rusefi.com

Many large V engines have two parallel exhaust systems. How do we handle things in terms of WBO/AFR tuning - one sensor? Two sensors? Do we need any rusEFI logic to validate sensor match? Do we need any logic to average two sensors etc etc?

A few things should happen:

1. Support a second wideband sensor
2. Select which bank each cylinder is on (we can't guess because different MFGs use different cylinder numbering schemes)
3. Run a separate short term fuel instance on the second bank/sensor

I've been planning to implement this for a while, as soon as I get a second wideband controller for the other bank on my car (so far just been using the right bank)

https://github.com/rusefi/rusefi/issues/2044

What are legit reasons to correct one of the banks? i.e. do we expect not same air flow or not same temperature or not same WBO age?

Same question about MAP or even MAF sensors. At what point do we have to run second ECU really to have a second of everything?

^^^ what he said.

Additionally, it's fairly easy to check if cylinders are assigned to the wrong sensor.
1. in closed loop, make a significant but non-life threatening change to the fueling on a cylinder (10-20%)
2. Observe which sensor(s) see the AFR deflection
3. Rinse, repeat #1 and #2 until all cylinders have been accounted for.

If the banks are simply reversed, an unchecked PID closed loop fuel control loop will result in one bank going to maximum lean correction and the other going to maximum rich correction because the changes the loop creates will have the opposite of the intended effect.

If the banks have a mixture of correct and incorrect cylinders assigned, all bets are off.

This reminds me that I think it's incredibly sane to "letter" cylinders (i.e. "A" "B" "C" "D") in terms of their firing order vs. number them like a manufacturer does. This is purely a subjective opinion.

It does indeed need per bank fuel trim - usually the exhaust is different for steering/firewall/starter clearance, which can cause a bit of difference at different load/rpm. All OEMs do this. One independent trim bank per upstream O2 sensor.

mck1117 wrote: ↑

Mon Oct 19, 2020 5:15 pm

It does indeed need per bank fuel trim - usually the exhaust is different for steering/firewall/starter clearance, which can cause a bit of difference at different load/rpm. All OEMs do this. One independent trim bank per upstream O2 sensor.

+1
That is the standard. Assume the sensors are correct and trim whatever cylinders feed the sensor. I think the OEMs also track the fuel and MAF readings to ensure the O2 sensor readings are reasonable but I don't think any aftermarket ECU I've seen does that. This is not a V engine only think, the BMW straight 6s use 2 as well, front and rear blanks.

Just cuz I fitted 18 bungs to my headers figuring it might help. Last run it had 2 WB in the front/rear banks and 4 NB in the groups of 3 and I'm trying to pull cylinder MAP of the multiMAP signal but didn't play with that enough to know if that's working or I need to physically being a signal from each sensor....this is because ITBs just don't stay matched and I'm hoping to use balance in the ecu a bit before lighting the warning light to tell me its time to adjust. This is why ths 12 O2 bungs too....just hopping to get the data on the inlet side where I already have 12 sensors.

blundar wrote: ↑

Mon Oct 19, 2020 5:08 pm

This reminds me that I think it's incredibly sane to "letter" cylinders (i.e. "A" "B" "C" "D") in terms of their firing order vs. number them like a manufacturer does. This is purely a subjective opinion.

Out of curiosity, why? The whole point of the ECU knowing the firing order is that you don't have to think about it. Coil 1, injector 1, and cylinder 1 are all connected together. I know my engine has all the odd cylinders go down one header, and the evens down the other. Don't have to think about firing order.