ECU tuning

[jambo_gtir]

Mainly a question for those of you who re-map ECU's really (Fusion Ed).....

Am I right in thinking that when you re-map say a NisTune ECU for example, you remove the wide-band sensor once mapping is complete and the standard lambda sensor (narrow-band) is put back in?
So under load the ECU is in open-loop and still doesn;t care what the sensor is telling it?

Or.... do some of these after-market ECU's run closed loop all the time with a wideband?

Reason being, I'm designing my own one at the moment and wondered whats out there at the moment.

Cheers.

 

[Trip]

I doubt any ecu (aftermarket or not) will run on Closed loop at full load.

 

[jambo_gtir]

Modern cars do for emissions reasons, the ECU is constanty adjusting fueling.
The benifit for our cars would be better fuel economy and to help keep the engine safe (closed loop even under full load would stop engine running to lean and causing problems).
Would anyone actually be interested in something like this?

 

[PobodY]

I'm with Trip. - I think you can leave it connected and read what it's telling you about the mxiture, but the ECU is ignoring it; which ECUs take a reading from a wide-band anyway, don't they rely on the AFM to set the injector map?

EDIT: Too slow. It's an interesting proposition; you're actually looking at making your own PLCs for this?

 

[jambo_gtir]

I'm with Trip. - I think you can leave it connected and read what it's telling you about the mxiture, but the ECU is ignoring it; which ECUs take a reading from a wide-band anyway, don't they rely on the AFM to set the injector map?

EDIT: Too slow. It's an interesting proposition; you're actually looking at making your own PLCs for this?

Yes but the maps are set up using a wide-band in the first place if you see what I mean.
I'm thinking that if i made it closed loop all the time, i can over-ride the base map slightly if it conflicts with whats actually going on in the engine.
Yer I'm doing this for a big uni project. Was just seeing if people would be interested really lol.

 

[PobodY]

So you'd need a 3D map?
Airflow Vs injector duty (or timing depending on which you're looking at) as standard, but with the lambda giving you a third axis of "fine" adjustment on each point in the base map?

How would it deal with a failure in one of the circuits? I guess if it's the lambda you'd just get the base map.

In the same way that you were saying reading the AFM is too early and so map adjustments should be delayed, isn't the lambda too late and any 'bad' mxiture events have already happened? - I'm just throwing it out there.

Interest, like most things, is probably dependant on price.

This sounds like the work of a process control engineer.

 

[jambo_gtir]

So you'd need a 3D map?
Airflow Vs injector duty (or timing depending on which you're looking at) as standard, but with the lambda giving you a third axis of "fine" adjustment on each point in the base map?

The wide-band sensor in the exhaust would 'read' the AFR. The perfect AFR for petrol engines (stoichiometric ratio) is 14.7:1; so i'd be adjusting the injectors constantly to keep to this value (or near as possible). The circuit would intercept the signals from standard ECU. If the circuit failed then like you said... it would just be the standard ECU carrying on as normal so it would be pretty fail-safe.

In the same way that you were saying reading the AFM is too early and so map adjustments should be delayed, isn't the lambda too late and any 'bad' mxiture events have already happened? - I'm just throwing it out there.

Yes spot on mate, thats why I thought moving the AFM too close to engine is a bad idea.... But like you say, on the exhaust side your dealing with the past (something that may have happended 2 cycles away) and you can't change the past. But it would just be a way of keeping the engine safe really, hopefully lol.

 

[jambo_gtir]

....Assuming that If you start running way too lean, the sensor would pick it up quick enough to do something about it. This is the only way as you have no way of measuring what is truely going on before the engine.

 

[red reading]

If you run a turbo engine at stoic you will melt the piston's etc, be carefull with what you are designing the principal is very good.

 

[PobodY]

If you run a turbo engine at stoic you will melt the piston's etc, be carefull with what you are designing the principal is very good.

Should it be about 11:1 when you're on boost? - I remember it being something like that last time my car was on a rolling road.

 

[Trip]

....Assuming that If you start running way too lean, the sensor would pick it up quick enough to do something about it. This is the only way as you have no way of measuring what is truely going on before the engine.

What about if one of the Injector is blocked ? the WB02 will detect lean condition and add more fuel to compensate. What it is actually doing is flooding fuel to the other 3 cylinders.

 

[Trip]

Should it be about 11:1 when you're on boost? - I remember it being something like that last time my car was on a rolling road.

More on the high11 low12. It depends on other parameters too.

 

[PobodY]

Hang about... are we proving free product development by means of peer review? :shock:

 

[williams]

The wide-band sensor in the exhaust would 'read' the AFR. The perfect AFR for petrol engines (stoichiometric ratio) is 14.7:1; so i'd be adjusting the injectors constantly to keep to this value (or near as possible). The circuit would intercept the signals from standard ECU. If the circuit failed then like you said... it would just be the standard ECU carrying on as normal so it would be pretty fail-safe.



Yes spot on mate, thats why I thought moving the AFM too close to engine is a bad idea.... But like you say, on the exhaust side your dealing with the past (something that may have happended 2 cycles away) and you can't change the past. But it would just be a way of keeping the engine safe really, hopefully lol.

ive bought 4 books on mapping as im wanting to teach myself to do it so i can map my emanage myself. and from what i understand an afr of 14.7:1 is the perfect 'burn', but this doesnt mean its best for power or reliability.
from what i have read best approx afr for power is 13.2 - 13.5, but on a turbo engine your probable better of going a bit richer even up to 11.5 higher up the revs, this will keep temps down and help prevent knock.

some ecu's can run with a wideband to try and meet target AFRs which you would set yourself.

im no expert this is just from what ive read.

