel map at normal operating temp. Then adjust the corrections as needed on days when you can start cold. You'll probably have to wait for the weather to cooperate with you.

Your test with trying to hit cells in neutral won't work. You need to be in the center of each cell and to get there, you will need a dyno that can apply a load and let the engine run at steady state. When you free-rev the engine with the throttle you are seldom in the center of cell, and mostly likely not in a cell you would ever be in while driving. You need to stay on the cell for 1/2 to 1 second before deciding what the mixture is and determine the change (ramp runs are done differently, but you can tune the low cam on your Honda up to about 4000 RPM steady state all the way.

Where are you located -- perhaps someone with a dyno near you can offer to help.

"When you free-rev the engine with the throttle you are seldom in the center of cell, and mostly likely not in a cell you would ever be in while driving"

Totally agree I don't think I'm in any of those cells when driving, may pass through one for an instant of time but not "in them". But would you say it's normal to be so lean free-revving? Sometimes people do free-rev. At at stop light to be an idiot or just in your garage doing whatever no?

I think if I could fix just the free-revving issue (if it is an issue) than I'd feel a lot better. I have this worry that I've spent so much time and money on this (I did all the building myself) and the thought of free revving it for whatever reason and blowing it really concerns me.

I don't think you have an issue. As long as the engine has oil pressure and isn't overheating - it's practically impossible to cause any real damage by free reving below the RPM limiter -- there is no load.

One reason you want to stay lean is to keep the plugs from fouling, which can happen with a rich mixture. The reason we used to rev the engines at the stop lights was to literally "clear the plugs" and keep them from fouling. But that was before EFI when carbs would run too rich at idle.

As David said, it is normal for the fuelling to be a little rich - sometimes a lot rich if it's cold enough - when cold because of the fuel condensing out on the port and cylinder surfaces, and sometimes poorer atomisation (mixing). Not sure about modern vehicles, but back in the day the lambda would often be inoperative, in the ECU, for fuelling correction below the point where the enrichment was occuring.

Also, as David said, it is unlikely that damage can be done free-rev'ing if the engine is warm, has oil pressure, and isn't taken beyond it's mechanical limits - but it doesn't really demonstrate anything useful.

Back in the carbie days, with 'healthy' camshafts, there was a phenomena called 'reversion', this was when the piston would push some of the charge back into the port and push that air back through the carb. What would happen is that that small proportion would actually pass through the venturi 3 times, so have 3 times the fuel, which could average out as being rather rich overall. When you also consider the narrower operating heat range of the sparp plugs back then, fouling could be a big problem at low rpm. With some vehicles the reverse pulsing through the carb' could be seen as a fuel mist around the carb entry, this was known as 'fuel stand-off', and was particularly dangerous when there was a risk of a backfire - here is a rather cool example - the Mini is using a mechanical fuel injection ahead of the slide throttles, and you can see the fuel pooling on the slides. https://www.youtube.com/watch?v=hHjf9zhQIRk&list=PLpu6GsSJ4YNokDfkLnBMzFIwWqIJHN0Fl

That is interesting. At the end of the video you can see it spraying the camera with wet fuel. And I thought carbies were simplier than EFI, sounds like just a different set of issues.

Also sounds like the verdict is lean during free revving is negligable. Again recognizing not a lot of practical applications for free revving but it happens at times. Bottom line I did another test where I held the throttle partly open until 4k rpm and held it there and the AFR sat at 17:1 and this is of no real cocearn? Of course there is no situation other than a silly test where I’d hold a free-rev that long.

Is the engine misfiring at 17:1? You can certainly change it, but that is only 15% lean of stoichiometric -- just about ideal for maximum fuel economy. You could cruise at that setting.

Nothing unusual under those circumstances except the AFR reading. Maybe I’ve been a bit overconcearned about being as close as possible to lambda 1 on non-wot areas.