Subaru Outback Forums banner

1 - 20 of 31 Posts

·
Registered
Joined
·
17,704 Posts
Discussion Starter #1
According to Subaru factory service manuals, the engine control module (ECM) is supposed to turn the fans on and off at specific engine coolant temperatures (ECT), depending in part on the vehicle speed and whether or not the air conditioning is on.

I've been monitoring both coolant temperature and fan operation using an engine data scanner (in my case, BtSsm). With the warmer summer temperatures, I more often see the ECT go up above it's usual level of ~188 F, when in stop-and-start traffic or parked but idling.

According to my 07 FSM, with the AC off, the fans should come on at low vehicle speed at between 205 and 210 F, and at high speed if the coolant reaches 212 F or more.

I'm finding that the fans come on at low speed or stopped only after the coolant temperature reaches 212 - 214 F. I've never seen it get higher, so can't say when the high speed fan mode would kick in. (Note: I have tested the fan operation using the "green" connectors -- the fans cycle properly between off, low and high speed.)

The ECT data is from the ECM, and that is what the ECM is using to control the fans, so it looks as if the ECM isn't turning the fans on as early as I was expecting.

I appreciate that the actual difference is small, but I'm wondering if this is typical.

I'm looking only at when the fans turn on when stopped in traffic or parked and idling, so that the effect of air flow through the engine compartment is minimized, and AC off.

If you're using BtSsm, or Romraider, or FreeSSM, Torque, or any other app/tool that shows engine coolant temperature from the ECM, at what ECT do your fans turn on?

[Note: some apps, such as BtSsm, Romraider, and I believe FreeSSM, and perhaps others, can also show the status of the fan controls (0 = off, 1 = on), but just listening for the fans to come on and noting the temperature indication would probably work, especially if it's repeated several times.]

The FSM specification for ECT versus fan operation varies slightly between models and years, but the issue is whether or not the start temperature is indeed what the FSM says it should be, or is it generally on the high side. See attached table.)
 

Attachments

·
Registered
Car: 2008 Tribeca, 2010 LGT, Sold: 2005 XT Limited
Joined
·
2,926 Posts
I haven't gotten around to setting up a logger for my iphone, soon, soon.

However, I wonder - does the ECU actually report temperature, or a digital number that btssm is converting to a temperature? If that conversion factor is off, that could explain the difference. It might be correct at ambient, but off when warm (ie btssm is using a simple linear conversion while the ecu is using a polynomial or vice a versa).

Can you change the units to report digital results?

Tom
 

·
Registered
Joined
·
17,704 Posts
Discussion Starter #3
Tom

I'm not sure what you mean. Can you expand on the idea?

What I do know is that both BtSsm and Romraider use the same parameter definitions, which include the parameter's unique "address" in the ECM, and a formula for converting the data to, in this case, F or C. In the case of C (which is what I normally use), the formula is C = x-40, whereas for an F readout, the formula is F = 32+9*(x-40)/5. From that it does look as if the data from the ECM, namely "x", is not C or F itself but a value which happens to be |40| higher than C. The conversion to "C" is obviously linear. These formulas are used both in BtSsm and Romraider, and perhaps other scanning tools.
 

·
Registered
Car: 2008 Tribeca, 2010 LGT, Sold: 2005 XT Limited
Joined
·
2,926 Posts
Most of the temperature sensor are NTC thermistors. (https://en.wikipedia.org/wiki/Engine_coolant_temperature_sensor) The ECM is reading a digitized resistance (or voltage drop). The voltage drop is then converted to resistance and then the temperature using either the Steinhart–Hart equation or a linear approximation of it or a lookup table. It was not clear to me if the ECM reported the converted temperature, the resistance or the voltage.

Based on what you state, it appears it is reporting a digitized offset temperature, with a gain of 1. Makes sense, as it doesn't ever want it go go below 0, and probably has a max of 255 DN (or 215C, and can read as low as -40C) to keep it to 1 byte. However, is it possible the ECM is using a gain that isn't exactly 1? That would cause the discrepancy you are seeing.
 

