rusefi.com

What do we call those Denso four-wire sensors which are not binary narrow not bosch wideband?
See https://github.com/rusefi/wideband/issues/12

See https://rusefi.com/forum/viewtopic.php?p=43610#p43610

@dron_gus says just a resistor divider and we are in business? What about heater control - is there any strategy or just constant power?

AndreyB wrote: ↑

Fri Nov 25, 2022 7:46 pm

What do we call those Denso four-wire sensors which are not binary narrow not bosch wideband?

planar o2 sensor

AndreyB wrote: ↑

Fri Nov 25, 2022 7:48 pm

@dron_gus says just a resistor divider and we are in business? What about heater control - is there any strategy or just constant power?

Huh? They require current sensing on a constant voltage source. Supply 0.4v, measure current source/sink.

Forgot to attach the schematics

It needs the voltage to be stable, and yet have enough sensitivity to actually measure useful information. The sensor's linearity is only assured when the voltage is stable. 0.36 to 0.50 volts probably doesn't count as stable. You could probably use the sensor if we had the characterization information about its behavior, but that's only really available for stable voltage, not variable voltage.

That's also a 0.14v range, most ECU divide that so it's actually 0.096v at the ADC, which is only 119 ADC counts. Not a lot of significant figures, but maybe enough.

an opamp to amplify voltage?

puff wrote: ↑

Fri Nov 25, 2022 9:23 pm

an opamp to amplify voltage?

yes, I did a design a while ago that uses a jellybean dual opamp and some passives. Never fabbed it, though. Will share it if I still have it when back from my trip.

Cold sensor, heater resistance 3Ohm. Lab power supply reports 3A starting consumption which falls to 2A in 20 seconds. Three minutes later current falls to around 1.7A and stays at 1.7A for next hour. Heater resistance when hot 6A-ish.

Looks like this heater just needs constant power and it does the rest itself.

Between two DMMs in 20mA mode I am failing to see any life from the sensing part of the sensor. Have I fried my sensor or do I have to heat it to get current flowing?

white wire: signal negative
blue wire: signal positive

it needs to be hot to work - extremely high impedance when cold

mck1117 wrote: ↑

Fri Dec 02, 2022 7:51 pm

it needs to be hot to work - extremely high impedance when cold

A bit of heater juice and It's alive! Thank you.

1.71mA on open air and goes down to 0.8mA with a lighter

glad to see you making progress on this.

https://github.com/mck1117/denso-wideband

Measures -0.8 to +0.8mA, constant 0.4v across the sensor within +-25uV across the current range. Outputs linear voltage from 0.4 to 4.6 volts.

About 70hz -3db bandwidth.

Goes to WA tomorrow

I usually see 2.5v output and definitely saw 5v a couple of times. Please advice.

https://github.com/mck1117/denso-wideband/issues/5

Looks like the S+ and S- labels on the PCB are swapped - S+ is really S-, and S- is really S+. Try swapping your sensor wires and see if that helps.

What about https://github.com/mck1117/denso-wideband/issues/6?

mck1117 wrote: ↑

Thu Feb 02, 2023 3:51 am

Try swapping your sensor wires and see if that helps.

Things are much nicer this way!

I now have 5v without a lighter, and with a lighter I can get voltage to about 3.7v

AndreyB wrote: ↑

Thu Feb 02, 2023 3:51 pm

Things are much nicer this way!

I now have 5v without a lighter, and with a lighter I can get voltage to about 3.7v

You can also try pumping gas from the lighter unlit in to the sensor, that should read even richer than the flame.