I've been doing some research into how to set up a proper cooling fan bypass switch.
What I've found is that in the 2000+ j-bodies, there is an issue with bridging the relay. Apparently, if you groud out the dark green wire to the relay (which makes the relay turn the fan on), the ECU will detect that it's been grounded and throw a code.
What I'm wondering is whether I can use a diode to prevent the ECU from detecting this ground. The preferred method of turning the fan on seems to be to install a second relay which sends power direclty to the fan.
My theory is that if I can prevent the ECU from detecting that I've grounded out the relay, I can use that method without throwing codes.
No I know I can use a diode to prevent voltage from flowing through a wire. I did this when installing my interior trunk switch. I'm just not sure if it works the same way when preventing a ground.
In my head this should work, but I'm no electronics guru.
If I put a diode between the ECU and the relay, will the ECU be able to detect that there is a ground on the other side of that diode?
When the ECU grounds that wire, will the relay properly detect the ground through the diode?
I'm not sure if current flows from the ECU to the relay or from the relay to the ECU, but if you could use a diode to block the current and have it work then it would probably work if you just disconnected the wire.
Do you have a set of prints with the fan circuit? Instead of trying to turn on the relay directly, couldn't you trick the circuit into thinking it needs the fan on. Some cars have a thermal switch mounted on the firewall. Depending on how it's wired you could either wire a switch in parallel to bypass the switch or in series to cut it out (whichever makes the fan run).
Wild Weasel - in theory, that would work. You'd need a big enough diode to handle the current though.
i think this would work, but like lan man said you will have to find a diode that can handl the power.
a diode, will act like a switch.
this is kind of hard to grasp but, it does. there are tons of diodes, but it sounds like you want a switching diode.
a switching diode will prevent voltage from flowing through a circuit so long as it is under 0.7V. as soon as the diode sees 0.7V, it will allow voltage to flow.
since you even admitted to not being an electronics wiz, dont waste your time with this as electronics are a thorn in the nuts (im in the process of becoming an electrical tech).
email me a copy of the schematic you have, and i will see if there is some where to wire in your switch.
now, on the other hand, from what i remember. if you install that diode backwards, it takes a @!#$ load of voltage to make it switch.
Injection is nice but id rather be BLOWN!
The circuit diagram is pretty simple.
The ECU triggers a relay by grounding the coil. So when you splice into that wire and put a switch on it that goes to ground, the relay coil will energize and the fan will come on.
The problem is that the ECU somehow detects that there's a ground there when there shouldn't be and throws a code. So I'm wondering if the diode will isolate that ground from the ECU.
This would all be easy if the ECU was applying 12V to that wire. I know the diode would work for that. I just don't really know the difference beteen applying power and applying ground.
It seems to me that applying ground basically draws 12V from the other end of the wire. So a diode turned around backwards should do the trick.
It's not a high amperage wire since all it does is energize the relay. The big wires are from the relay to the fan.
I'll try to draw up a diagram...
yes, a diode can do that. basically, you're using it to stop current flow in a given circuit. from what you're saying, the ECU senses there is a ground there because of current flow caused by the relay energizing and completing the circuit. since diodes will only flow current in one direction, you can use it to stop the current flow for this particular circuit through the ECU. (depends on how you look at current flow... negative to positive, or positive to negative.)
post up the schematic, and I can give you a better idea of how the circuit works, and how to apply the diode.
Desert Tuners
“When you come across a big kettle of crazy, it’s best not to stir it.”
Ok, here's the stock schematic with my proposed changes in red.
My question is this... can the ECU in any way know that the switch has been flipped, providing the ground for the relay?
not with the diode like that. the cathode side is going to block the ground back to the ECU. if the ECU sends a ground, it will complete the circuit, and energize the relay.
Desert Tuners
“When you come across a big kettle of crazy, it’s best not to stir it.”
btw, the diode isn't going to have to be that big. best bet is to look up the relay and see what the current draw is. I don't think it's going to be that much.
Desert Tuners
“When you come across a big kettle of crazy, it’s best not to stir it.”
That's exactly what I did with my old silver Z24. I can't remember how I had it, but that's the wire I used.
I used a 4005 power diode
if/when this is figured out i want detailed instructions
i auto-x so i want to be able to turn my fan on btwn runs
i found the double relay method, but this sounds like less work once we know how to do it
it's quite easy. use Lenko's diode (the 4005 power one, probably available at Radio Shack), then a DPDT switch. the switch should be able to handle 12 V at around 1 amp. I don't see the coil in that relay pulling a lot of current. hell, the big RF relays I work on here at work only require 500mA actuating current.
wire it just like WW has in the schematic up there, and you should be set.
question is... where is the location of this wire, and what color is it?
Desert Tuners
“When you come across a big kettle of crazy, it’s best not to stir it.”
