hamhawk
Member
What info you looking for? Ive been running the ecm controlled fan on mine since feb 2024 with no issues.
Using PCM to control aftermarket electric radiator fan?
- Thread starter Steel City 06
- Start date
What info you looking for? Ive been running the ecm controlled fan on mine since feb 2024 with no issues.
is it a TJ ECU? I thought you were running on an XJ ECU?
Is this PWM variable speed, or just a lo/hi 2speed setup (or maybe just simple on/off)
I see "Pin C2" and lo/hi set to 199* in HP tuners with the 56041465AC suite.
my goal is to hook the Camaro Fan directly to the PCM for PWM control and then set A/C override, cutoff MPH, on/off temp - all that.
I appreciate the clarification - got a little confused reading through the thread.
hamhawk
Member
Its a tj pcm with a xj tune file. Its a simple on/off. I played with all the settings for a hi/low setup, but it didnt seem like anything but the on/off worked. I havent played with any pwm stuff, not sure if the jtec pcm supports it.
are you sure your running a PWM capable fan? some of what you were describing with cycling and odd behavior with any settings other than 199 on/off might be a non PWM fan trying to do PWM (I'm looking for a small hope it might work with pwm lol)
hamhawk
Member
Its just a standard fan. I seem to remember reading the pulsing of the output was for "soft start" on the fan, i want to say it was a grand cherokee feature but cant remember. Set up the way it is works for me, so i havent dug any deeper into it.
I gave up and went aftermarket.
During my testing I was unable to determine if the PCM actually had the hardware for PWM control, and if so, what pin it would be on.
Unwilling to fry a new PCM testing it, so I went aftermarket instead.
There very clearly is software in the PCM to control it, and with the proper tune, you can track “desired fan state” as a percentage, and it ramps in and out continuously as expected rather than just on-off (or 2-stage). However, it is unclear if there is any hardware in the PCM to actually provide this signal (and if so, what pin).
Theoretically, one could network in to the PCM through the OBD-II communication protocols and use an Arduino to create a signal, however this is as complex and possibly less reliable than just using an Arduino to control it independently.
So I did find my '99 XJ auto JTEC+ PCM and used some credits to be able to write to it. First off, I threw codes 1698 for no TCM communication which was expected because I row my own gears. The other was for EVAP pump something blah blah. I disabled the evap pump code with HP Tuners. The TCM error code won't go away despite the fact that I have disabled every single trans diagnostic code available in HP tuners. Frustrating. Also for reference the oil pressure gauge is dead with that PCM although I've been told how to fix that.
Anywho, got the fan settings enabled and pin C21 from the PCM now shows switched ground when the ignition is on, but no setting I try does anything other than give me an ignition switched ground. I'm going to reread this thread and see if I can find anything to try. But at least I tried trying today!
Anywho, got the fan settings enabled and pin C21 from the PCM now shows switched ground when the ignition is on, but no setting I try does anything other than give me an ignition switched ground. I'm going to reread this thread and see if I can find anything to try. But at least I tried trying today!
So I did find my '99 XJ auto JTEC+ PCM and used some credits to be able to write to it. First off, I threw codes 1698 for no TCM communication which was expected because I row my own gears. The other was for EVAP pump something blah blah. I disabled the evap pump code with HP Tuners. The TCM error code won't go away despite the fact that I have disabled every single trans diagnostic code available in HP tuners. Frustrating. Also for reference the oil pressure gauge is dead with that PCM although I've been told how to fix that.
Anywho, got the fan settings enabled and pin C21 from the PCM now shows switched ground when the ignition is on, but no setting I try does anything other than give me an ignition switched ground. I'm going to reread this thread and see if I can find anything to try. But at least I tried trying today!
If it does work, it is going to be switched ground at approximately 100 Hz. There is no "pull-up" inside an ECM normally, so it can be very difficult to discern if it is switching or not. The fan's internal controller usually has a pull-up circuit to 5V (sometimes 12V) with an inline resistor somewhere in the range of 2,200 ohms for 5V.
Without a pull-up circuit, the pin will probably just appear to have continuity to ground regardless of PWM state.
Interesting. Not familiar with the "pull up" term, but you are saying it will look like it's seeing switched ground just from ignition power because it's pulsing too fast for my mark 1 eyeball? But in fact it's trying to send a varying PWM signal? Because I'm just trying a simple voltage probe light thingy.
And if anybody has any input on the P1698 TCM code I would greatly appreciate it
And if anybody has any input on the P1698 TCM code I would greatly appreciate it
Possibly, yes. You'll need an oscilloscope to differentiate between a switched ground and a 100hz pwm waveform.
Interesting considering hamhawk seemed to just get a switched on/off ground. Or maybe he didn't but it functioned as one with the PWM pulsing? Hamhawk?
Also, my 99 XJ PCM has a 493AD OS FWIW.
Also also just for reference, my stock '98 TJ PCM shows no ground at all on C2 or C21. I was just out there swapping PCMs back and forth and noted the difference.
Also, my 99 XJ PCM has a 493AD OS FWIW.
Also also just for reference, my stock '98 TJ PCM shows no ground at all on C2 or C21. I was just out there swapping PCMs back and forth and noted the difference.
wellp that is just promising enough for me to start budgeting for parts - guess I have to join the party.
There are 3 different generations of PCM in this conversation making it more confusing. 05 - 06 running the NGC3 PCM have huge differences in the config options compared to the JTEC/JTEC+ found in 97-03 TJs.
