hauker

My '08 ML350 (105K miles) started displaying one of the coolant messages ("thermometer floating on ocean" - you get the idea) in the vehicle status message display. ... I double checked the manual: "Coolant too hot". (no surprise there) - FYI, ambient temperature: 96 F (hello Texas).

I checked: (all at rest 5-10 min after turning engine off)
- coolant temp in the display: A little over 80 C (that's not much past the middle)
- coolant level: about a cm over the plastic bracket in the expansion tank (as it should be)
- hooked up OBD-II: no codes stored, real time sensors show coolant temp @ 177 F slowly falling
- V-belt: belt seems to have decent tension (gives a little), but water pump pulley is spinning, no weird noises

According to the manual, a coolant replacement isn't due until 143K ... and not sure know how to check for good vs bad coolant ...?!

So, I am running out of ideas ... bad coolant, sensor, radiator, fan ...!? Any guidance is highly appreciated.

-----------------------

Update: An hour later with open hood (but in 96 F) the coolant temp is still at 152 F. (read via OBD-II)

Interesting observation, though: When I did my tests an hour ago, the radiator fan did not run (which actually surprised me a little at the time). Now, with the ignition "on", just like before (to pwr the OBD-II without starting the car), the radiator fan runs at what seems to be "full speed" pulling the temp down to 130 F in minutes.

Oddly enough, though, twenty minutes later, the coolant is now 107 F (15 deg above ambient temp) and the radiator fan is still running with only a little (if any) drop in speed. (a few minutes later the fan stops ... coolant now at 105 F)

Sooo ... is there a switch that sometime does and sometimes does not work to start the radiator fan ...? And can 'not starting the fan when the car thinks it should' by any chance trigger the warning message I experienced ...??

hauker

update #2:

I drove the car this morning (a few minutes of city traffic @ 40-50 mph): Temperature rose quickly to 80 C (~180 F), warning came on, temp stayed around 90 C. When the car is moving, temp drops to about 80. And after a while, the AC lost it's cool ...

At this point I am convinced, the radiator fan is the culprit as is wasn't running during the whole experience.

I looked in the fuse/relay box in the engine compartment, but don't know what I'm looking at. Does anyone know which relay in the box (or any of the 3 others) is for the radiator fan?

Or should I be looking at a temp sensor? (unlikely as neither coolant nor AC seem to trigger the fan (?), but fan ran yesterday after I let it sit for a while)

Can't find a diagram ...

hauker

Okay, after some research on the interwebs, I am more confused than before:

It looks like there is no relay for this (!?) ...
... so if the fan is the problem, a fully assembly swap will be in order (?)

"Normal" operating temperature" people say is at 80-100 ('C', I assume) ...
... I don't go above 90 C, so why does the "Coolant Temp Hot" message come on in the first place?

After driving the car home today (ca. 15 min), I put a room fan in front of the car and turned the engine on ...
... the fan cooled the radiator down (to the touch), but the coolant temps stayed around 80-90 C with the warning popping up again and again.

I also turned the AC back on (I had left it off to not heat the radiator more) - it wasn't getting cold (as expected) ...
... but neither did the condenser get warm - shouldn't it get hot quickly? BTW the coolant temp rose to almost 200 F (OBD-II) (ca 95 C)

So I am wondering: Could it be the thermostat?
But if so, what does that have to do with the AC not get cold? And why does the fan not start? And why isn't the radiator/condenser getting any hotter? Arghh ....

I have the feeling there is a simple solution, if I can just figure out how all of this is connected and trace it back to the one thing causing it (thermostat, fan controller, water pump, AC pressure, ...)

Help, please ...

hebdog

Your problem could be one little thing, or a whole lot of little things, or one big thing. The reason your A/C compressor is shutting off is because cooling system and HVAC systems are sort of linked. They are dependent on each other working properly for one system or the other to work properly.

The relays for the cooling fan are sometimes built in to the fan control module and are not serviceable and the control module or the whole assembly needs to be replaced.

Start the vehicle and turn the A/C on and the fan(s) should come on. If not there is most likely a problem (obviously), and you need to figure out why.

You can also start the vehicle and get the temps up to operating temp and the fans should come on as well. Driving over 35 MPH will do the fans job hence why you need to be idling to diagnose the problem.

hauker

Thanks hebdog ... I'll try that today. I mostly had the AC off as I was worried about the condenser heating the radiator more ... well, and the fact, that it didn't seem to work. But now that I understand the temps to be "normal", I turn it back on. My wife (it's actually her car) said the AC usually takes a while (at least a minute or so) to kick in (surprised to me), so I'll drive it with the AC on today. Maybe that will kick the fan on ...

Hmmm ... "cooling system and HVAC systems are sort of linked" - I was afraid of that ... any chance there is a good explanation of that relationship out there somewhere (what sensor tells what component to do what, etc.)?