·
Registered
Joined
·
17,704 Posts
Discussion Starter #5 (Edited)
However, is it possible the ECM is using a gain that isn't exactly 1? That would cause the discrepancy you are seeing.
If that is causing the discrepancy, wouldn't it affect most, if not all, engine PID readers across most if not all ECMs of the same year/generation, or more? If we get numerous responses that have similar discrepancies, that could well be the case. But if most, if not all, of the responses indicate the switching taking place in the middle or low end of the FSM range, rather than at or beyond the high end, as in my case, it would suggest the "discrepancy" is unique to my situation and not due to a software design that isn't 1, which, in effect, is what I was wanting to determine at the outset.

There are complications to this. Romraider, FreeSSM and BtSsm use the Subaru-specific ssm data stream to derive their PID readouts. Generic scanners generally read on the OBDII stream. There could be differences between the two such that the fan switching in one could appear to be off, while the other is not.

I believe Romraider logger can be switched between the ssm and OBD data streams. I'll try to have a look at this tomorrow, and if it works on my 07, see if the temperature readings when warmed up are consistent. [Update: the adapter cable I have to connect to the ECU via the OBD port does not connect to the OBD data -- it's ssm-specific. Although Romraider could read OBD data, it requires a different cable.]

I do appreciate your interest.
 

·
Registered
Joined
·
13 Posts
This might be going above my head. But my Subaru also does not turn on the fans, with the AC off, till 212F (According to my Ultra Gauge).

Does this mean something is wrong with my Subaru?

Or that my reader is not displaying the correct temp?
 

·
Registered
Joined
·
17,704 Posts
Discussion Starter #7
This might be going above my head. But my Subaru also does not turn on the fans, with the AC off, till 212F (According to my Ultra Gauge).

Does this mean something is wrong with my Subaru?

Or that my reader is not displaying the correct temp?
Interesting that you have the same apparent "turn on" temperature.

However, in my case I haven't seen the instrument panel temperature gauge go above the normal position (meaning the coolant temperature is above 212 F) as seems to be happening in your case.

Your specific car's issue is being discussed in http://www.subaruoutback.org/forums/109-gen-3-2005-2009/379177-2009-outback-limited-cooling-questions.html.
 

·
Registered
Joined
·
17,704 Posts
Discussion Starter #8 (Edited)
A learning/update:

I found that the ECU ROM (the memorized/stored data that the ECU uses to determine what/how/when to set controls for different situations) has three tables for patterns of radiator fan control (this applies to my 2007 2.5 non-turbo ECU -- it might not be the same with others). I've attached a screen shot ("Fan modes") of the three tables plus another relating fan control to vehicle speed.

I don't know why or how the ECU would select between the three tables. But I believe that in each table, M0 is fans off, M1 is fans on low speed, and M2 is fans on high speed. What's I don't understand is why there's more than one table. It seems rather significant that in the first one (Radiator Fan Mode A . . .) the temperatures at which there would appear to be a change in the fan status (between M0, M1 and M2) are about 8 degrees F lower than the other two tables. (The tables Mode B and Mode C are identical.)

Nevertheless, given my earlier observation noted in post #1 above, where the fans were coming on at slow speed (M1) at 212 F (100 C), the ECU was using Mode B or Mode C.

Today I ran a simulation, substituting a variable resistor for the ECT sensor, and monitored the ECT readout and the ECU fan control status using BtSSM (and also watching the fans themselves). With the key at on (engine off), I varied the resistor from high (~4.6 kOhms) which corresponds to about 46 F, to low, simulating the coolant increasing to above 212 F. Interestingly, the fans kicked on at slow speed at around 203 F (95 C), and switched to high speed at around 212 F (100 C) just as the "Mode A" table suggests. The test was repeated while logging the data, and is presented in the attached chart ("Rad Fan response"). In the chart, I also slowly simulated the coolant temperature dropping and noted where the fans changed from high speed to low to off.

It appears that my earlier observations were made when the ECU was in Mode B or C, but sitting in my garage, with everything cooled down, and a few months later, it was using Mode A.

The questions that goes begging for now are:

What criteria does the ECU use to "decide" whether to follow Mode A or Mode B/C?

Why would it use the latter at all, where the coolant temperature is allowed to get noticeably higher?

