This is on my '89 IROC in the signature...

The subject pretty-much says all. Basically, everything else on the car is working as it should (electronically) except the fans. I've got the same wiring diagram that everybody seems to float around here [attached]. I understand it enough to know that the fan circuits both work (i.e. everything between the battery, relays, and fans). I know that because, if I short terminals 'E' and 'A' on a relay socket (see diagram), either of one of the fans will come on (depending on which socket) even if the key is off.

I have also tried invoking the diagnostic mode on the ECM which, to my knowledge, should turn the primary fan on; the fan does not come on.

I have a new ECM, new fan motors, and new fan relays.

What concretely can I do to try to fix this issue? I know that I should probably "check the wiring" between the relays and the ECM, however I only know whats going up until the wires run into a black plastic wiring sheath or some electrical tape. After that, I have no way of what's going on...

I appreciate any and all assistance!


BTW -- I know that the diagram has manual switches in it, but I would like to try to get mine working as they would have been from the factory.

 

That circuit coming off the batt positive pole to the relays is a fusible link on my '88 GTA harness. There's a 1-pin Weatherpak connector right near the batt pole which connects the fusible link, if the link is blown it'll be pretty obvious.

Whip out your voltmeter or test light and check for 12VDC at the relay side of that fusible link'd circuit.

When mine blew, I replaced it with two separate fusible links (one for each fan) fed directly from the starter pole where the other fusible links are.

 

Sancho,

If you can get the fans to operate by jumping the 'A' and 'E' terminals at the relays, you probably don't have problems with the fusible links. You are effectively bypassing the relay contacts in doing that.

You can test the operation of the relays and circuits by grounding the 'F' terminal of each relays to turn on the fans. If that works, the relays, sockets, and their wiring are correct. If not, you may have problems with the relays or their connections, or the ignition power supply to the relay controls (C/H FAN fuse and circuit). The C/H FAN circuit is hot in RUN only, so the ignition will have to be ON to test with that method.

 

Run power directly to your fans first and make sure the motors arent dead. Unlikely, but you never know.

 

Well, I at least figured out why the primary fan wasn't comming on in the diagnostic mode: the "new" relay that I got from O'Reilly was bad right out of the box. Seriously, I'm really getting hesitant to put anything from AutoZone or O'Reilly on my car anymore.

Anyway, I changed the relay to another relay that I had laying around, and now the fan comes on when I invoke the diagnostic mode. So, what does that tell us? I already know that the wiring between the fans and the relays is good--can I also assume now that the wiring between the ECM and the primary relay is good as well? I haven't driven the car yet, so I don't know if the primary will still come on when it should, but I at least know that the wiring ought to be good.

As for the secondary... The secondary relay is good. I'm curious: the secondary is supposed to automatically be on if the A/C is on, correct? I'm wondering if I could try to turn the A/C on when my car is running and test for voltage across the terminals on the socket that go to the relay coil.

Also, I understand that the secondary fan is controlled by a sensor in one of the heads, correct? Could I try shorting the wire from that sensor to the block and seeing if that turns the fan on?

 

Okay, another update.

I just went out and took my car for a spin. Of course, since it was moving the whole time, the temperature was very low the whole time. But, when I got back home, I let the car idle in the driveway for a little bit. Over the course of a few minutes, the temmperature gradually pulled past 220, then past the next line (240?), and then it started to drop back down. I turned the car off, then put the key in RUN, and I could tell that the secondary fan was on. So, it looks like the secondary fan is working as it's supposed to.

The primary, however, still will not turn on while the car is running--regardless of the temperature.

So, let's do some "process of elimination" with the primary fan:

* The wiring between the fan and the relay tests good
* The wiring between the ECM and the relay tests good (b/c the fan turns on when I invoke the diagnostic mode as it's supposed to)
* The socket, then, is also good
* The motor, of course, is good as well
* The ECM is new (well, it's been remanufactured, but I bought it only a week ago)

What does that leave? I'm thinking that it might be the temperature sensor, but then again, I can see the temperature being properly reported on the gague on the dash.

Anyone agree/disagree with my logic? Any ideas as to what the problem might actually be?

Thanks again!

 

The temp gauge on the dash is fed from a separate sensor. Dash gauge is the sensor in the driver side head between 1-3 plugs, whereas the ECM is fed coolant temp data from the CTS located @ the front of intake...I'm assuming you have a TPI engine.

If you can get access to a scan tool or buy an ALDL cable for your laptop (you'll want to get ALDL data later when you start software tuning...resistance is futile) you can see what the ECM is seeing for coolant temp.

First, I'd get the engine up to temp and check primary fan operation w/ the engine running, rather than stopped w/ the key in RUN position. On some TPI harnesses the relay pull voltage is only provided when the ECM is seeing dist ref pulses...even if you close ground on the relay, it still won't trigger.

