Subaru Outback Forums banner

engine temp gauge

18K views 45 replies 5 participants last post by  plain OM 
#1 ·
Does anyone have suggestions on what to do regarding the temp gauge on the dash showing a full needle width cooler than before? This is after a new OE thermostat and Subaru coolant replaced on the '05 outback.
Perhaps still air in system or something else?

Thanks.
 
#2 · (Edited)
Where was the needle before? Perhaps it was on the high side and now that's been corrected with the new thermostat. The only way I can think of to get a better idea of what's happening now is to connect a scanner tool to the OBD port and monitor the engine coolant temperature. The "normal" range is between 185 F and 212 F (85 to 100 C).
 
#3 ·
Attached are the pics that show where the gauge normally use to sit driving down the road. Now it sits lower as reflected in the second pic. When sitting at a light it goes back up to the "ideal" reading but when i am off the pedal like going on a rolling road it goes down to the lower picture.
Prior to the fluid change it never fluctuated. Also it never goes about the halfway mark on the gauge luckily.

I can plug in and read ECT when driving to see if it drops below 160.
 

Attachments

#7 ·
I really don't see a problem, seeing as how the gauge doesn't really show temperature, but range. You'll be fine, as long as the "range" doesn't go into the upper, or red, level. The reason it's lower is because you just flushed the system and replaced the thermostat. All looks well.
 
  • Like
Reactions: SlperOB
#9 ·
The gauge is designed to stay at/near the 9:00 o'clock "normal" position for coolant temperatures between 185 and 212 F. If the coolant temp is lower, or higher, than that, the needle moves away from the "normal" position. In other words, the pointer doesn't move if the coolant temperature varies between, say, 190 and 210. If the ECT is monitored over a range of driving conditions, it will be apparent that the ECT varies while the gauge appears to remain unchanged. However, when as the ECT is above 98 F (roughly), the radiator fans should turn on. If the gauge pointer itself goes up by a pointer-width or more, meaning the temp is above 212, then the fans should be on at high speed. The fans should be able to maintain the temperature within the "normal" range. Anything above that, even if it's not all the way to the red at the top of the gauge, is an "overheating" situation.

I don't see a significant difference in the gauge pointer position, but that might be the difference in the camera capture area. If a scanner is available, it's simple to verify if the coolant temperature is in the normal range, or is low.
 
#8 ·
pics look the same.

I doubt you have anything to be concerned about.

if you still have the old t'stat, you could pull the new one and do a stove-top comparison. use a good thermometer to compare opening temps.

frankly, I don't think it would be worth the effort.

do the t'stats both look like the large wax capsule type that Subarus require?

 
#10 ·
1Lucky Texan---yes it is the OE Subaru Tstat just like the pictured on. Not a fan of aftermarket products.

I will drive vehicle with scan reader and see what ECT is when the gauge is a bit lower than before. I guess with aluminum heads I would rather be running cooler just not sure why this changed on the gauge.
 
#11 ·
Here is the new info...Drove vehicle with scan tool to read ECT. With outside temp ~35F at 40mph in city when the gauge was at the original position (a needle width below "9 o'clock" position) the ECT temp was 181-183F. When approaching a stop light or a slight down pitch in the road, so off the gas, the ECT would drop to 174-175 and the gauge would be 1 to 1.5 needle widths below the ideal position. Once accelerating again the temp rose back to 181-184 and gauge position was as before tstat and coolant change. (I never recall seeing the gauge fluctuate like that before the new thermostat. Only when going down a long hill at highway speed in the winter would the gauge drop down several widths then stabilize and climb back up when back on the gas on flat ground.

No sure if the radiator is functioning more efficiently now after flush and new coolant or if new tstat is working more opening and closing than the old one??
 
#12 ·
Sounds to me like you did a satisfactory service. I'd also say that the new thermostat opens sooner than the old one, and reacts quicker, because it's new. I'm glad you where able to test it out, because now you know for sure. Be happy that everything is working as intended. Congrats!
 
#13 ·
The ECT is at the low range of the "normal" position of the gauge. As noted earlier, the gauge remains at the same place starting from about 185 F. As yours is running around 183, it's on the lower edge. When the car is coasting, the coolant temperature drops and, because the gauge system was already at the lower edge of "normal", the gauge pointer starts to track downward with the lower temperature.

But might I ask: Was the test drive done with the cabin heater running? If the HVAC fan was running, it could be adding cooling causing the ECT to be low, especially if the outside air is down around 35 F.

My 07, once fully warmed up, seems to be remain above 85 C (185 F). I have the ECT displayed almost all the time and I don't recall it dropping down below 85 C when slowing to a stop or going down a hill off-throttle.

It would be nice to have comparable data for other, similar, engines.

It certainly appears that the new thermostat is opening earlier. This might not be a problem as long as it doesn't drop too low. The ECM uses the ECT to determine fuel trims and timing, and a too low running temperature could lead to increased fuel consumption.
 
#14 ·
But might I ask: Was the test drive done with the cabin heater running? If the HVAC fan was running, it could be adding cooling causing the ECT to be low, especially if the outside air is down around 35 F.