(Edit: Back in March, I was caught in a traffic tie up, and noticed that the fans would turn on at around 203 F (95 C) and then go off when the temperature dropped below that. At that time the outside temp was slightly below freezing. But in August, when it was much warmer outside, I noticed the fans were starting to run at around 212 F, and that led to starting this thread. So it does seem that the ECU is selecting different Mode tables under different conditions.)
 

Attachments

·
Registered
Joined
·
13 Posts
That is exactly what I wanted to try but lacked the equipment/know how to do. Thank You.

And with that info I am going to assume the ultragauge is displaying a temp that is correct (Or at least matches what the ECU is thinking)

And since you saw those temp levels in the ECU... I am going to leave well enough alone unless I ever get in the red.

It seems like the Modes should be dependent on vehicle speed. I know the maintenance manual mentions some changes based on that also. I wonder what mode "parked" is considered.

I wonder if there is a way to get a Subaru Engineer to explain how this works as far as the modes go, and maybe why they decided to do it this way.
 

·
Registered
Joined
·
17,704 Posts
Discussion Starter #11
I would still be interested in a meet up if you wanted to pull what my Rom's tables look like.
I didn't pull my ROM. I have always used Romraider (romraider.com) and the good folks over there had "defined" the ECU that is in my car. That provides the ability to display the ROM tables in the Romraider Editor, which is the first attachment. Romraider is an excellent program and covers most current Subarus. However, it's normally used with a Windows laptop which is okay for occasional troubleshooting, but not to have running whenever the car is used (as a "dashboard") similar to the way Ultragauge works.
 

·
Registered
Joined
·
17,704 Posts
Discussion Starter #12
An update:

Been watching the ECU fan control over the summer; the fans continued to be turned on (slow speed) at around 212 F (100 C), i.e., Fan Mode B. However, a few weeks ago I was in a long line waiting to get through Customs. The outside temperature was low-mid-70s, and cloudy. We were moving forward intermittently, and just leaving the engine idling in gear was enough to advance the car the one car-length at a time.

During this time I noticed (and was surprised) that the fans were being turned on at around 203 F (95C), i.e., Fan Mode A.

That observation and the one back in March (see Edit to post #8 above) confirm that the ECM does switch between the Modes.

If the two experiences are at all indicative, it seems as if the combination of extended idling accompanied by some slow vehicle movement might be among the factors that determine which Fan Mode the ECU uses. It's as if, after some time, the ECU assumes that the car isn't going to resume normal driving any time soon and the risk of overheating from lack of air naturally flowing through the radiator and around the engine, combined with the relatively slower circulation of the coolant, justifies running at a lower temperature. (Higher temperatures are used to help reduce emissions.)

At this point it's just an early hypothesis, but it does demonstrate that the engine management is far more complex than we might imagine . . .
 

·
Registered
06 Outback 2.5i
Joined
·
397 Posts
@plain OM, have you revisited this since last fall? Perhaps it's a bug in the firmware or there's a TSB about this?

To me, 212F seems way too late to be kicking on at low speeds, especially since anything above 212F causes the gauge to go above the normal position.
 

·
Registered
Joined
·
17,704 Posts
Discussion Starter #14
@plain OM, have you revisited this since last fall? Perhaps it's a bug in the firmware or there's a TSB about this?
No, I haven't.

I don't think it's a bug; there's two fan control tables in the ECU ROM (post #8 above), one of which clearly calls for the fans to first come on at 212 F. Subaru built that into the operating system. The question, however is: When, and why, would the ECU use the 212 F table instead of the 203 F table (which I have also seen in operation)?

To me, 212F seems way too late to be kicking on at low speeds, especially since anything above 212F causes the gauge to go above the normal position.
I've seen the ECT indication on the tablet go above 212 F (actually, I'm monitoring in C, so that's 100 C) by a degree or two, but at that point the fans are on and the temperature falls rapidly. In those cases I never saw the cluster gauge pointer noticeably off the "normal" position.

I suspect that a pointer width above normal isn't a concern as long as the fans are on and the actual coolant temperature is indeed falling back. It's when the pointer is clearly above normal that I'd be concerned.

In your case, what puzzles me is that the car seemed to consistently run much hotter on the uphill slopes, whereas few, if any, others have reported similar regular high temperature reactions. I haven't seen the ECT on my 07 do that, but admit I'm rarely driving in areas with long hill climbs in hot weather. The few times I was, it might have gone up a few degrees. e.g., from 87 C to 90 C, but that's still well below any concern.

