How to make the move to an aftermarket ECU

In reply to Keith Tanner :

Probably.  As I've mentioned before, I've still not broken one of my instrumentation UEGO's after a few decades of testing.  And recently, I've taken to turning on the on board sensor before even cranking to replicate newer sensors, and I've still not broken one of those, either.

The two times I have damaged a sensor- the first was on a race truck that ran leaded fuel, and the second was a test where the exhaust extractor failed on full blast- pulling oil through the turbos and killing my exhaust.

Add to that- the last time I drove my Alfa below freezing was when we bought it and drove it home from Columbus in 1996.

My LC1 (LSU4.2)  heated in about five seconds.  The LSU4.9 systems seem to take a lot longer.  Maybe they are being more delicate with the sensor?

I do like that the AEM utilizes the calibration resistor. Innovates require a fresh air recalibration with a new sensor, they ignore the calibration resistor.

There is a heating applied voltage vs time curve for the 4.9 sensor in the data sheet and its not crazy fast....but I guess why would you need it to be?  OEMs need to meet tight emissions specs so I can see how they might work hard to figure out just how fast they can start getting O2 data, and I imagine try different sensor location and lots of different warmup curves.  I'm not sure what the value on a race car or hotrod of that would be though vs taking a conservative approach and knowing you aren't damaging the sensor?

I've not played with LSU ADv sensors, I guess the read faster but not idea if you can heat them up faster without fear...but still it would seem a conservative approach to heating on a custom install is mostly upside.

In reply to mke :

Being closed loop ASAP is better than not.   Trying to calculate the lost fuel estimate on a cold engine in a consistent manner takes a very long time and a whole lot of starts.  Even if it's conservative, it doesn't take much of a mistake to be a problem.  And for my "fantasy" application, I will add a catalyst- so making sure I have no misfires by being closed loop is a high priority.

Yes, emissions are much more robust, but one rule I've learned- if the a/f control makes really good emissions, it also drives really well. 

Pete. (l33t FS) said:

My LC1 (LSU4.2)  heated in about five seconds.  The LSU4.9 systems seem to take a lot longer.  Maybe they are being more delicate with the sensor?

I do like that the AEM utilizes the calibration resistor. Innovates require a fresh air recalibration with a new sensor, they ignore the calibration resistor.

That would be odd- the larger the number, the newer the part- and the newer the part, the faster we have been pushing them to use.

And just to kind of restate, for people that like to mess with stuff like how and when the WBO2 sensor heats up enginelab might be for you.  The AEM infinity HW is 0, 1, 2 onboard controllers in the 3, 5, 7 HW respectively and the enginelab software can be loaded at no cost, just register the HW on the enginelab site and then the world is yours. 

Right click, add if-else, and select the things you want to happen before the heater starts

Right click add table and define whatever fuel control you can dream up. or add 12 tables or 20 tables.  Everything I've been able to dream up hasn't used even 1/2 the available memory space I don't think.

Right click, add CAN packet, define the packet

No actual C programming, its a menu driven logic level, so If yo can say, I need to read the MAP sensor then smooth the reading, then convert from voltage to kPa,  then you can probably do it.   No compiling and you can change the model on the fly in the ECU.  There is also a scripting language for those that understand how to use it, which does not include me so I use the menu stuff.  Writing a whole model is a lot of work, but I'm happy to share mine with anyone who wants it as a starting point....its pretty frikin cool once you get the hang of it.

alfadriver said:

Pete. (l33t FS) said:

My LC1 (LSU4.2)  heated in about five seconds.  The LSU4.9 systems seem to take a lot longer.  Maybe they are being more delicate with the sensor?

I do like that the AEM utilizes the calibration resistor. Innovates require a fresh air recalibration with a new sensor, they ignore the calibration resistor.

That would be odd- the larger the number, the newer the part- and the newer the part, the faster we have been pushing them to use.

The innovate brand controllers are kind of know for frying sensors so  maybe not a surprise.

In reply to alfadriver :

Funny, I almost never use closed loop on anything....just 1 more thing to go wrong.  I just err on the rich side during warm up and it always drives nice.  I have used closed loop under cruise conditions but it seemed more trouble than it was worth.  I love the idea, at least for lower power,  just not the work to really make it work well.

I'm not sure what other ECUs also have it these days, but there is misfire detection capability in enginelab.  I've kind of got it working but just haven't spent enough time to really get it right.  It shows me when all it well and an occasional misfire, but once things are misfiring all the time it kind of stops seeing them, but at that point lambda shows it clearly, at least in open loop.

Closed loop makes it much easier not to go wrong. It is the key to everything. You just have to make sure the ECU is getting good data.