 

[iomegalinux]

some ecu's can run with a wideband to try and meet target AFRs which you would set yourself.
.

this is a fonction in the emanage ultimate, but i did not tried it.
you target what afr you want, so it constantly tune the map for that target.
it help getting a base map quickly.

 

[watoga]

As has already been said, an A/F ratio of 11-12 is the "ballpark" figure, but will depend on other parameters such as engine load and ignition timing.

Dave

 

[Fusion Ed]

Do not try closed loop wide band from e-manage, it will get its fuel table in a complete mess.

As for GTi-Rs yes closed loop at light throttle, and open loop at medium heavy throttle. I.e. its closed loop where it needs to be 14.7:1 but open loop all other areas.

You should be able to cruise at say 80mph and still be in closed loop at light throttle at lambda=1. You attempt to get as much light throttle in closed loop as possible to get best fuel economy off boost.

I will always leave the narrowband in place if at all possible when tuning. Then I can check that the ecu is not having to make much compensation or correction to the main map to ensure the injector/maf scaling etc is correct.

Its now more common than not that modern cars will run closed loop all the time and in fact cars like the Micra/Primera/Almera run closed loop pretty much all the time as well, and that's going back ~18 years... But some modern turbo cars will also run closed loop and even at Lambda =1 for the entire fuel map (Focus ST IIRC), and will do this without melting etc. Obviously there is a negative performance impact, but a positive emissions one.

You can also map leaner than lambda = 1 for part and light throttle and gain economy but you need to watch EGT but its usually fine for off boost/ almost no load and can really pick up some mpg there. As an example I tried this on my P11GT and averaged 47mpg on a 280 mile journey, where usually Id get about 40-43.

Modern after-market ECUs can do full time closed loop too. The aim is to build up a correction map/value to the main one table, but it takes a while to establish and you have to give them boundaries to stay within. To get this working right however the main maps have to be good. Don't be fooled it would be capable of doing the mapping for you.

As for e-manage you have to remember that everything you do on one of them is NOT absolute. For example injector resizing is done by entering old vs new injector sizes, then the e-manage makes a calculation which is and can only ever be approximate. Its not unusual for the real values to not work right at all, and then have to compensate one way or the other to get it right.

Another thing with the e-manage. Less is more, its rare IMHO that I see an e-manage map where people seem to properly know how they work or what they are doing. They forget that they are simply correcting the main map, not re-writing it, and as its correction values you need, they tend to be quite simple in order to get them working perfectly, they only get more complicated when you have to clamp the main ecu (to get around load/boost cuts) then the fuel/ign maps begin to look like normal ones as you have essentially locked the main ecu from that point on. But again, you have to remember this and only work with the parts after the point in which you clamp (if its even necessary) in this way.

Ed

 

[PobodY]

Thanks Ed, that has cleared a few things up for me.

 

[williams]

Do not try closed loop wide band from e-manage, it will get its fuel table in a complete mess.

As for GTi-Rs yes closed loop at light throttle, and open loop at medium heavy throttle. I.e. its closed loop where it needs to be 14.7:1 but open loop all other areas.

You should be able to cruise at say 80mph and still be in closed loop at light throttle at lambda=1. You attempt to get as much light throttle in closed loop as possible to get best fuel economy off boost.

I will always leave the narrowband in place if at all possible when tuning. Then I can check that the ecu is not having to make much compensation or correction to the main map to ensure the injector/maf scaling etc is correct.

Its now more common than not that modern cars will run closed loop all the time and in fact cars like the Micra/Primera/Almera run closed loop pretty much all the time as well, and that's going back ~18 years... But some modern turbo cars will also run closed loop and even at Lambda =1 for the entire fuel map (Focus ST IIRC), and will do this without melting etc. Obviously there is a negative performance impact, but a positive emissions one.

You can also map leaner than lambda = 1 for part and light throttle and gain economy but you need to watch EGT but its usually fine for off boost/ almost no load and can really pick up some mpg there. As an example I tried this on my P11GT and averaged 47mpg on a 280 mile journey, where usually Id get about 40-43.

Modern after-market ECUs can do full time closed loop too. The aim is to build up a correction map/value to the main one table, but it takes a while to establish and you have to give them boundaries to stay within. To get this working right however the main maps have to be good. Don't be fooled it would be capable of doing the mapping for you.

As for e-manage you have to remember that everything you do on one of them is NOT absolute. For example injector resizing is done by entering old vs new injector sizes, then the e-manage makes a calculation which is and can only ever be approximate. Its not unusual for the real values to not work right at all, and then have to compensate one way or the other to get it right.

Another thing with the e-manage. Less is more, its rare IMHO that I see an e-manage map where people seem to properly know how they work or what they are doing. They forget that they are simply correcting the main map, not re-writing it, and as its correction values you need, they tend to be quite simple in order to get them working perfectly, they only get more complicated when you have to clamp the main ecu (to get around load/boost cuts) then the fuel/ign maps begin to look like normal ones as you have essentially locked the main ecu from that point on. But again, you have to remember this and only work with the parts after the point in which you clamp (if its even necessary) in this way.

Ed

yeah, i was only planning on using the wideband for datalogging not running full time with the ecu.

 

[jambo_gtir]

cheers for the info everyone, was what I was hoping for. I will have no problems with the design and build of the circuit as this is what I do; the problem for me is I am new to engine tuning so I can only really go by what I read off the net and you helpful people. I'm not going to market this or anything, just wanted something nice and complicated to play around with and get me a good grade. But if the circuit does work I will be fitting it to my car.
The whole idea I have is mainly for protection of the engine, not necessarily more power, although that would be nice too lol.