It's dark green, and it's in the engine bay fuse/relay box. Take the bottom of the plastic out, and you can access all the wires.
i also added extra plugs for fuses for added stuff... .like my Caddy horns, extra fan, and HID's.
Hey Lenko! What's up, bro!
You were involved in the conversations I was sifting through in researching this. In those conversations, people were saying that it didn't work and that the ECU still detected the ground from the switch and threw a code.
I just couldn't figure out how it could with the diode in place.
So you're saying you did this and it didn't throw codes? That would make for some great peace of mind before wiring it up myself!
Yes it works just fine. The people who said it didn't work must have been putting the diode in backwards
Maybe I'll try it on my car today (I have to put another fuse in that panel today, so...) and let you know how it is supposed to go.
Why not just use a 2 position toggle switch with the common hooked to ground? When the switch is in one position it connects the ecu to the circuit and everything acts as normal. In the other position it opens the connection from the ecu to the relay and connects the relay to ground causing the fan to turn on.
Like this
that would still cause the ECU to ground out, not solving the problem of the code trip.
the diode idea would probably work the best.
or, if the motor is a 30A motor, which i am assuming because of the fuse, why not just wire in a switch between the LT BLUE wire and the RED, after the fuse of coarse.
i would think that this would make the fan come on, and have the ECU not trip the code because it has nothing to do with the control section of the circuit. the motor is already grounded on one side, so all it needs is the power to turn it on.
the relay is just one big switch that turns on when the ECU says it should, so why not by pass one switch by putting another switch in parallel with the relay.
a single pole single throw (SPST) should do it. just get a 12V switch rated for however many amps the motor is or higher, and you should be fine.
i drew a schematic in autoCAD, but cant paste that into anything else but Word. if you want the schem, just let me know and i can email it to you.
Injection is nice but id rather be BLOWN!
i would like the pics please
liquidfire007@yahoo.com
whitegoose wrote:that would still cause the ECU to ground out, not solving the problem of the code trip.
No, the it would be impossible to connect the ECU directly to ground. The switch connecte EITHER the ECU to the coil OR the ground to the coil but never the ECU to ground.
Labotomi: That might throw a code when the ECU completes the circuit and no voltage comes through. The reason for the diode is to make the ECU completely unaware that anything has changed. When the ECU grounds the wire, power will flow through it as normal.
For a scenario... if you're in traffic and the temp gauge goes high, you might throw your switch to turn the fan on. Then, as the car gets hot, you might turn on the A/C. The ECU will ground the wire to turn on the fan and will probably throw a code when no current flows. If it can detect another ground, I'd be certain it will detect a lack of current.
I hooked mine up yesterday. So far so good. I can turn it on and off while driving, and turn the A/C on with the switch on, and no codes were thrown. I'll keep playing with it for the rest of the week before calling it done.
Whitegoose: Your solution would work, but the whole point of the relay is that you don't want the high-amperage wires running all over the place if you can avoid it. The original solution to this was to add a second relay that does exactly what you're suggesting. That works. No questions there. The whole point here though is to avoid all of that by just using the existing relay. The ECU detecting the change was the only hurdle and the diode seems to solve that.
I think I'm in agreement with Lenko in that whoever was throwing codes probably had something hooked up wrong. In the original threads on this topic though, nothing was ever concluded so hopefully now we can clear it up once and for all.
Mine's hooked up exactly like my diagram above and working fine so far.
Curiously though, I bought a switch with an LED indicator on it and there doesn't seem to be any way to wire it up so the LED lights up when it's on.
I can light the LED... and I can ground the relay and turn on the fan... but I can't do both at the same time no matter how I wire it up.
hmm, do you have a ground hooked up to your switch in the right spot?
might be the internal wiring of the switch, could be ment to act in a different way.
get a switch from a fog light application, that should work, or wire an LED in series with it since this seems to be a low voltage circuit anyways. the LED shouldnt blow.
four switch should have the three prongs on it for GND ACC and PWR, depends on the internal position of the switch. sounds like if you flip where your PWR and ACC wires are, it might work. keep playing with it.
Injection is nice but id rather be BLOWN!
Yeah, it's got those three connectors but every which way I switch the wires, it won't turn the fan on and the LED on together.
I think it's the internal wiring of the switch and there's probably nothing I can do about it but to wire up an LED separately.
either that or get yourself a new switch.
like i said, one ment for fog lights should work, as my rocker switch illuminates when i turn on my fogs.
or, just say to hell with it! how important is it to know that the fan is on, so long as you dont exceed the duty cycle of the fan motor.
Injection is nice but id rather be BLOWN!
There's a duty cycle for a fan motor?? The fan is on constantly when you run the A/C so if you go for a road trip, it could be on for hours at a time.
It's really not important. The fan is loud enough that I'm not going to forget that it's on and walk away from the car and kill the battery.
watch, now youve cursed yourself and you'll leave er on and drain the damn thing!
lol
Injection is nice but id rather be BLOWN!
