Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

LPG ECU mapping

Status
Not open for further replies.

TheKnight

Member
Hello,

I have a Landi Renzo LPG system on my car, and using Landi Renzo Omegas software. After a bad work of a technician the consumption increased a lot and became laggy and jerky. I auto-calibrated it with the software, but I don't know what the values in the mapping mean - or if I have any improvement at all. Does anyone have experience with this or can help me understand what to the numbers in mapping mean - or more specifically how can you tell which numbers are good for optimum performance? I have attached pictures of the software and mapping so you can see what I mean.

Thanks and have a nice day.
 

Attachments

  • 3.PNG
    3.PNG
    65.4 KB · Views: 1,185
I have no clues about programming an LPG system but with the standard all mechanical mixer systems the general setup is to lean them out to where the engine bogs and then richen it up a tiny bit at a time until the bogging effect goes away. That's where your most efficient fuel energy to mechanical energy conversion point will be.

Anything much over that is just wasting fuel to which effect a typical PLG fueled engine can run at over 2X richer mixture settings than it needs and show absolutely no performance problems other than excessive fuel consumption and higher exhaust temperatures.

That's the best I can offer.
 
I have no clues about programming an LPG system
Me either - this is the first time I'm trying anything like this so sorry for any dumb questions.

Anyway, doesn't the mapping determine the mixture - because my car became laggy and jerky I guess it's lean and needs more fuel? So that would mean increasing the values in the mapping?
 
It's far from a dumb question in my book. Especially given much of the stuff that gets asked here on the forums. :eek:

Yes, mapping determines the mixture so as far as adjustment goes It's hard to say.

Typically with my pickups and other propane powered machines running too lean never made the fuel consumption noticeably worse.Or at least not an extreme extent.
It usually dropped it a bit but the power loss was substantial at that point.

For your setup I would say try going one way a fair amount and see what happens then go the other and see. Unfortunately, it's likely going to be a bit of a trial and error method to learn how your system works.

If you go leaner and you have no power then you need to go the other way. If you go richer and nothing changes or it gets worse then obviously you need to go leaner. How far I have no idea.
 
Yes, I was afraid that the only method would be the trial and error and not a method to be sure that you've got the right values.

Thank you for the answer.
 
The real problem is the actual fuel quality and that can have a huge range of effect on things.
One tank you may get excellent MPG numbers and the next is terrible in comparison despite no major change in driving conditions or routines.

I have no idea how the computer system deals with that but on the mechanical systems I just leave it set where it's best with good fuel and accept that many places sell some pretty low-grade stuff at times.
 
I havent setup lpg systems, however I have dabbled with petrol injection and megasquirt.
That looks like a fuel map, the values pertain to an open time for the injectors, for an 8 bit system 255 would be the max.
For a 2d map which that is one axis is engine rpm and the other engine load, load can be calculated from various sensors or sometimes its just the throttle position.
The ecu selects values referencing load and rpm on the map, then trims those values by reading the oxygen sensor to get as near perfect mixture as it can.
If your engine is running poorly I'd expect a warning light to be on, as in this situation the trim constant goes too far one way and triggers an alarm, so I'm not sure whats happened in your circumstance, something has changed and the ecu hasnt detected it.
Theres usually 2 trim factors, long term trim and short term trim, long term is just an integrated version of short.
The actual values are usually determined by tuning the car on a rolling road, not something you can play with as a rule, unless the ecu can do Ve mode where it calculates fuel amount from the air density and amount used.
Hesitancy and jerkyness on a petrol is often caused by a dirty mass air flow sensor, sometimes a manifold vacuum leak, though a leak brings up an alarm as a rule.
Hope this helps.
 
Thank you for the answers.

I always drive with the same fuel, from the same place, although yeah I suppose the fuel is never the same but that is not up to me.

dr pepper - I'm sorry I don't understand what you mean by trim constant - which value is that? And no, there are no warning lights in the car, nor there are any leaks.

My question is, is there like a general rule from which you can determine if the values are good, or just like as mentioned above, trial and error?

I have made a few more auto calibrations with the engine hot and with the engine cool and have attached them for you to see, maybe it helps.
 