It certainly appears that the new thermostat is opening earlier. This might not be a problem as long as it doesn't drop too low. The ECM uses the ECT to determine fuel trims and timing, and a too low running temperature could lead to increased fuel consumption.

The test drive was done with the heater/ac off. I have manual controls and the fan was off and the temp knob all the way set to cold.

During the drive i did watch the STFT and LTFT; the STFT was hovering at 0 both sides and the LTFT about 3.5 - 5 each side.

It does seem like something is causing the coolant temp to run cooler than normal. The fans were not running at any point either. I did mix the new coolant to about -38F with distilled water. The old coolant was -25F not sure if that alters the ECT or if its not significant enough.
 
#15 · (Edited)
No, the small difference in the mix should not be a factor here.

Given that even my 07 is running near the lower range of "normal", I wouldn't be surprised if small differences in production would result in even my engine temperature being lower (or higher) if I were to change the thermostat. I'd probably keep an eye on the temperature for a while, and then accept it as good, especially if, during colder weather highway driving, the temperature doesn't drop down a lot. (In this situation, there's lots of cold air flowing by the engine and a lot of heat is dissipated from this natural flow. The thermostat should still be able to maintain the temperature.)

Edit: A review of the FSM revealed that there is an OBD check for coolant temperature malfunction. The code is set if, in two consecutive driving cycles, the measured coolant temperature remains at or below 167 F when the ECM determines it should be higher. Also, the FSM indicates that with a fully warmed up engine, at idle, the ECT should be no less than 167 F. Given this threshold, it looks as if your situation is within the acceptable limits.

During the drive i did watch the STFT and LTFT; the STFT was hovering at 0 both sides and the LTFT about 3.5 - 5 each side.
This is off-topic, but I'm interested in what your scanner is reading; does it indicate fuel trims for each side (Bank 1 and Bank 2)? I didn't think that was available from the ECM.
 
#17 ·
Edit: A review of the FSM revealed that there is an OBD check for coolant temperature malfunction. The code is set if, in two consecutive driving cycles, the measured coolant temperature remains at or below 167 F when the ECM determines it should be higher. Also, the FSM indicates that with a fully warmed up engine, at idle, the ECT should be no less than 167 F. Given this threshold, it looks as if your situation is within the acceptable limits.

This is off-topic, but I'm interested in what your scanner is reading; does it indicate fuel trims for each side (Bank 1 and Bank 2)? I didn't think that was available from the ECM.
Well at least i never see ECT that low during normal driving. I will wait and see i guess. I just know the gauge has never responded so quickly before.

Yes fuel trims are being read from first o2 sensor on each bank.
 
#16 ·
Just to throw this one out there. Of course...keep in mind, modified turbo car...but I also have an aftermarket race radiator, all aluminum. The trans cooler is also an external aftermarket.

I run an OEM t-stat. I have both a temperature gauge tapped into a coolant line, right after it exits the radiator..and the stock coolant sensor in the block. The stock gauge in my car reads...dead on 9 0'clock. Like your second picture of your gauges. If I am logging with RR, the stock coolant sensor reads usually right on with the Innovate gauge. Sometimes it will be 1-2 Celsius hotter.

Frank consistently runs at 88 C when driven normally...That's 190 F. I suspect it was close with the factory cooling, the needle has always been where it is now. It just never gets really hot like it used to. The stock gauge used to creep up under had driving. I have an additional oil cooler also, my main goal was to fight heat soak during / after hard drivng.

I have never seen her hotter than 101 C...and I was flogging her then.
 
#18 · (Edited)
I was out today in ambient temp just below freezing and was specifically watching to see if/how the ECT varied. After the engine reached operating temperature (gauge at it's "normal", about 1/2 pointer width below 9:00 o'clock), in city and highway driving the ECT was between 86 - 90 C. When decelerating off throttle, including a long highway exit lane and ramp, the ECT didn't show any sign of dropping.

During the summer with outside temps up around 30 C and in slow stop and start driving, the ECT did increase, beyond that range, but this was controlled effectively by the radiator fans.

Again, I'm not sure there's anything to be concerned about, but more comparisons would be reassuring.

In regard to the gauge, the attached might be of interest.
 

Attachments

#19 ·
I was out today in ambient temp just below freezing and was specifically watching to see if/how the ECT varied. After the engine reached operating temperature (gauge at it's "normal", about 1/2 pointer width below 9:00 o'clock), in city and highway driving the ECT was between 86 - 90 C. When decelerating off throttle, including a long highway exit lane and ramp, the ECT didn't show any sign of dropping.

During the summer with outside temps up around 30 C and in slow stop and start driving, the ECT did increase, beyond that range, but this was controlled effectively by the radiator fans.
Thank you for the comparison...Yours runs between 186F-194F then and gauge showed no sign of dropping upon deceleration. I cant even get the ECT that high now! That is how my previous setup was before the coolant and tstat change. How old is your tstat and coolant? I wonder if my new tstat is a faulty tstat that isnt closing properly.

Good info on the ECT sensor.
 