 

kevinc: I forgot all about that sensor in the front of the intake manifold! (yes, I do have TPI)

I have tested the primary fan operation with the engine running, and it simply never comes on. I have been wondering why the temperature likes to climb just up to the redline, but it never crosses into it--nor does the radiator ever boil over. The reason is that the secondary fan is operating as it should and keeping the temperature below redline. The problem is, if the primary were working as it was supposed to, the secondary should never need to kick in (because it shouldn't get that high in the first place).

Regarding getting a scanner put on the car: I just talked with my mechanic about what I want to do (just have him take five minutes to put the scanner on), and he's just quoting me his standard price for a diagnostic (and I won't pay it just for getting one reading off of the scanner). He also said that he didn't think I needed to put a scanner on it anyway, because he claims that if the CTS was bad, the computer would be throwing a code. Well, I just checked the codes on my computer, and it's not throwing any.

I still think it could be the CTS, though... I mean, what if the CTS had a problem where its resistance wouldn't change--how would the computer know that the engine was actually operating at a higher temperature if the CTS is broken and only reporting one temperature? I don't know if that ever happens, but I don't think that just because the computer isn't throwing any codes doesn't mean that a sensor might not be bad.

I think this just gives me a good excuse to want to go make my own data cable so that I can get ECM data by myself... Anyone know of a link to instructions for building one that will work on a TPI car?

Thanks

 

Sancho,

Since the primary cooling fan operates in the diagnostic mode, the relay, wiring, and ECM are probably all O.K.

The CT sensor can be "bad" but still not set a DTC. If the sensor is not dropping the 5V reference signal enough to operate the fan, it will certainly not set a high temperature (low resistance) fault for the ECM. It MAY set a low temperature fault (high resistance) DTC, but only it the circuit resistance is above 120K ohms. You can test the sensor with an ohmmeter. Check the resistance across the sensor (fun to get to, isn't it?) at a presumed temperature and compare it to the chart:



You should also test for the 5VDC reference signal across the sensor connector in the wiring harness with the ignition on. Anything below 4.7VDC is going to indicate a connection problem. Inspect the connector for corrosion, terminal contamination, damage, etc.

If my data is correct, the primary fan should turn on at 233°F indicated coolant temperature and off at 218°F. The secondary fan should not come on first, and since you've proven a good fan circuit and hardware, the CTS is about all that's left. Of course, if you're using an aftermarket or custom EPROM in your ECM, that is a possile culptit as well, but not as likely. Most aftermarkets actually lower the turn-on points, but if there was a write error to the PROM when it was "burned", you could have an issue.

 

replace the cts. it's very cheap and easy to do. being the original, i wouldn't trust it.

 

Well...

Vader and Mystikkal_69: I ended up taking both of your suggestions. After I did some testing, I decided to just get a new CTS. Everything *seemed* to be in order: I got 4.98V across the harness that went to the CTS, and the resistance at room temperature was 2.311 kOhms. However, since the sensor was probably the original--and since Mystikkal made it sound so easy--I broke down and bought a new one. Just for grins, I tested both the new and OLD CTS side-by-side and measured their resistances at varying temperatures, and they both responded generally the same.

So, I put the new CTS on anyway, drive the car around for about 15 minutes. Of course, the car was moving that whole time, so the temperature was excellent (below the first line). So, I decided to let my car idle for a few minutes in the driveway to elminate normal airflow and hopefully force at least the primary fan on. Slowly, the temperature climbed--it got past 220, and the primary still wouldn't come on. I let it climb a little higher, but I then turned it off at around 230 because I know that the primary should have been on before that.


But here's where it gets weird...

I went ahead and manually jumpered both fans to stay on and let the car idle for a little while more (just to bring the engine temp down). While I was waiting, I had the multimeter out and decided to directly test the voltage across the coil leads on the socket for the primary fan (at this point, the temperature was still above 220). Guess what--I read 14V across the coil leads!

I thought: hmm, that should mean that the relay would be activated and thus the primary fan should be operating now. So I went ahead and took the jumpers off of the socket and plugged the relay back up. I'm expecting to hear/feel a click from the relay, but I get nothing. The fan didn't even come on. (Remember, I just read 14V across the coil leads and the relay didn't even click.)

So I thought mabye the socket is just dirty or something. But, before I went to clean it, I went ahead and invoked the diagnostic mode on the ECM by shorting the appropiate pins on the ALDL. The fan shouldn't be comming on if the socket is truly dirty, right? Well, I turned the key, and sure enough the primary is on. I pulled the relay, tested the voltage across the coil contacts, and read about 12.8V.


So, this is *really* confusing me. The connectors are <b>not</b> not corroded if the fan clearly is powering on as it should be in diagnostic mode. My only guess at this point is that, when I read 14V across the terminals but the relay still didn't activate, and then I read 12.8V across the same leads in diagnostic mode and the relay did activate... that must mean that, for some reason, there wasn't enough current on the 14V test to activate the coil on the relay. Is that possible, and what might explain it?

