PDA

View Full Version : safc mapping



psbarham
20-09-2009, 07:22 PM
I've finally got round to fitting my apexi safc to the vr4 today. i did some runs with the wideband to see how far out my fuelling was to start with, and i have to say i was surprised how lean it is as stock.

uploaded/1031/1253469899.jpg

we then played with the safc a bit, mainly adding a bit of fuel in the 3-4K rpm range and taking just a little bit out at the top but only 3-6% to be safe and then it looked like this, this was a run through the gears but WOT in 3rd & 4th with boost at 0.60 bar

uploaded/1031/1253470071.jpg

again we fiddled a bit, but nothing drastic and ended up with this, again done with WOT in third

uploaded/1031/1253470679.jpg

and this one was done with 0.85 bar boost in 3rd

uploaded/1031/1253470855.jpg

any suggestions will be taken seriously, and gratefully received as I don't want to blow the bloody thing up

bradc
20-09-2009, 07:53 PM
I don't beleive there is any need to put fuel IN at 3000-4000rpm. The engines never detect knock there in my evoscan experience with a number of VR-4's and if you keep it a bit lean there you will save in fuel and get slightly better performance. Above 4000rpm of course you'll be pushing it so there is no need to worry about fuel economy that high up, but in normal driving it is quite common to get above 3000rpm.

Turbo_Steve
20-09-2009, 08:31 PM
First things first, you need boost on those graphs, as this is a pretty fair indicator of engine load.

Your AFRs should really be mapped to load, not to RPM (though as the SAFC isn't load based, you need RPMs on the graph as well).

On boost your AFRs should really be below 12.
Off boost, as discussed, you're fighting closed loop mode. Forget it.
The area you'll make the biggest difference is, as brad says, between 2500-4000rpms. This is where you're building boost, and you have a range of options.

You can either go rich (the extra fuel means the exhaust gas has a lot of mass....good for spooling...bad power sapping once you're actually on boost, and really bad for fuel economy, and as you're on a piggyback it will probably pull some timing too)


Or you can try and take fuelling out. This is a more interesting proposition: without any monitoring of exhaust temps for both banks, I'd advise against going too far in this direction, however if you keep your AFRs below 12.5 I doubt you'll have any problems. The trick here is that you get lots of lovely timing, and as it's a bit lean, you're getting lots of very hot, high velocity gas. Brilliant for spool, good economy. It can, however, feel unresponsive, as everything has to get hot for it to work. The gas has less mass as you're leaner, and you run the risk of both det and excessive heat build up. It's hard to hold the car in this zone for any length of time without it flipping back to closed loop, so heat build up isn't so much of an issue. Det, however, will lose all that timing. And, of course, you're on cast pistons, so every little knock event significantly ages them.

In your circumstances (i.e. very little in the way of monitoring) I would be inclined to always err on the side of richness.
You're mainly only going to be concerned with the fuelling when you're on boost, as the SAFC is more about power than economy.

Dom B
20-09-2009, 08:48 PM
I fitted a unit to a volkswagan golf running a grafted on motronic a few years back and at low loads you could set a manifold pressure and rpm (by pots only, nothing digital back then) but what it would do is at the selected point it would click the lambda signal over to a false squarewave signal crossing as expected but generated inside the unit itself. This fooled the ecu into allowing a very crude air meter mapping unit which is kinda what the a piggyback ecu is, this meant that the closed loop mode didn't fight the changes. I dare say some of the fancy piggybacks can have this function built in and do it with graphs instead of old potentiometers.

psbarham
20-09-2009, 08:58 PM
its very difficult to get the lm1 to read the 5v signal from the apexi MAP, we wasted about 4 hr's doing this on the brick'o'doom last year with no success even after i built a pressure chamber and regulator to give a measured constant pressure to map the MAP with, which is why i put the rpm on the graph as an indicator.

so a bit richer on the lower end and leave the top end alone then? so if i aimed for the high 10's/low 11's upto 4 k and then leave alone as the little 'ole tubs are running out of puff after then anyway, and tbh i very rarely stray above 5k anyway due to driving style and mechanical sympathy.

Gowf
21-09-2009, 12:36 AM
What you want paul is around the 11.2-11.5 mark. Its safe enough at that. Right up the top end, ok if you wanted to be safe then a very high ten would be ok.

When lower down you can leanspool these and it does help to reduce fuel cut, but be careful how lean you go.

At 0.6bar, you'l be fine (you'd be shocked what i run afr-wise, but thats for me and im fully aware and prepared if it lets go, but currently i have no knock) at maybe a low 12, but dont go any leaner than that, if you want it to last. Are you monitoring it with det cans?

What you'l find is that the ecu detects knock when there isnt any, its a PITA, and so then retards the timing.

On cruise is a totally different matter, but if your not having a load vs rpm map then its a bit pants. Although, are you using an alpha- n (throttle position as load) map?
You can go bloody lean on cruise, but 14.7 is what your after although if you have the stock lambda sensor still plugged in it will do that anyway. But you can go leaner, it only lean misses at about 18afr, but thats maybe a little extreme for most, but its the only way i get any economy out of mine, and even then its only about 28mpg

psbarham
21-09-2009, 07:51 AM
Cool, much appreciated everybody, :afro:

So Gowf how far out looking at the graphs would you say the fuelling is and where would you fiddle with it? at the moment the only monitoring i have is the lm1, which is not a permanent install, which is why I'm veering on the side of caution.

Turbo_Steve
21-09-2009, 09:08 AM
around the 11.2-11.5 mark.

Do these run especially hot? (let me guess....the back runs hotter than the front?). Low-mid 11's would seem cautious to me, but then your knowledge (and more importantly, experience!) of these is far more extensive than mine?

Eurospec
21-09-2009, 11:08 AM
They do build heat quickly- yes.

I tend to set the road cars up around there. I know there is power to be had, but its not that much in my experiance and it is a good bit of saftey.

Cheers,

Ben.

Turbo_Steve
21-09-2009, 02:05 PM
That's interesting: thank you.