Ethanol Content in Australian 98 Octane

Hi All

I'm progressively working on by e85 Turbo build and got the ethanol content sensor working on a M150. The Fuel Temp reads fine but I get about 30% Ethanol content reading on 98 pump fuel (Melbourne, Australia).

Could 30% ethanol be right? I'm trying to figure out if I've either misconfigured the sensor/sensor error or if the fuel you buy at the pump could actually have 30% ethanol in it.

Thanks

i run BP 98 and its under 1% ethanol, and i believe maximum they can add to unleaded is 10%.

water can give a reading as though its ethanol, has the tank had water or ethanol in it?

the gm sensors fault frequency tends to be 180hz i believe so it reads 100% ethanol, my old dead sensor did anyway (temp still worked).

you getting 30% is strange

Does the reading go below 30%?

Which Sensor and calibration are you using with the M1?

I have a Continental 13577379 and used the calibrations from the M1 dropdown selection for fuel composition (reads 30% which doesn't seem right). I don't have a way to make it move by changing fuels - it hovers between 30.0 and 30.1 %

For fuel temp I entered a manual calibration from 1V (-40C) to 5V (125C) and got a sensible reading of about 20C which seemed plausible.

Hi Adrian,

It sounds like you have a setup issue with that sensor, I have attached screen grabs of how the sensor should be setup for you to use.

Other than the filter and delay settings that's what I have (using Dig 1 as the input). I'll update those filter and delay settings and report back.

You could swap the input pin to be one of the Udig inputs and then do a capture of the pin to see what data is coming across from the sensor.

Thanks Stephen, I'll try that.

There aren't any UDIG inputs left (on the M150 FT86 package) but I'll temporarily re-assign one from elsewhere and grab some data to report back.

The full part number is: 13577379 1209401768

I used a scope and put a 1K pullup to 5V and the sensor output (which is a switch to ground) to see what frequency it was running on.

It read 78.5 Hz (which when I put a manual calibration of 50Hz = 0% Ethanol and 150Hz = 100% Ethanol into M1tune) both give a 28.5% Ethanol figure. So this means:

- I have a bad sensor

- I actually have a tank of 30% Ethanol! Car has only ever had 98 in it and no water that I know of (I've owned the car since new and don't have any reason to think there is water in the tank), or

- This model has a calibration that is different from 50Hz=0% and 150Hz=100%. As mentioned the Fuel Temp Pulse width component works.

Anyone in Melbourne have a good known working sensor lying about I can borrow?

Hi Adrian,

Why are you using the pullup resistor? the signal out of the sensor is a straight digital square wave signal. You need a 5V+ and 0V to power the sensor, and then the output goes directly to the M1 Digital input pin. I have just scoped my desk one, and sitting in free air I have it reading 50hz, if I add some water into the tube I can get a higher reading. This matches the values being read in the M1.

Mine has the same part number as yours (the first 8 digits) as far as I am aware, the second lot of numbers are the manufacture data and batch number.

Hi Stephen

When wired into the ECU I have +12V and Gnd and the output going to the Dig In. No pullup resistors. The Motec doco says +12V supply, not 5V, so that is what I used. My understanding is that the ECU Dig Inputs have a pullup resistor and hence measure when the input switches to ground.

For testing purposes (stand alone, not wired into the ECU) I needed the pullup resistor 5V to emulate this. I have not used a pullup resistor when the sensor is wired into the ECU.

Does the Ethanol sensor 'generate' a 5V square wave or does it 'switch input to ground' at the specified frequency?

I tested in free-air and it reads 24.5% Ethanol using the default calibration, down from 30% with 98 Octane in it. This sensor is just not sending 50Hz=0% -> 150Hz=100%. It's either a bad sensor or there is another calibration needed. I'm thinking bad sensor.

In free air you should get a 50hz output, anything else is looking like a fault.

I am aware of the M1 Flex Fuel document, as I wrote it... I have been testing these sensors again lately, and I have found that you can run them off of a 5V ECU output if needed, and they will provide accurate readings. The only downside with doing this is that they have a higher current draw and can overload the 5V supply if there are a number of sensors on the bus.

I purchased another sensor from a different supplier. It reads 30% Ethanol on 98 octane too. I put some e85 in and it reads 105% Ethanol. I'm thinking there is either another calibration for these or a batch of bad sensors. I'm just going to pick a sensor, manually calibrate it with a few batches of fuel (both 98 and e85) and hope it is linear. I guess as long as the tune matches the sensor I should be OK.

Hi- i just had a quick look on line and you can do a very simple test at home for measuring the alcohol content of fuel . It was a vid by Delphi sensors . It only needs a calibrated measuring test tube . Good luck .

I ended up buying another sensor from a third supplier. Reads 0.6% ethanol on 98 Octane and 86.5% on e85 - so basically perfectly calibrated.

Appears that there are a batch of dodgy sensors out there.