Simulated Narrowband/ 39f/44f lambda signal voltage tolerance

[vanepico]

Hi guys,

I am planning to wire a 14point7 SLC OEM simulated narrowband into my non lambda CA18DET (which to what I have read so far is possible, as I don't need the heater transistor).

My problem comes from the fact that when the module goes into a calibration mode, it pulls the simulated narrowband output up to 5v (which seems like a weird design decision to me, considering the ECUs should only see 1v max). Has anyone looked into how much voltage the ECU pin 29 lambda input can tolerate? The signal pin appears to go to the heatsink of what looks like a power transistor marked 27 92 and a pin of a module marked HRA-15 but I couldn't trace any further.

I was going to try a resistor and zener diode on the input to limit to 1v but you can't get a 1v zener, the smallest I've found is 2.7v. Will this voltage be tolerated?

I am also thinking about how to deal with when the wideband isn't up to temp, or is having errors, as when the ECU is set to ignore sensor temperature (on account of not having access to narrowband heating element for measurement), it would blindly follow the lambda voltage for fuel trims, possibly leading to problems, so it would be nice to know what voltage on the signal input would cause the ECU to stop trimming (I'm assuming 0v, if it thinks it is shorted to ground, or unplugged), so I can build in a failsafe system.

Thanks

TL;DR

• Can non lambda 44f ECU still read sensor voltage and apply fuel trims?

• Will circuitry on pin 29 lambda signal die from 5v? What about 2.7v?

• What signal voltage/continuity on heater circuit will cause the ECU to stop trimming?(for failsafe purposes)

[Matt]

1. Don't know... haven't looked that closely at that variant. Would need to run on the bench and monitor it to see what it does

2. I've seen people accidently put 12V into this input and not kill it. The ECU uses a 0 - 2.55V AD converter to display 0 - 2.5V (for the 0-1V input). Over this voltage doesn't seem to damage the input, but is not recommended. Haven't looked into the circuitry from the input to the converter. It may go through one of the ceramic opamp boards

3. ECU doesn't look at the heater circuit, only the temp sensor input to see at minimum temperature before using the O2 sensor input

[vanepico]

Thanks for the reply, hopefully I can get the ecu software side of the 44f to work like the 39f lambda version, if not I might just put in a 39f ecu.

Seeing that people have put 12v in inputs by accident and not fried anything is good to hear, 2.7 for very short periods should be bearable, I suppose being in an automotive application the ECUs were probably designed to be pretty resilient to worn through wiring looms etc. I presume the HRA modules are the ceramic op amp modules you mention.

Ahh, the widebands control sensor temp in a closed loop way, I saw temperature and thought it was something like that. I thought it might avoid trimming if the sensor wasn't up to temp.

Does the ADC get pulled to 2.5 or 0v when the sensor is disconnected on lambda ECUs? If you've got the part number for the ADCs commonly used I wouldn't mind searching for it in my spare ecu. I did look up the processor datasheet and it didn't seem to have any built in ADCs so I assume it is an external chip?

Thanks!

[Matt]

AD converter is HD46510

Normally signals to the AD converter (0-5.12V inputs) would have an internal weak pull down. Further in the circuitry there may be other filters and pull up/downs on ceramic coated boards which are not documented