Braineack

iTrader: (62)

In the same load/rpm my ignition map is around 29°. I was averaging around 27mpg city with rx7 460cc injectors driving 10 miles a day, 5 days a week; never getting above 50mph.

I wasalways hapy with that, and I've never tried to tune for more since I harly drive it on the highway or extended periods.

But yeah, generally speaking, as you advance spark you can reduce fuel. This is because the flame speed is reduced as you lean out, and you add advance to put the flame front back to the same spot realitevly speaking.

I think you'll be most fuel efficient if you can maintain 16.0:1 AFR with the lowest PW as determined by the spark angle.

hornetball

iTrader: (4)

Joe, you mentioned not having a dyno in the garage. Have you tried Virtual Dyno? It yields repeatable results which is what you need for tuning. I've been meaning to use it for spark tuning, but have been waiting for a mod that outputs the calculated torque so I can divide by MAP and smooth things out for boost variability. Given that you are NA, this would not be an issue.

I knew I had seen that stock map somewhere. Peak advance above 40 certainly suggests that our spark MAPs are overly conservative. I'm leary of directly equating the non-dimensional load numbers to pressure values though. It may not be a simple ratio conversion.


WARNING, what follows is a super-simplified spark theory 101 intended for noobs reading the thread. Worth having a few rules of thumb handy to follow the discussion.

/Begin Spark Theory 101

The objective of spark advance is to achieve peak cylinder pressure at a rotational crank position that yields best torque. The two variables are RPM and how fast the fuel burns. RPM is easy to deal with. How fast the fuel burns is affected by several factors. The main ones are:

1. What is the fuel? Some fuels are engineered to burn slowly (like AVGAS), some quickly (like race fuel).

2. What is the AFR? For gasoline, your fastest burn happens in the range of 12.5:1 to 13:1. Go richer or leaner and the burn speed slows down.

3. Charge Pressure. The higher the pressure, the faster the burn speed. That's why you retard for boost. That's why Hustler's spark maps use less advance on higher compression engines.

So what do all these variables mean? You have to tune for it and it's a PITA. Most (like me) are lazy, burn in a safe spark tune and dream about having the time and equipment to actually do it right someday. And to really do it right takes a lot more than a few full-throttle blasts on the neighborhood dyno with a "tuner."

/End Spark Theory 101.

Ryan_G

iTrader: (8)

It takes a lot more money and time then most people are willing to throw at it considering most people are averaging high 20's low 30's combined driving and that is satisfactory, especially for the turbo cars.

I feel as though to really do this right would take many many hours on a load dyno and would cost a lot of money.

hustler

iTrader: (6)

Not really. You can run lots of authority on EGO with two different AFR tables (if PID EGO is set-up right) and then adjust spark and fuel to measure output with MAP constant (brick on throttle).

You can also lock wheel speed and go for the smallest duty or PW to maintain that speed, but that takes some throttle finesse and I try to eliminate as many human variables as possible. Realistically, you can fine-tune fuel in cruise in ~30minutes. You have to go in with a plan though

richyvrlimited

iTrader: (1)

I was always under the impression that adding fuel reduced the flame front speed, that's why it helps starve off knock by being too rich.

More advance means you're igniting the fuel earlier, which also means you can reduce the amount of fuel, said fuel then burns faster - which doesn't matter as it's been ignited earlier.

hustler

iTrader: (6)

MAx flame speed is ~12.5-13.5 AFR, flame speed is reduced everywhere else. The reason you advance spark angle with a lean AFR is to adjust cylinder pressures at the target piston speeds.

Ideally, you run a longer duration burn at low pistons speeds, shorter duration at peak piston speeds. There are other variables that affect this though, so I run static AFR targets from 0-4000rpm, 5000-7000rpm.

hornetball

iTrader: (4)

AFR vs. flame speed looks like this: ^

The tip is around 12.5 to 13:1. Go richer or leaner and flame speed slows. Starting from 14.7:1 and going leaner slows flame speed. Starting from 12:1 and going richer also slows flame speed.

Apparently, Hustler types faster, hence the name.

Braineack

iTrader: (62)

here:










that posted OEM map vs. a typical basemap you might end up with:



I'm guessing on the load values, but it makes sense to me. or shoudl I assume the row I marked at 60% should be 100% load? I need to find the old thread where someone was taking obdII data and populating a spark map...


but there's also a huge thing to remember.... hold on while I set it up...

Leafy

iTrader: (1)

Trouble with using VD to tune cruise spark is that its just as useful as looking at injector pulse width in a log afterwards. It takes a long time. If you're on a proper load varying dyno you can literally lock into a wheel speed hold the throttle so you stay in one cell, and keep hitting the +1* button in your tuning software until the torque on the dyno display plateaus (or it knocks), then back down one. Then move to the next cell. Or if you are trying to max it out, once thats done, you can lower commanded afr and add more spark while keeping the torque number the same or higher. Basically what hustler is saying.

richyvrlimited

iTrader: (1)

Thanks for the clarification, makes sense

hustler

iTrader: (6)

I should note that the goal is almsot always the longest duration burn, hence the awesomeness of the soft-head piston design. However without mechanical octane in that type of combustion chamber, we run a shorter duration burn at higher piston speeds because it creates more peak-dynamic cylinder pressure.

