Ready to ditch that old stock o2 sensor and run your Renix beast off a high tech wideband sensor? Well today we’ll learn how with the trusty REM!
Note: Images and graphs shown below are from my 1989 Jeep Cherokee running a Stroked 4.6L engine and aftermarket injectors. A stock setup may look a little different, but information is still the same.
What is an Oxygen Sensor?
So, for starters let’s take it from the top. The Renix Fuel Injection System uses an Oxygen Sensor (o2) in the Engine Exhaust Stream to monitor the combustion process. It can detect the amount of oxygen present after combustion to determine if the engine is running efficiently or not.
In simple terms, it senses Oxygen so a High Voltage = Higher Oxygen Content, which we infer as less fuel which we call Lean. Low Voltage = Less Oxygen which is more fuel which is Rich. We can represent and compare the amount of Oxygen in the system with an Air Fuel Ratio (AFR) to understand the current state of the engine better.
The ideal Air/Fuel Ratio that gives us a “perfect” burn is 14.7 parts air to 1 part fuel. This is where the Stock Narrow Band o2 sensor is set to operate at to help the engine maintain this Stoichiometric AFR. AFR is an interesting spectrum though because you won’t make the most power at 14.7, but you will have some of the cleanest emissions there. Lower AFR of around 12 to 1 is where the best power will be found, and in fact, this is around where the engine runs in OPEN LOOP which we’ll learn about next.
What is Closed Loop?
When the engine first starts up, the ECU runs in OPEN LOOP which means it uses a base fuel curve to run the engine and just guesses how much fuel to use. Generally it will run pretty rich to make sure the engine stays running which is good for power, but it’s not great for economy. During this time, extra fuel helps to warm up the o2 sensor and catalytic converter so that it can do its job properly.
After 30 seconds or so, the o2 should be ready so the ECU will attempt to run in CLOSED LOOP which monitors the o2 sensor and adjusts the fuel curve until it runs at the target AFR (Air/Fuel Ratio) of 14.7. In this mode the ECU will start pulling fuel and wait for the o2 sensor to raise above 2.5V (Lean) which means it has crossed the target AFR. Then it will add fuel until the sensor drops below 2.5V (Rich) and this cycle continues so that the engine bounces right around the target value.
The ECU has a fail-safe so that if the o2 sensor does not respond after a certain amount of time, it will revert back to OPEN LOOP to keep the engine running without leaning it out to the point of stalling. Using larger injectors, bored throttle body, or a larger engine displacement may offset the fueling needs enough that the engine can not run within the factory closed loop limit so it’s important to keep your fuel injection system balanced!
What is a Fuel Trim?
An essential tuning value to understand in the closed loop system is the Short Term Fuel Trim or STFT for short. The Short Term Fuel Trim tells us how far off the base fuel curve the engine is currently running to maintain stoic. For Renix, the base value is 128. Lower values mean less fuel is needed, and higher mean more fuel is needed. This tells us the exact amount of adjustment needed to run correctly which can help diagnose the health and balance of the system.
In figure 1 above, we can see that STFT actually went all the way down to 0 before recovering and achieving a balance. If it hits 0 or 255 for too long, it will detect this as an issue in the system and go back to open loop to try again later.
The Long Term Fuel Trim (LTFT) is a companion to the STFT and helps us get a better picture of the system under all conditions. As the name would imply, this gives us a long term view or average fuel trim to see overall how out of whack it is. Again, a target of 128 is desired, but a perfect 128 usually means the system has not run long enough in closed loop to actually get a reading. A high reading means it usually needs more fuel then stock, and lower needing less. This can help indicate if things like injectors are over or undersized, if it’s even able to stay closed loop long enough to work.
Wideband vs Narrowband
So now that we understand the basics of the stock system, we can look into improving it and using it to our advantage. When tuning a system, the AFR is a critical component in making sure the engine is running correctly in all conditions. Run too rich and you’re just flooding the engine. Run too Lean and you risk damaging the engine with knock while making no power.
Ok great, so how do we get an AFR reading then? Well from the o2 sensor of course. Generally, you can map the o2 sensors Voltage Output to a specific AFR and now we’ve got a reading to play with, but there are a few limitations with the Stock o2 Sensor.
Most basic closed loop systems like Renix work in an on/off or rich/lean setup. The computer doesn’t care about the specific AFR of the engine, it’s only goal is to get the engine to run at the cross over point and that’s it. As such, the sensor has been designed to help give the computer a nice On or Off signal in a very narrow window. This works well enough for the computer, but it’s not very helpful for accurate tuning.