I'm following your rebuild thread (your overheating thread being paused for the time being), and will look forward to the result of the rad replacement.
 

·
Registered
06 Outback 2.5i
Joined
·
397 Posts
The interesting thing is the FSM has no mention of the second table. It simply states if the rad fans do not kick on at 205F (98C) / under 19 MPH, they need to be troubleshooted.
 

·
Registered
Joined
·
17,704 Posts
Discussion Starter #16
The interesting thing is the FSM has no mention of the second table. It simply states if the rad fans do not kick on at 205F (98C) / under 19 MPH, they need to be troubleshooted.
Based on the fan operating table (post #1 above) I too thought that the fans should come on at around 205 F; that's what led to me starting this thread!

However, unless I've missed it, there's nothing that actually says to troubleshoot if the fans don't come on at 205 F.

There's only one trouble code that deals with fan operation based on speed and temperature, namely P0483, and it uses 212 F (see attached).

I should note that the fan Mode tables in post #8 above are from the ROM in my 2007, 2.5, non-turbo, Fed-spec engine. The fan tables in the ROM in your 2006, CA-spec could be different (although I doubt it).
 

Attachments

·
Registered
06 Outback 2.5i
Joined
·
397 Posts
I actually have a federal emissions car in California. It was originally sold in Colorado :) But that's a moot point, as the cooling system FSMs are the same between California spec and Federal spec vehicles.

Take a look at the 2nd page of the attached PDF:

B: INSPECTION
DETECTING CONDITION:
• Engine coolant temperature is above 96°C (205°F).
• Vehicle speed is below 19 km/h (12 MPH).
TROUBLE SYMPTOMS:
Radiator main and sub fans do not rotate under the above conditions.

It then goes on to say if both fans are turning on low-speed mode and high-speed mode when directed by SSM, then the fans are good.

So this second 212F map is either a bug, or an "undocumented feature" :laugh:
 

Attachments

·
Registered
Joined
·
17,704 Posts
Discussion Starter #18
mea culpa I missed that page and appreciate you coming back with it.

The "feature" might be undocumented, but is implied. If the fans don't come on at 205 F, but work in the forced test, then when would they come on? There must be some point but that's left out. P0483 implies 212 F is another possibility, because there it seems to say the fan control goes from off to on at that higher temperature.

I chalk it up to poor technical writing.

What I do know about my 07 is that under certain conditions Mode A is indeed used; that includes waiting in traffic for an extended period (see also post #12), and when the engine is off, and cool, as was demonstrated in the "simulation" reported in post #8.

I'm wondering if you ever logged the fan operation when stationary in gear (D), or at best very slow moving, stop and start, situations. Perhaps your ECU will also employ the lower temp control pattern. Was this done in the overheating thread? ("Fan operation" meaning the 0/1 off/on control signals from the ECM for the two fan relays. They are under the "Switches" tab in Romraider.)
 

·
Registered
Subaru Outback 2.5 CVT (2012MY, for Europe)
Joined
·
6 Posts
@plain OM, hello!

I apologize in advance for my English...
You could understand in what case the ECU uses table A or B (C)?

In my car Outback 2.5CVT (2012 MY, for Europe), two tables are used - A and B. Moreover, the values in table A are different from yours, and table B is the same as yours.
In order to slightly lower the temperature of the engine, I replaced the standard AA200 thermostat with AA230, which has a full opening temperature of 91C (196F) instead of 95C (203F).
In the ECU Subaru firmware for the thermostat AA230, the temperatures in table A are 3-5C lower than for AA200, but the values in table B are almost the same (only two parameters are lower by 1C).
I have been traveling by this my car for 3 months and during this time I noticed that table B is used to turn on the fans.
Trying to figure out how best to change the temperatures in the Radiator Fan Modes A / B (ECT) tables.
 

Attachments

·
Registered
Joined
·
17,704 Posts
Discussion Starter #20
You could understand in what case the ECU uses table A or B (C)?
No, I have not found out in which case the ECU uses table A or B, other than what I reported earlier, namely, that it seems to depend, at least in part, on the driving conditions at the time.
 
1 - 20 of 31 Posts
Top