During my observation over the last two few days, the fan seemed to not turn on at all - with the exception of the first days when it turned on after I turned the ignition "on" (w/o starting the engine) in the driveway for a OBD-II readout 30 min after the drive. Then it blew full throttle for about 30 min, cooling the engine from 180 to 105 F and stopped.

But then again, with temps not exceeding 90 C, maybe it shouldn't ...!?

Any idea why the system would show my the warning while also showing a temp range of 80-90 C?

hauker

Okay, I now have 2 diametrically opposed data points in one day of driving the car to work and back (ca. 15 min drive, city & highway traffic):

This morning: (let's call it the "best case scenario")
- ambient temp: 88 F
- coolant @ start: < 40 C
- at start: AC on, radiator fan runs at moderate speed
- during drive: AC works great, coolant 80-90 C, no warning light
- at end of drive: AC still works, radiator fan is running
- after drive (engine of): radiator fan turns off and spools down
(... perfect, the way it should be ...)

This afternoon: (let's call it the "worst case scenario")
- ambient temp: 100 F
- coolant @ start: 50-60 C
- at start: AC on, radiator fan does not run
- during drive: AC does not work, coolant 80-90 C, warning light comes on
!!!: at the same time of cooling light, AC airflow drops significantly !!!
- at end of drive: AC still not working, radiator fan still not running
- after drive (engine of): radiator fan does not turn on either

So, hebdog, you're obviously right with "cooling system and HVAC systems are sort of linked" ... question is
a) how, and
b) what's causing this?

Temp sensor (s)? Radiator fan? AC system? ... (which, and how can I test?)
(based on the coolant range, at least I think the thermostat is okay)

Any good suggestions ...?

hauker

Okay, based on everything so far, I think I can rule out the coolant system. (belt, pump, thermostat, temp range, ... all checks out).

I also checked the AC / fan side:
- compressor (pulley spinning, no idea how to check if clutch is engaged)
- refrigeration (works, small line very hot, big line ice cold)
- fan connector (cleaned, no visible damage)
- wiring to fan (no visible damage)

The AC hasn't worked at all over the last 48 hrs, so I'm guessing I'm dealing with either:
- refrigerant (pressure, charge, blockage)
- fan controller,
- sensor(s), or
- the electronics in between (including the ECU - hopefully not)

How would I test for any of these?

Could it be as easy as simply low pressure in the AC lines preventing the fan to kick in, thereby heating up the refrigerant and that triggers the warning light? (of course then I'd probably have a refrigerant leak)

Help, please ... ... I am out of ideas.

hebdog

I will look it up tomorrow when I get to the shop. Those systems are linked by various different things, way to many to list. I just spent 4 hours diagnosing an AC problem on a car that I replaced a suction line on. The low pressure switch that came with the new line was faulty and the connector was corroded. Put the old sensor back on and new connector and all is well for now.

The AC clutch will spin on the front of the compressor if it is engaged. If the fans are not being commanded to turn on for what ever reason the compressor will not engage. It is a fail safe to protect both systems.

Sounds to me like you have a fan motor or fan control module going out which is causing your sporadic issues. Both those problems unfortunately are expensive and you will need to replace the whole fan assembly, usually. They typically don't sell the parts separate.

If you have a multimeter or test light, see if the fan control module is getting proper power and ground.

hauker

Thanks hebdog - I look forward to what you can come up with tomorrow ...

I looked at the compressor and what I think is the clutch (inner disc) is spinning as well, but at standstill it doesn't look like a clutch to me (and somewhere I read that these cars have a "clutchless" compressor ...?)- also, wouldn't the fact that the coolant line gets cold imply that the compressor is actually working ...? (coming to think of it, could the fact that despite the large temperature differential between the two lines, the AC still isn't working, imply that there is either a blockage or a low charge ...?)

I do have a multimeter and I looked at the plug - it's 4 prong. I assume I'm looking at 12V, ground, and two other dynamic signal line of sorts ... but no idea, nor which is which. How would I go about checking/measuring this? (thx)

N_Jay

I am guessing since I don't have any direct experience with failures like this on MB (yet)

Cooling system is good, since any airflow (fan or moving) keeps the coolant in or near the normal range 80-100C

AC is system is good since when the cooling systems seems normal, the AC works.

Problem is probably in the fan controller since the fan is capable of running (eliminating the fan as a problem) but is not running when it should.

The AC is probably being shut if by its controller when it "Thinks" that car is overheating, or maybe even proactively when it can't get the fan to run.

The overheat signal on the dash is probably triggered by the same problem.

Remember, our cars are networks of computers, and one controller either ignoring signals or sending out erroneous information can (and often will) show up as failures of other systems.