Anyways, sounds like you a making some headway on your JTEC+ config. I wish I had a good case for a picoscope for just this sort of project. They're a bit too expensive for a home mechanic. https://www.picoauto.com/products/automotive-oscilloscope-kit/overview?model=4425A+Starter
I'm still on the fence about the electric fan. The math all says do it. The Engine Masters episode on mechanical fan power consumption was shocking. On the other hand the clutch is dead simple.
I tried to do a highway run in hot temps with my GoPro on a magnetic mount under the hood to figure out if the fan clutch was engaged or not, but the rotation speed in the dark is high enough that you can't tell.
If somebody has relevant info on ngc I'd love to have some.
Same. I'm confident I can do something cool with an Arduino and have the software close enough to test, but I'm concerned about the reliability and how to make the electronics as robust as a mechanical fan clutch. I don't want to have a failure 20 miles into the woods.
So, a few things after some tinkering:
1- EVIDENTLY if you don't get the additional PCM wire/plug inserted all the way and it isn't touching the pin on the PCM at all, no matter what settings you have, the PCM cannot provide that wire a ground. I don't remember the dumbass who neglected to tell me this, but thanks asshole
B) At least for a '99 XJ PCM, it's C2, not C21 (and C21 is always an ignition on switched ground from what I can tell)
III. I now get a dimly glowing led test light before the trigger temp, them a full boogy ground exactly at the trigger temp. So progress, but confused on why I see any ground before hand. And it's an instant thing at the trigger temp, not a rapid ramp up. Maybe going to test it with a relay tomorrow.
Also, the 99 XJ factory service manual says it's just a switched ground, not PWM. I suspect that's a later years thing if at all.
1- EVIDENTLY if you don't get the additional PCM wire/plug inserted all the way and it isn't touching the pin on the PCM at all, no matter what settings you have, the PCM cannot provide that wire a ground. I don't remember the dumbass who neglected to tell me this, but thanks asshole
B) At least for a '99 XJ PCM, it's C2, not C21 (and C21 is always an ignition on switched ground from what I can tell)
III. I now get a dimly glowing led test light before the trigger temp, them a full boogy ground exactly at the trigger temp. So progress, but confused on why I see any ground before hand. And it's an instant thing at the trigger temp, not a rapid ramp up. Maybe going to test it with a relay tomorrow.
Also, the 99 XJ factory service manual says it's just a switched ground, not PWM. I suspect that's a later years thing if at all.
Interesting. Not familiar with the "pull up" term, but you are saying it will look like it's seeing switched ground just from ignition power because it's pulsing too fast for my mark 1 eyeball? But in fact it's trying to send a varying PWM signal? Because I'm just trying a simple voltage probe light thingy.
And if anybody has any input on the P1698 TCM code I would greatly appreciate it
So the signal is literally just the ground switch opening and closing 100 times per second. It contains the information on the percent state target based on the length of the closed state as a function of the length of the pulse. So for example, 50% closed and 50% open means 50% fan state.
The reason you can’t see this with a multimeter is twofold. One, the multimeter has capacitance and has no idea the ground is switching on or off at that speed. Two, your eyes can’t see anything near this fast and make sense of it.
So to get a voltage signal out of it, you need a supply voltage (usually 5V) with an inline resistor (2,200 ohms is common). This supply has a small amount of current flowing when the switch is closed and not when it is open. Then you can use a simple oscilloscope to monitor the voltage at the PCM end of the resistor. When the switch closes, this voltage snaps to zero, and then snaps right back to 5V when the switch closes. (The resistor is just there to keep the current at sane levels and not fry the power supply or ECM.)
This signal will show up as a square wave waveform on an oscilloscope. The duration of the grounded state to the total wave period is the PWM %.
For reference, even if you had an insane voltmeter, you wouldn’t be able to see it. Most TV programming flashes at only 30 cycles per second, and computer monitors usually about 60. So 100 cycles per second would just be a blur.
Imagine you’re trying to figure out what a switch on your car does. You think it might go to a cigarette lighter outlet. If you just stick a voltmeter in the outlet and wildly flap the switch up and down at superhuman speed, it will still show up as 12V regardless of whether or not the outlet is actually being switched because of the capacitance of the meter and even the wiring itself. So you need something that can quickly purge that residual voltage (such as a resistor to ground, in this case a pull-down resistor) and also a voltmeter that can keep up with your superhuman switch flipping capabilities.
Last edited:
I'm tracking you are saying it's a PWM signal I'm seeing. If that's the case though, why does it flip instantly to a hard ground at trigger temp? Also the FSM doesn't mention PWM. Would it be so much slower flashing at room temp ECT that maybe I could see it?
I did just connect a relay with the dim ground and it's enough to trigger the relay FWIW.
I did just connect a relay with the dim ground and it's enough to trigger the relay FWIW.
PWM is a setting option in the programming. I believe certain Jeep models of the era (WJ if I recall correctly) used a PWM electric fan, though they used a rather oddball solid state relay and just amplified that PWM as the fan motor's actual power supply.
Modern PWM fans use the same signaling protocol that was originally designed to run a SSR without requiring any solid state relay. So something like the Camaro fan can theoretically read the PWM signal directly from the PCM.
Sone of the very late NGC4 PCMs do also use the PWM signal to control electric fans.
The idea is that if some models of the JTEC PCM and some models of the NGC4 have the hardware and the software to run a PWM fan, and the NGC3 has the software, there is a good chance that all three PCMs are just blanket equipped with that capability.
Note you'll be seeing a relay on-off signal if that's what you have selected in the tune. You'll only see PWM if you switch it to the PWM fan setting.
I did set the "DC fan" setting to disabled per hamhawk's experience. I'm going to enable it and see what happens.
No change, but I also checked with the ignition off and I still get the same weak ground?
Similar threads