Attachments

  • 5.PNG
    5.PNG
    30.7 KB · Views: 625
  • 6.PNG
    6.PNG
    30.9 KB · Views: 392
  • 7.PNG
    7.PNG
    30.8 KB · Views: 424
  • 8.PNG
    8.PNG
    30.6 KB · Views: 383
I assume if the ecu will auto callibrate that it uses Ve or volumetric efficiency mode, the tables all seem consistant, that makes it unlikely to be a leak, the values would be up and down.
Looks like the engine was cold in the first table and warm in the last, presumably you calibrate when the engine is warm.
I dont spose you have tables from when the engine was running well?, that would be usefull.
The correction constant or trim is a value determined by the ecu, it chooses a value from the table, then looks at the exhasut 02 sensor, if the mixture isnt perfect it will trim the values from the table, then store this trim value and use it next time, the table isnt changed the ecu just modifies the value pulled from it.
If the trim is high or low because there is something wrong with the engine, usually higher than 10% you get an alarm.
Another thing that can cause driveability issues is a bad 02 sensor.
Are you able to check for fault codes?, are they even supported, some gas ecu's are very simple and may not do it.
You could try a drive cycle check, the ecu tests various stuff during a drive cycle, drive along the motorway at 60 for at least 5 minutes, then pull off and let the engine idle for 5 more minutes then switch off for 5 minutes, this should make the ecu go through its drive cycle, if something is wrong the check engine light (like I was saying if supported) will come on.
 
Unfortunately I don't have the maps from the optimal engine, and my setup doesn't support fault codes. :banghead:

The o2 sensor is fine, in fact it's a new one, so that's not it.

About the correction constant - then how does it manage to give me new values after I immediately make a new calibration? Does it due to the change of temperature, maybe different reading from the sensors?
 
You are making me glad I have the old school systems where I have one screw for idle A/F mix and one bigger one for open throttle A/F mix. :joyful:

It's so simple an environmentalist could do it. (maybe) :angelic:
 
Alpha N systems favoured by a lot of manufacts is actually quite good, esp later ones with obd2, its not all that difficult to understand.
However they are getting silly these days with multiple networks and ridiculously complicated gizmo's, plus sometimes I wonder if a certain amount of money making is engineered into these systems.
You used to be able to fix a vauxhall nova with a leatherman, nowadays a fairly simple thing on a car over 5 years old can end up with it being on a scrap heap, I know I've seen it many times, how can this be good for the environment, as well as lowering emissions they should make vehicles more repairable.
 
Just curious as to what type of O2 sensor you are using being that a properly set up IC engine running on LPG will have too cool of exhaust temps for normal O2 sensors to work plus has a stoichiometric ratio of ~16:1 - 17:1 which is just outside of a standard narrow band O2 sensors working range to get an accurate and reliable measurement from.
 
Interesting and odd at the same time.

From what I can find that is a one wire unheated narrow band type sensor of which typically do not work at all on LPG fuel engine applications.

The only type I am aware of that do work properly are the multi-wire heated wideband units which are not directly interchangeable with the older simpler one wire types.

Given that I am half wondering if your O2 sensor does anything at all relating to the engine control being on all of the vehicles I have done dual fuel conversions to the stock one wire (unheated) and stock three wire (heated) narrow band O2 sensors always generate codes relating to either being too cool to function or out of range (too lean to read) or both.
 
Yes me too, I would have expected a 3 wire heated 02 sensor on that application.
Is the one wire spec'd by the gas conversion manufact?
 
Hi and sorry for the slow reply.

Yes, the wire was factory. I attached some pictures from the car pdf for you to see if it helps in any way:
 

Attachments

  • o21.PNG
    o21.PNG
    69.3 KB · Views: 558
  • o22.PNG
    o22.PNG
    41.6 KB · Views: 292
I see. They are dancing right on the edge of its ability to sense the leaner A/F ratio.

I've done that cheat before to richen up normal gasoline A/F mixtures by putting a low forward drop germanium diode (.3 - .4 volts) in series with the O2 sensor output to making thing that a ~12 - 13:1 A/F ratio (most power and fuel efficiency opposed to 'cleanest burning'/ not so efficient.) is the correct ratio it normal wants to see.
 
I built this kit some years ago for checking EGO sensors, It has come in handy as ive put long leads on it so it can be sat on the dash. I use it to check that the sensor is working properly, & also check for proper operation of engine when road tested. It will show if there is a too rich or too lean condition. The circuit was in Febuary 1994 Electronics Australia. You might find this usefull in your testing.
EGO TESTER.7.JPG
EGO SENSOR TESTER.jpg
 
Theres an epe version which uses a lm3914 bargraph chip, very handy item, esp if you dont have a scanner, even if you do they sre sometimes not quick enough to see a wide open throttle blip, esp if your looking at other pids as well.
If idling or constant load running a narrow band sensor will continuously show rich/lean, wideband sensors are pretty stable around a lambda of 1, wideband sensors use a different control method and are not as easy to play with.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top