I am ruling out the main temp sensor (in the thermostat housing) since the temp reading seems right. There might be other temp sensors or wiring to the temp sensors that are part of (or the whole) problem.

Since it is all happening together, I would look for one problem before looking for a bunch of problems.

Start with the fan controller and/or its wiring.

hauker

Thank you, N_Jay ... I agree with your line of thinking.

The wiring has no visible external damage ... how do I best go about testing the fan controller and the wiring?

For the record: The AC does not blow cold air (but warm) ... all that is "working", is that the pipes get hot/cold - so I assume the compressor is doing something (thereby working), but I am wondering if a low refrigerant charge could be the culprit as well ...?

N_Jay

I would get the fan issue fixed first, the go back to dealing with the AC.


Seems most of the controllers on our beasts are "Toss in the trash and buy a new one" repairs.


Maybe this could help, (Maybe not)
http://www.peachparts.com/shopforum/...erheating.html

hauker

Thanks for the link, N_Jay - unfortunately, it's a different car, different engine, different fan ... and no real answer to the question posted.
But yes, my car's fan assembly includes the controller, so if it turns out to be the issue, it'll be a replacement of the whole assy. ($700+ OEM, ~$350 after market from I can tell)

I drove the car with no air on at all and the warning did not come on (neither did the fan as far as I could tell) - further proof, that the coolant system seems to be okay.

I think I should test the power signal(s) to the fan, and (if possible) the fan itself next - does anyone have a good procedure to do so and/or pin-diagram of the plug leading to the fan, so I know what to test for (and not make things worse by shortening something out)? Thx.

hauker

... got any new suggesting for me, hebdog?

... hmmm, this should be a "variable displacement compressor" (Denso part-# 471-1594) if I am not mistaken, so it actually does not have a clutch. However, when I turn the AC on, the refrigerant lines do get hot & cold despite the fan not running, so I think the compressor is working ..!?

At this point it's the likely culprit. Fan's not been running for days now ... then again, I've had the AC off as well. The AC is only blowing hot air at this point, the fan not turning on, and this kicks off (or at least greatly accelerates) the coolant temp warning.

... I hear you. I believe an OEM part is ~$700 (+$300 labor), but the jobs looks pretty straight forward, and I found "OEM replacement" fans online for as little as $300 ... anyone ever have any luck with these?

That said, while $300 is much better than $1000, I'd hate to order one, wait 5 days to get it, only to find out, that wasn't it. So, how can test the fan/signal to confirm it's the fan?

update: I figured out that the big two prongs are 12V & ground and measured 12.47V (with everything off - 14.09 with engine running) and very little resistance to ground.

Here are my measures for the little prongs (looking at it with the little ones top, big ones bottom):
left/right: 0 V / 0 V (off)
left/right: 12.47 V / 0.36 V (on/ no engine)
left/right: 12.47 V / 0.36 V (on/ no engine / AC)
left/right: 14.09 V / 3.2 V (on/ engine)
left/right: 14.09 V / 3.2 V (on/ engine / AC)

What does that tell me? Is there a way (via the two small prongs) to "tell" the fan to run in order to test it for real ...?

N_Jay

If you want the controller, maybe buy a fan with broken mounts from a user parts place and do some surgery.


car-parts.com for example.

hebdog

Bud I am sorry I didn't get back to you sooner. My wife's A/C quit and have been dealing with that, as well as everyone else's A/C problems lol! I will check for you tomorrow when I get to work.

I still feel like your fan control module is the culprit. The wires don't have to be damaged. It's an electrical part and sometimes they just fail with out warning. I would have to try and find a pin-out diagram and see what pin should be seeing what voltage and when. Hopefully it will be in our data base at work. If not as someone else suggested, you can always buy a used unit and just rob the module. Doesn't garantee that your fan motor itself isn't what could be wrong either. The motor could be what caused the failure of the module, if in fact that is what is wrong. If we have a failure on one or the other we will sell the whole assembly. It's more in the long run but causes a lot less headaches.

Going aftermarket is ok, just try and do some research and make sure it is truly compatible with the picky MB electrical system.

hebdog

1. Check all fuses. Inspect connections on the Engine Coolant Temperature (ECT) sensor and A/C pressure sensor on the receiver-drier.

2. Connect a M-B fully functional compatible scan tool and retrieve/record/clear DTCs in the Powertrain Control Module (PCM), Signal/Activation Module (SAM), Automatic Air Conditioning (AAC), and Instrument Cluster Module (ICM).

3. Attempt to access the activations menu in the scanner under the PCM. If available, use a scanner to activate the engine cooling fan and control it using the scan tool. If the fan is not controlled, check voltage signals at the Fan Control Module (FCM) (M4/7) located on the fan shroud.

