RX8Club.com

RX8Club.com (https://www.rx8club.com/)
-   Rotary Swaps (https://www.rx8club.com/rotary-swaps-217/)
-   -   Tricking ECU to get rid of oil level warning light. (https://www.rx8club.com/rotary-swaps-217/tricking-ecu-get-rid-oil-level-warning-light-265390/)

stickmantijuana 05-29-2017 11:41 AM

Tricking ECU to get rid of oil level warning light.
 
I have a swapped RX8 without the stock oil level sensor, and I'm having trouble getting rid of the oil level light.

I tried to ground it to chassis and leaving it cut, but the light stays on. In the wiring diagram, it looks like it should be grounded to the chassis to trick the computer to think the sensor is tripped....?

I noticed a few swapped RX8's here that got rid of that light, and was curious how that was achieved. Thanks all.

btw, is there any way for me to get rid of the check engine light?

Reoze 05-29-2017 12:01 PM

The oil level sensor is just a switch iirc. If grounding the sensor output did not work, then the next thing I would try is shorting it to positive (the one on the sensor). One or the other should produce a result. If it does not, then it's possible that this is a sensor with fault modes built into it. Some sensors use 0v or 5v to signify a fault. While valid ranges are something like 0.1v to 4.9v. If neither grounding, nor shorting, end up working. You might want to try a small voltage divider.

stickmantijuana 05-29-2017 12:07 PM

1 Attachment(s)
Thanks for the fast reply.

Not really feeling too good about randomly trying to supply positive voltage to a ecu wire lol...

from the diagram, it looks like it's grounded... and from the pictures, the sensor is 2 wire. I wonder if it has some internal resistance...? do you have one laying around and a multimeter? :fingersx:

Reoze 05-29-2017 12:13 PM

I've always been pretty terrible at reading automotive diagrams but this sensor is just a simple switch. You could think of it like a light switch. If breaking the connections did not produce the result, then you need to keep them closed (short to input power). As long as you're not exceeding the input voltage of the switch, I don't think there's any issue at all.

Now if you tried shorting a 5v sensor to a 12v line (battery), then I'd be a bit more concerned.

If you're still worried, try bench testing it. Hook up 5v to the input, and ground the other side. Open and close the switch and look at the voltage in each state.

dannobre 05-29-2017 04:37 PM

You don't want to back feed the ECU with that circuit. That is a grounded switch...it is either grounded...or not so you only really have 2 choices ;)

Neither of which involve applying ANY power to it :)

stickmantijuana 05-29-2017 06:12 PM

cool, thanks man.

well I had the car on, tried grounding the ECU lead for oil switch grounded to the chassis and just left cut, and the light is still on.. tried turning car on and off, and no difference..

I did a search, and from what I read, the light stays on even when you just unplug the oil switch sensor harness, so I think there's something little more complicated to it... internal resistance when the level is good? not too sure.. :Wconfused

Reoze 05-29-2017 08:26 PM


You don't want to back feed the ECU with that circuit. That is a grounded switch...it is either grounded...or not so you only really have 2 choices
Fair enough, as I said I can't read circuit diagrams for shit. But the switch could've only worked two ways. Applying power wouldn't have had a negative effect at all though.

What I don't understand is how the ECU is reading this value? If the output of the switch runs directly to ground, then the only way I can think of that the ECU can read this value is to measure the current draw on that specific pin. Which is a pretty terrible way to handle boolean logic.



I think there's something little more complicated to it... internal resistance when the level is good? not too sure..
You're probably right, as I said before, a lot of sensors treat absolute min and absolute max as a fault. There's a good chance it's looking for something in between.

Do you have an actual sensor, or just the wiring that's left over? I imagine you could measure this by throwing it in a glass of water or moving the trigger float by hand.

I found an image of the sensor that says it's a Hitachi LS-22A. I can't find anything remotely close to a datasheet on it though.

stickmantijuana 05-29-2017 09:42 PM

1 Attachment(s)
here's the factory service manual.

It looks like it's grounded when the level is fine. It's weird... the light stays on even when I ground the wire to the chassis.

oh well, I guess it isn't a big deal... I just have to live with the light lol

Reoze 05-29-2017 10:14 PM

The service manual states that when the dipstick is at or above the full line, it reads a 1.0 or less. If it were designed to simply read zero, it would have stated such. You likely need to allow some voltage to remain passing through the sensor line.

I'd try a simple voltage divider.You need two resistors of different values. It's about as simple as you can get, the voltage is split proportionately between the two resistors. If you bridged negative and positive with an 11k ohm resistor, wired a 1k resistor from that to ground. Then fed it 12v, you would see 1v on the other end of the voltage divider. With 11v being sent to the grounding point.

It does not have to be these values specifically at all either. As long as you can handle the current being applied to the wires, you could use almost any combination of resistors you have on hand.

A 47k ohm resistor and a 5k ohm resistor would give you 1.1'ish volts for example.

Voltage Divider Calculator

dannobre 05-30-2017 12:13 AM


Originally Posted by Reoze (Post 4821126)
Fair enough, as I said I can't read circuit diagrams for shit. But the switch could've only worked two ways. Applying power wouldn't have had a negative effect at all though.

Applying power to a grounded reference will likely blow up the ECU...or at least any safeguards that are in place to prevent that scenario

If you connect the 2 wires together in the plug it should go out :)

Reoze 05-30-2017 12:59 AM

I'm not sure if you quite understood what I was saying. The switch is either normally open or normally closed. If the switch is acting as a current sink or not doesn't really matter. By connecting the two wires you aren't doing anything that isn't happening during normal operation. You certainly would not under any circumstances fry your ECU. Unless you start shorting random wires together. On top of that, this is really only an issue when you are shorting VCC to Gnd. This is a sensor line, the current passing through is most likely negligible.

dannobre 05-30-2017 01:12 AM


Originally Posted by Reoze (Post 4821147)
I'm not sure if you quite understood what I was saying. The switch is either normally open or normally closed. If the switch is acting as a current sink or not doesn't really matter. By connecting the two wires youaren't doing anything that isn't happening during normal operation. You certainly would not under any circumstances fry your ECU. Unless you start shorting random wires together.

You were the one talking about voltage in the first place 😎

Reoze 05-30-2017 01:26 AM

Well now you lost me. There's still voltage involved?

Not to mention...


If you connect the 2 wires together in the plug it should go out
Is exactly what I was saying to do.

TeamRX8 05-30-2017 09:43 AM

3 Attachment(s)
...

see attachments, bridge a resistor across the two plug contacts to achieve 1 volt or a little less, use 14.7 Volts as B+



.


All times are GMT -5. The time now is 02:57 PM.


© 2024 MH Sub I, LLC dba Internet Brands