Adding a Switch to a Jeep OE Aux Fan Circuit

I need some help with the wiring in my Jeep, and I think I've got a solution, just want some feedback before I do anything...

The auxiliary fan is controlled by the engine computer (ECM), and I want to be able to activate the OE fan relay without the ECM knowing (that would give me a Check Engine Light - CEL). The relay's coil has constant +12v, so it just needs a ground to activate, but the ECM needs to see either +12v (when it thinks the fan is off) or a couple mA (whenever it thinks the fan is on). I need help fooling my ECM.



Here is how the system is currently wired, the failed option I'm currently using (that will turn on the fan, but I get a CEL), and what I think might work (the cap is there so that the ECM sees the +12v before the current all flows to ground, I'm afraid that split second while switching might give me a CEL).

Please give me any advice/opinions you have, and thanks for any help you can provide.

Have you thought about controlling the fan instead of the relay? The ECM shouldn't be able to "see" that.

You can try replacing the SPST switch with DPDT switch.

This will enable you to "fake" the relay when you switch to manual fan ON. The diode is there to prevent the ECM from seeing a short disconnection. You can use diode like 1N4004 to 1N4007 .

Try first using 1KΩ for Rx and if that doesn't work, reducing its value to half and try again. If that still doesn't work, try measuring the real coil resistance of the relay and use that instead.

Ron H said:

Have you thought about controlling the fan instead of the relay? The ECM shouldn't be able to "see" that.

Click to expand...

Yes, it would certainly work, but I don't want to add more relays/wire to an already cluttered engine bay. I'd prefer to run a couple wires and call it a day.

eblc1388 said:

You can try replacing the SPST switch with DPDT switch.

This will enable you to "fake" the relay when you switch to manual fan ON. The diode is there to prevent the ECM from seeing a short disconnection. You can use diode like 1N4004 to 1N4007 .

Try first using 1KΩ for Rx and if that doesn't work, reducing its value to half and try again. If that still doesn't work, try measuring the real coil resistance of the relay and use that instead.

Click to expand...

I like this a lot, but my only concern is that during switching, the ECM would 'see' a lack of voltage. Would adding a capacitor where I indicated decrease this possibility? If so, what size should I use? Thanks again for your help.

no one?

10 char

lost honda pro said:

no one?

10 char

Click to expand...

The diode takes care of the momentary dead time when both switches are open. The only potential problem I see is if the contact to ground occurs before the other pole contacts the resistor. It might be a non-issue. A cap to ground where you indicated might help that situation, but you would have to try it. You would have to experiment with values. I would start with 100nF. If that doesn't work, go up in decade steps to maybe 100uF max.
Below is an alternate scheme that never presents an open to the ECU, but will temporarily present a double load during switching. If eblc1388's scheme doesn't work, you could try this one. If you don't return the dummy relay to switched +12V, it will always present a small load on the battery. That may present a wiring problem.

Ron H said:

The diode takes care of the momentary dead time when both switches are open. The only potential problem I see is if the contact to ground occurs before the other pole contacts the resistor. It might be a non-issue. A cap to ground where you indicated might help that situation, but you would have to try it. You would have to experiment with values. I would start with 100nF. If that doesn't work, go up in decade steps to maybe 100uF max.
Below is an alternate scheme that never presents an open to the ECU, but will temporarily present a double load during switching. If eblc1388's scheme doesn't work, you could try this one. If you don't return the dummy relay to switched +12V, it will always present a small load on the battery. That may present a wiring problem.

Click to expand...

Thanks Ron, that's another great option. As for your concerns, any of the powered wires would be ignition hot only, so a small additional load while the engine's running won't be a problem.

Thanks again for the help guys!

Results:

eblc388's method works flawlessly. Ron H was entirely correct that the diode takes care of the switching time. I have no error lights, and it should be dead reliable.

You guys are awesome, thanks again for the help!