The stock o2 Sensor used is known as a Narrowband Sensor because it’s AFR window or Band is very small. On top of being a small range, it’s also inaccurate, varying a bit with exhaust temperature and unburned fuel. It’s band reading is also not fully linear, looking more like an S curve. All of that combined means we have a fairly useless sensor for accurate tuning needs.
Below in Fig 2 you can see what a narrow band signal looks like at idle. Notice the steep cut-offs at the top and bottom which help achieve an On/Off effect. Through this plot we can see that the useful band is from 14.3 AFR to 15.1 AFR. The horizontal floor and ceiling vary too much to properly use, but even that would only give us 13 to 16 AFR.
We can overcome all of these issues with an aftermarket Wideband Sensor usually known as a UEGO (Universal Exhaust-Gas Oxygen) sensor. The main advantage of a Wideband sensor as you may have guessed is that is has a much larger AFR window which gives us a much better picture of how the engine is running in all conditions, especially under heavy loads. This is where a good tune matters the most, as a small adjustment could make a noticable difference.
Below in Fig 3 we can see the limitations of the stock narrowband signal and how much more can be seen with a Wideband kit. Here it’s easy to achieve an 11 to 18 AFR reading and see what’s really going on.
A good UEGO Wideband sensor such as the AEM X-series kit that I used boasts a laboratory grade o2 sensor which provides a fast and accurate reading from 8.5-18 AFR, quick warm-up time, and most importantly a linear AFR output which we can use for tuning and logging purposes. With this sensor in place, we can finally start to tune with confidence.
Just for fun, below in Fig 4 we can see the 2 sensors pitted against each other during a 15 min drive. The Wideband sensor is providing the baseline AFR reading which is mapped to the stock sensor at that time. We can see that the Narrow band varies up to an entire AFR unit or more in the narrowband operating range, and we won’t even talk about how far off it is at the floor and ceiling. The worst part though is the area circled in Red, as this is where the stock sensor got flooded out from a WOT pull and just stopped working entirely until it burned off. In a High Performance situation this could cause some serious trouble with a wildly incorrect value reported.
Using a Wideband Sensor in a Narrowband system
So now that we have a good idea on the benefits of monitoring a wideband sensor and seeing the AFR, let’s get to actually hooking this thing up! The first thing to understand is that this is NOT going to be a simple plug and play setup because a Wideband sensor typically needs some sort of Driver to make the sensor work. The old Renix ECU is not programmed to take advantage of a wideband sensor reading, nor is it designed to even drive a wideband sensor, so we will have to do some middle man converting to make this work.
For this, we will use an REM to read in the AFR signal from the new sensor, which will convert it into a narrowband signal and then be routed into the factory wiring. This is known as a piggyback tuning setup as we are feeding the computer modified sensor data to encourage it to run the way we want.
If you’d like to follow my exact install, I have a youtube video on the whole AEM kit here:
And this video goes over hooking it up to the REM. Feel free to skip all the trial and error of me trying to figure this out for the first time.
The First Step is driving the sensor and getting a reading. The AEM X-Series 30-0300 kit includes a Wideband Sensor and a Gauge which also functions as the sensor driver. Cool, so we can thread this wideband sensor directly into the stock o2 exhaust bung and make it work with the included gauge/driver. I added a 2nd bung for dual sensor testing, but that is not required for this conversion.
The gauge is also where we will get our AFR reading from in the form of an analog signal. The manual specifies it as a “0-5V Output Gasoline value from 8.5:1 to 18.0:1 AFR” which is perfect for reading in with the Renix Engine Monitor. We will connect the gauge analog output to a free REM analog input which can be accessed from the In/Out port of the REM Diagnostic Adapter or from the optional Bonus Mod Jack on newer models. It’s important to make sure everything is wired up properly and in spec, so if you have any questions at all please feel free to ask away here.
Once the output has been interfaced with the REM, we can set it up and read it in. In Renduinix v0.089, go to Options > More > Settings > Advanced Menu and scroll to the Wideband options. Here you will find Wideband Read which sets the AFR Profile. Check your specific gauge to find out the profile, but for this kit we will pick “8.5 to 18 AF.” Next we will move to Wideband Pin so the REM knows where to look for the signal. This will depend on how you have yours hooked up, but using a 0-5v analog pin with yield the best results. For REM pin support, you can check HERE.
Now that the REM has been configured, we can look for our new AFR reading. In the gauges page, you can now scroll to AFR which should report the same reading as your wideband gauge. In the diagnose menu there is now a Wideband Page which we can use to actually tune the setup. If your REM is reporting an AFR value then we can move to the next step which is interfacing the REM with the vehicle.