In reply to mke :

You should try closed loop.  It was a pretty amazing revolution for OEM's to have non-stoich closed loop- make more power and at the same time keep the exhaust cool enough with less fuel.  We almost have closed loop exhaust temp control with that.  My experience with that is why I want closed loop ASAP and all of the time.  Adding the ease of calibration just makes it better.

Sensors only fail when the heater cracks and when the surface is poisoned, other than that, they work really well and are incredibly reliable.  

And it's kind of ironic you worry about one more sensor going wrong when you have 12 MAP sensors.  Just pointing that out.

Keith Tanner said:

Closed loop makes it much easier not to go wrong. It is the key to everything. You just have to make sure the ECU is getting good data.

And that is why I like the FAST, it is closed loop all the time and VERY good at keeping to desired lambda.

At least it works well for single O2 applications like turbo Buicks.

alfadriver said:

In reply to mke :

You should try closed loop. 

As I said, I have used it in the past but just just never found it to add value.  Once the calibration was done, it was making +/- a couple %...mostly to pull out the fuel is just added vice versa so it was much more stable with it off than on.   I do have a closed loop lambda roughed into my model so if a need arises I can finish it up and give it a try, but I'll wait until I see a need but for now I'm just logging lambda error to speed up calibration work.  Its not that have anything against letting the ECU change things, I do have the ECU trimming the cylinder fuel based on MAP.....and that seems enough for now.

alfadriver said:

Pete. (l33t FS) said:

My LC1 (LSU4.2)  heated in about five seconds.  The LSU4.9 systems seem to take a lot longer.  Maybe they are being more delicate with the sensor?

I do like that the AEM utilizes the calibration resistor. Innovates require a fresh air recalibration with a new sensor, they ignore the calibration resistor.

That would be odd- the larger the number, the newer the part- and the newer the part, the faster we have been pushing them to use.

My neanderthal brain sees that the 4.9 sensors are smaller and more expensive, which equates to "more fragile" to me.    Certainly I never had any problems with the 4.2 sensors from cold shock, since it was SOP for me to turn the ignition on, wait for the sensor to warm up and start reading, and THEN start the car, so I could see what it was doing on a cold start.  The only thing that would kill the sensor was rapid and flamboyantly spectacular ingestion of a half quart or so of engine oil when I'd let the oil level get too high, the sensors don't like that very much.  Then again neither did the local mosquito population, or the other people on the road... even coal rollers would be all "Dude WTF".

#wankelissues

mke said:

alfadriver said:

Pete. (l33t FS) said:

My LC1 (LSU4.2)  heated in about five seconds.  The LSU4.9 systems seem to take a lot longer.  Maybe they are being more delicate with the sensor?

I do like that the AEM utilizes the calibration resistor. Innovates require a fresh air recalibration with a new sensor, they ignore the calibration resistor.

That would be odd- the larger the number, the newer the part- and the newer the part, the faster we have been pushing them to use.

The innovate brand controllers are kind of know for frying sensors so  maybe not a surprise.

Because innovate is too cheap to use the bosch control chip. 

To go with alfa's q a few pages back I have an odd control strategy on mine that works for me. I use a separate relay, and turn it on at rpm > 500 or tps >90 so if I want to calibrate it I can just floor it with the car off. On ms3 likely you can define a custom loop with coolant temp vs time as well to define the output logic. I would have to think a little on how to implement but it should be do-able. 

Agreed on the i/o and features always nice to have spares available. 

Paul_VR6 (Forum Supporter) said:

On ms3 likely you can define a custom loop with coolant temp vs time as well to define the output logic. I would have to think a little on how to implement but it should be do-able.

I can't help myself, I'm going to show you how easy custom can be

Target>Add Channel>NoOperatioon>

TRhe winow opens and you give it a name, I called it WBO2_Sensor_Delay_Time and in the inial value I set 5 for 5 seconds

Then Target>Add Channel> If-Else Condition> RunTime>WBO2_Sensor_Delay_Time ( You don't type the variables, you just select from the menu)

Then Target>Add Channel>Hardware IO> select the if-else you just added, select the driver you want to use

Now the ECU is literally reprogrammed

 If you want to see the new variable and be able to set abjust in on the tuning software,  

Right click on the page and select  whatever tuner page you want to add it to, select Text, selected WBO2_Sensor_Delay_Time for the menu and that's done.  You can set the display up however you want.  Its REALLY easy to get used to having the ECU do whatever please

A close friend of mine is currently trying to find a good aftermarket ECU for his current built and was really lost on where to start. Since reading this he has started to make a plan and looked into a few options.  

Paris Van Gorder said:

A close friend of mine is currently trying to find a good aftermarket ECU for his current built and was really lost on where to start. Since reading this he has started to make a plan and looked into a few options.  

It's going to really depend on the car and engine; what sort of project is this?