Braineack

iTrader: (62)

so while the stock timing map is intersting to look at, this is the AFR map it also follows:



so I mean, really, not the best point of reference... because that timing drop at 5K correlates to all that extra fuel being dumped in during open loop.

hornetball

iTrader: (4)

I don't have a steady-state dyno sitting in the garage. Wish I did.

For us cheapos, it would seem that a good strategy for street-tuning spark is:

1. Tune part-throttle on a long flat road at either constant MAP or constant RPM. If using constant RPM, you can tune spark for lowest PW or lowest MAP. If using constant MAP, tune for lowest PW or highest RPM.

2. Tune WOT with timed runs or something like VD. Use your part-throttle runs to make sure there is no law-enforcement present ahead of time.

I suspect this is a lot easier to do in W. TX than in Southern Cali or the East Coast.

Leafy

iTrader: (1)

I used to want to do it the cheapo way. BUT 1 speeding ticket pays for most of the dyno time you would need. AND the cheapo way takes forever, FOR-EV-ER, the dyno takes 4 hours tops and gets you better results. Oh yeah, I can road tune a car in decently in 4 hours, but I can make it perfect on the dyno in 4 hours. It'll take months to get it perfect road tuning.

hustler

iTrader: (6)

I'm tuning in Rowlette on 12/15 if you want to come play, very cheap dyno time on an MD-250.

Joe Perez

iTrader: (8)

All this talk about lean AFR, rich AFR...

Rule for the thread: I am running 14.7:1 AFR in all cells except at the very top of the MAP range, and perhaps the very far end of the RPM range. Assume this to be an unchangeable constant. For the purposes of comparing to STOCK OEM IGNITION TABLES, this is the only valid scenario.


That's the big problem I'm having. I found something similar (from a 1.6) in the last page of the "Socketing" thread, and it had the disadvantage of being ambiguous on both the load and advance scales.

I don't even know if a direct, linear correlation exists between load (as computed from airflow) to MAP, much less what the conversion factor would be.



Mind sharing the ignition map?



That's quite interesting.

If I take that table at face value, it means that the OEM ECU on a '97 car is running spark advance in the high teens / low 20s at the very top, which is roughly 20° less than I have on my engine at the moment.

That's what makes me question the data.

I bumped the non-idle rows another 10% this morning, and it's still making great power with no knock.




Just added another 5% (after taking that screenshot) and we'll run that table tonight.

I think I'm going to install a table-switching switch so that I can do instant A/B comparisons between tables.






Two methods:

1: Turn down the radio and listen.

2: If I crack a piston in half, then I'll know I was getting knock that I couldn't hear.

hornetball

iTrader: (4)

Thanks for the offer. I'll definitely try to do that (assuming I've got the car back together by then). I'll PM as we get closer.

I rented some dyno time once before, specifically to tune spark, and it really didn't work out too well. First, the operator gave me a blank stare when I described what I wanted to do with spark advance. His idea was to just download maps from the internet, as if I hadn't already been doing that. Then, after a couple of runs the dyno broke down (was not communicating with the operator's laptop). Ended up wasting a day for no real gain. Whole thing kind of turned me off. Suspect things would be night and day different with a knowledgeable operator.

Braineack

iTrader: (62)

it pretty much mirrors the table I made out of the stock map posted in this thread that was supossedly pulled from eprom data.

both my 2.2L i4 and 2.4L i4 only put out mid-low 20° advance in cruise when I monitor on an odbII scanner.



And that thread is privvy to my awesome quote:

hustler

iTrader: (6)

Yeah, you're in good hands on this deal.

JasonC SBB

37.6° is therefore closer to MBT than 26°.

Just in case nobody has stated the obvious yet, timing at MBT will yield best fuel economy. This is true whether you are running at 14.7:1, or leaner. (note that leaner will move MBT point more advanced due to slower burn).

So one can simply find MBT on the dyno for the cruise range, and be done with it.

If a given motor and octane can run at MBT at WOT, it sure as hell can run at MBT (or even past it), at part-throttle, without knocking.

As for AFR, I do believe 15.5:1 (at its MBT) cruising will yield better BSFC than 14.7:1 (at its MBT). 16:1 may start showing inefficient burning, without special swirl/tumble features in the combustion chamber.

How to find MBT at part throttle? This was my proposal. Set dyno to constant RPM mod, sweep slowly through MAP/throttle position, and plot torque vs. MAP, at various RPMs:

https://www.miataturbo.net/ecus-tuni...-can-do-46160/