The REM has a Signal Out pin which will output our new narrowband signal for the computer to read. Check the REM pin support page to properly access your Signal Out pin. We will be connecting the Signal Out pin to the factory o2 signal wire so that the REM can directly feed the ECU the new signal to use. If you have a dead o2 laying around, feel free to clip the sensor and use the old plug so you can swap back to factory later if you need to.
If everything was wired up properly then you should notice the o2v reading updating while the engine is running. Due to some silly factory glitch, the o2v reading WILL NOT report a correct value when the engine is off, as instead it shows a scaled version of the TPS instead.
Piggyback Tuning Capabilities
Ok, so now that all the wiring has been hooked up we can enjoy the fruits of our labor and actually start tuning. Currently the REM only provides an Adjustable Target AFR which is available in the diagnose page. Here you will find the currently reported AFR as well as the ECU reported voltage on the first row. Both of these should be changing correctly to verify the system.
The second row shows the Target AFR (TG) and the Target Range (RG) which are what we can actually modify. Target AFR is the most important as this is where the ECU is attempting to be. Stock target is 14.7AFR, but you can set it lower for more power or raise it for better fuel economy. This value shouldn’t be changed too far or too fast unless you are familiar with AFR readings and their effect on the engine.
Next up is the Target Range which tells the REM how to simulate the signal. The range is how far the lowest AFR value is from the highest AFR value, the stock sensor having a range of about 0.5 AFR. A smaller Range will make a tighter signal which will make the engine vary less but too small of a value may cause the computer to reject the signal or miss the cross over. A larger range will increase how far the computer has to adjust to balance the reading, but too big will probably make it run unstable or hunt.
Target AFR = 14.7, Target Range = 0.5.
Simulated Signal: 0v = 14.45 AFR, 5v = 14.95 AFR.
Now that you’ve got it dialed in, you can make some general adjustments to how the engine runs and know that it’s running off of a more superior sensor, so congrats! Generally if I’m fooling around offroad I’ll set the Target down to 14 AFR for more power, or if I want to hyper-mile on the freeway then I’ll set the target to 15.3 AFR to save a little on gas.
Downfalls and Future Ideas
So, that sounds like an awful lot of work just to slightly adjust the engines target AFR, and you’re right. One of the limitations with piggyback tuning is that you are still confined to the properties of the sensor you are messing with. The Renix ECU is designed to run off of a hard programmed fuel map and fooling with the o2 signal only lets us bounce back and forth within a new specified target AFR, but if the computer doesn’t enter closed loop then all this work is useless.
The best case scenario is that I can work on programing a variable AFR mode that will set a changing target for the intended driving mode. If you want to horse around then we will have it slowly dial the AFR down so you can get some more bang for your buck with small pulls. If you’re looking for better mpg then we can have a cruise mode that will slowly dial the AFR up when it notices light, stable throttle and high speeds so you can save a little gas under low power situations. This will take some testing to see how quickly I can vary between modes but it would give us a slightly more useful tuning situation. In a perfect world we could force it into closed loop permanently, but that’s not gonna happen without altering the factory code.
One of the biggest downfalls of this setup is when the computer enters Wide Open Throttle (WOT) mode, which is probably where most users would like better tuning in the first place. Under the stock Renix setup, the computer reverts to Open Loop under WOT mode which allows the system to run at a much richer AFR then it normally would in closed loop. This does help us a bit, because otherwise we wouldn’t make very much power when we step on it, but this also doesn’t provide us with any sort of control over it either since it’s just using the stock tune.
In order to really feel the effects of a full fledged tune, we will have to mess with a sensor that is always in charge of running the engine. My number one target is the MAP sensor as that is directly linked to how much fuel is being commanded by the ECU. If we look at a sensor plot of the engine running, we can see that the injector pulse directly follows with the map sensor, which gives me a good indicator that we can use if for piggyback tuning.
Now, fear not, because all this work to setup a wideband sensor wasn’t for nothing. We will still need to know the proper AFR for tuning, and now that it’s in the REM it is easy enough to play with. The next step will be using the AFR input and a fake MAP output to try and directly adjust the fuel curve of the computer and make it run the way we want to all the time. This will take a lot of time to code and test, but it seems like the best plan of attack if we want to keep the stock Renix System in tact.
Now, there is one other options, but it is a lot more involved. This would involve opening the Renix Computer and replacing the chips that contain the factory fuel and timing maps with ones that we can program. Then we could actually program a proper tune without mucking with any of this silly piggyback stuff, but that is also not something that the average motorhead can pull off so you win some and you lose some. That also requires finding replacement chips that would work and figuring out how to program them properly so the computer will accept them, but i’m sure with enough people it could be figured out in no time. I’m pretty sure there are some french folks that already offer this service but I haven’t looked into it much yet.