4. Use a Digital Multimeter (DMM) for voltage testing at the connector on M4/7. Pin 2 Red (RD) should be battery voltage (B+) at all times. Pin 3 Pink (PK) should be B+ with the key in the on position (#2). Pin 1 Brown (BN) should be constant ground. Pin 4 Yellow/White (YE/WH) should be the control signal from the PCM, depending on the control signal from the PCM the voltage will vary between 3 V (low speed) and 11 V (high speed) approximately.

5. Check the version coding of Central Gateway Module (CGW) using the scan tool. Check the version coding of SAM module. Check the version coding on the powertrain control module.

Never mind the scan tool part, unless of course you have one lol. You want to pay special attention to #4. If the electrical voltages check out that means the module is getting what it needs to function. At that point you will know for sure that it is the fan control module and or fan motors.

hauker

Awesome ... thank you for the elaborate response, hebdog!
Done, (all relevant) fuses seem okay. Didn't "check" the ECT (I assume that's the one close to the thermostat?), but given the temp readings, I assume its okay.

How do I go about that?

lol, indeed ... don't have one and am trying to avoid going to a tech as long as I still harbor a faint hope that I can DIY the job (kind of think one should do as much work as possible on ones car themselves ... I'm funny that way)

Now, we're talking ... I'd say that very well aligns with my readings (see my last post - and wire colors match with pins) ... would you agree? Can I take then, that this is a sure-fire sign the signals are good, but the fan/controller may not? Does a bad fan motor/controller create a feedback that prevents the AC from running and trigger the coolant temp warning despite acceptable coolant temp? - In which case my entire issue may trace back to (and could be fixed by) the fan assembly ...!?!

See my comments on #4 ... am I blinded by wishful thinking or is it time to order a new fan assembly? The one I am currently looking at is aftermarket (< $300) and on ebay (from a guy with predominantly positive feedback (982 score) and a good amount of sales activity. The pictures look like the real deal (like the original, in fact - let's hope what I would get is what is posted and good quality ...).

hebdog

Yes. If the readings from the ECU side are in spec then the fan control module is being told to engage, but it can't cause there is a problem. I would probably go with replacing it.

hauker

I'm actually not sure I've fully determined that yet ...

Two thoughts:
a) If the fan is the issue, then why is the AC not working at speed (say 60 mph), and
b) I got 3.2 V on pin 4 when idling ... what about with more heat load? (which by your information should be up to 11 V)

After driving home stop-and-go in 96 F, the coolant temp symbol came on towards the end (no AC on, but temps still read 80-90F in the display). I kept the car running in the driveway and measured the voltage on pin 4, then turned the AC on to create more "load" (coolant temp now read ~105 F) and measured again: Either time, the voltage on pin 4 was 3.2 V. (... sigh ...)

Shouldn't the signal be higher at that point? ... then again, if the fan was working, shouldn't it have kicked on (slow) at 3.2 V?

I hate to even think this, but could it be that the fan isn't the issue after all, but either the PCM or a temp sensor?

N_Jay

I'm betting the fan controller is a smart box and it is bad.
It is either sending a signal saying it is bad,
or not properly responding to the signals it is sent,
and that is triggering the high temp warning and not allowing the AC to run.

hebdog

^correct. The module isn't really smart, it just does what it is told to do by the ECU.

Your A/C compressor will not enagage if certain conditions aren't met, like the fans not enagaging. The ECU will send the signal then monitor the voltage to see if they turned on. It is detecting that they are not so it then tells the compressor it is not ok to enagage. Unless you are back probing with the connector plugged into the module the readings will be what they are according to the steps I posted earlier. Those diagnostic steps I posted were to ensure the module is getting the proper signals, power, and ground. It is and is failing to work as designed. That means to that connector everything is within spec and everything after needs to be tested, except in this case because everything past that connector is the fan assembly.

hauker

Thanks N_Jay & hebdog - I think I see what you're saying, and it kinda makes sense ...

So ... getting ground on pin 1 and 14 V on pin 2 & 3, combined with 3.2 V on pin 4 is what is expected, telling us the ECU is doing it's job. But as the fan isn't spinning (which it slowly should at 3.2 V), it is the culprit and therefore needs to be replaced.

And as long as I have the plug in my hand measuring an open circuit of pin 4 against ground, I won't be getting more than 3.2 V from the ECU regardless of the "load" condition as it is not getting any "feedback" via voltage fluctuation or amp draw. And it won't simply send a "dumb" 6...8...11 V signal down the line. (did I get this right?)

And even at speed, the ECU is preventing the AC from kicking in, because it monitors the fan (even though it doesn't plan to turn it on at that point). (correct?)

Which probably also is why I am getting the warning light after a while, even if the AC stays off ...!?

hebdog

Yes.

hauker

Thanks ... I'll assume that goes 4x (for each of my annoying questions) ...

I'll go ahead and place an order for the fan and will let you know how it went once received & installed.

Thanks again for all the help.