This only makes less sense to me now, so hopefully one of you electronics gurus can make sense of it!

 

Sancho,

Reading a voltage at the relay socket coil terminals with the relay unplugged is not unexpected. The circuit uses a transistor driver to operate the load relays, and unless there is a resistance in the circuit (a.k.a., the relay coil) the voltage will float up to the system voltage. You can verify this by checking the voltage again, except this time install about a 1K ohm resistor to simulate a load. The circuit voltage should drop to near 0. (Don't confuse this with voltage to ground, since the circuit always has 12VDC and only gets a ground from the ECM.) This is completely normal on sinking type ouputs such as the ECM uses.

Since you can force the primary fan to operate by grounding the dark green/white wire (as the ECM driver transistor would), you can safely assume that the fan circuit is intact - at least up to the point of the ECM wiring.

What is still happening is that the ECM is not operating the relay. The reason is beyond me, unless you have an aftermarket EPROM or faulty ECM circuit connections.

The EPROM is your issue. If you have an aftermarket, and still have the original, try installing the original to test the fans.

The wiring I may be able to help with. It would seem that the ECM fan relay circuit is suspect. However, what negates this is the fact that in the Field Service Mode, the primary fan will operate via the ECM, as it should. This indicates that the driver and relay circuit are intact. That only leaves the CTS wiring to the ECM, and the ECM power grounding. Again, if all other systems are working as they should, the grounding is less of a likely problem.

The final answer to this may be a data scanner, so you can read the actual temperature that the ECM is getting from the CTS. If the dash gauge reads 235°F, but the data scanner shows a CTS of only 200°F, you may have problems/resistance in the CTS circuit to the ECM.

One thing you can do to try to "fool" the ECM into thinking the coolant is hotter than it really is would be to install a resistor in the CTS connector. I know it's a PITA to access the connector, but a 27 ohm, ¼W resistor plugged in should be enough to operate the fan without setting an error code. If that does the trick, you likely have some resistance in the ECM wiring harness. Finding it could be a real ******. There are several hidden splices in the factory harnesses, especially in the grounds.

One other thing you can do is measure the resistance of the grounding conductor at the CTS socket to a good engine ground or battery negative terminal. You should have 0 ohms. Any resistance there will only skew the CTS indicated reading lower, and report a lower temperature to the ECM. With only 20-30 ohms, the CTS will never indicate a temperature high enough to turn on the primary fan before the secondary fan operates at 240°F.

Incidentally, the primary fan may not be programmed to turn on until the CTS reports 233°F and back off at 218°F or so. There seems to be a little variation in programming over the years, with different engines and emissions requirements, and such. Given the possible calibration discrepancies of the dash gauge (dash gauge could be a little off - most are), and variation of the CTS, you may have been right on the edge of the fan operating and just "chickened out" a little too soon. Not saying you did, but operating at idle at 230° or even 240° isn't the worst thing, at least for a short time to test the theory.

 

A rough idea of the relay driver circuit, and why you can read voltage with no relay coil load connected:

 

Vader: Thanks a lot for the detailed reply!

Regarding the EPROM: I have no way of knowing if it is the original or not as the previous owner didn't tell me. However, with some of the hack jobs that have been done on this car electrically, it wouldn't suprise me if it is aftermarket or custom burned.

The only wiring that could *possibly* still be suspect is the wiring between the CTS and the ECM. While I have checked for proper voltage across the CTS leads (4.98V), I suppose that that there is a possibility there might be some corrosion or something (how, I don't know).

I definately want to interface with the ECM in some manner beyond simply scanning for codes (i.e. pull data off of it in real-time). This is giving me the perfect excuse to make an ALDL cable and read from the ECM with a laptop. I already have all of the parts, so as long as I don't fry one of the transistors or something, I should be in business. That way, I can actually see what the compter is "seeing" and, hopefully, also be able to view the stored constants in the EPROM (i.e. primary fan turn-on temp). I think that's what will really tell me what's up...

However, I am still curious about one thing... You brought up the issue of the grounding of the CTS. I assume that you are talking about the grounding of the CTS casing to the mainfold and, thus, the block. I didn't think that the CTS needed to be grounded to operate (I thought the two leads comming out of the socket were all that were needed). I can tell you that I put Teflon tape around the threads on the CTS to avoid leaking. Would that potentially cause any problems?

Thanks again for the replies and I'll keep you posted as soon as I can get my cable working...

 

Sancho,

The CTS doesn't have to ground through its case to the intake. The reference ground I was referring to is the grounded wire in the CTS connector. If there is resistance in that wire, the ECM will "read" a lower temperature from the CTS. You can test the resistance of that wire in the connector to the battery negative terminal. Short of scanning the ECM for data, checking the harness resistance may be your best bet.