#20 ·
My thermostat is factory original. Coolant has been drained and refilled every two or three years. The original factory coolant was the Subaru "green", but the replacement has been the newer Subaru "blue" in the premixed form.

In your first post the replacement thermostat was referred to as OE, but the coolant was specifically "Subaru". Is the thermostat the Subaru supplied one, or an aftermarket that claims to be "original equipment"?

While I hesitate to go that far, I can't discount my tendency to think that, indeed, the new thermostat is on the low side.

Do you still have the old thermostat? If so, is it marked with the operating temperature (usually stamped somewhere)? If the new one is Subaru, I'm wondering if the replacements being supplied now, even for the "older" engines, are set lower.
 
#21 ·
The one that i removed was Factory Original as well. I only changed it as a preventative maintenance since i was draining and flushing coolant. (Did not know you could use the "blue" super coolant without issue in 05-09 vehicles. I would have switched to that had i known darn!) I used the Green Subaru OEM coolant mixed 50/50 with distilled water and the T-stat I got was 21210AA030 OEM Subaru from the dealer and comes with gasket. I only visually looked at the 2 side by side and they were identical but did not look at the stamped numbers on it. Now its deep in a landfill.
 
#22 ·
The blue coolant can be used in 05-09, but it doesn't achieve its long-term replacement benefit unless it's the only coolant. If it's mixed with the green, the coolant still has to be changed more frequently.

Okay on the thermostats.

some more "research":

According to the 2005 FSM spec for the thermostat:

Starting temperature to open: 76 — 80 C (169 — 176F)
Fully opens 91C (196F)

But my 07 spec:

Starting temperature to open 80 — 84°C (176 — 183°F)
Fully opens 95°C (203°F)

According to http://opposedforces.com/parts/, earlier 3rd gen. would use 21210AA030, but later 3rd gen production would use 21210AA080. But the 080 was, apparently used, as far back as 04 October, so some 2005s might have this one.

I also checked OEM Subaru Parts Online | Subaru of America, and it seems to list the 21210AA030 for all when searched by year/model etc. However, when I enter the 080 part number in "search", it does show it as applicable to some 2005s.

I wonder if the different part numbers reflect different opening temps, and your car might have had the higher one, originally. I doubt we can find specs for the two part numbers, but if I have a chance on Monday, I'll call my dealer contact to see if they have both in stock and, if so, if they would check to see if they have different stamped temps. Or, maybe someone following this thread can provide information.
 
#25 ·
Thanks. That's consistent with what I found, although the other two sources suggest a lot of overlap of coverage based on model/trim and production date.

IF we can find that there is a difference, and perhaps that the 080 is indeed the higher temp one, there might be a basis to suggest replacing with the 080, but let's see what we can find.

Do you know the production date of your 2005? (Month/year should be on a label on the front door.) Also, what model/trim level is it?
 
#27 ·
10/04 production date. It's base i model.
That does seem to come within the "crossover" period where the car seems to be listed under both part numbers in some on-line sources.

I searched the two part numbers here and on Google. There's a fair number of hits relating to the operating temperature of the 080 and the consensus seems to be that it opens at 176-183 F (~82 C). (I didn't find any original documentation, so it's possible that one post led to others using the same information.) There's some threads in subaruoutback.org, including http://www.subaruoutback.org/forums...rent-set-temperature-between-4-cyl-6-cyl.html.) It appears that while the 080 opens at ~82C the 030 opens at ~78 C, or 4 C lower (a difference of 7 degrees F, which is also reflected in the FSM temperature specs in post #22 above).

I wonder if the parts data system that the dealers have can better identify the thermostat that was used in the engine based on the VIN.

What a puzzle!
 
#28 ·
I wonder if the parts data system that the dealers have can better identify the thermostat that was used in the engine based on the VIN.

What a puzzle!

And you guys are digging deep. This one, where the OEM data is so close...I would just be like NBD. plainOm, thats a reason for you to find out. I wish I had your brain...but we are definitely wired different.
 
#35 ·
Waiting on a call back. (Might just take some time . . . if not today, tomorrow.)

In the meantime I went back through a number of logs I have where I was on a highway for a few miles, then exited to a stop. In each case I graphed the ECT and throttle or equivalent indicators. There is a visual correlation that suggests the ECT did drop. I didn't see it in the drive I reported on above (not logged), but now it's possible I missed it somehow. I should also note that the logs were made on warmer days (Summer and early Fall) so that introduces a potential differentiating factor.

The logs seem to show a drop of 2 and 3 degrees C. My ECT data is in whole units, so there's some rounding involved, but the trend appears to be there. You had observed a 7 degree F drop, from ~182 to ~175, which converts to about 3 C. So it looks as if the change might not be unusual.

This is hardly a scientific analysis; it's just some observations needing further testing. (More logs is in the works . . . )
 
#36 ·
Waiting on a call back. (Might just take some time . . . if not today, tomorrow.)

The logs seem to show a drop of 2 and 3 degrees C. My ECT data is in whole units, so there's some rounding involved, but the trend appears to be there. You had observed a 7 degree F drop, from ~182 to ~175, which converts to about 3 C. So it looks as if the change might not be unusual.

This is hardly a scientific analysis; it's just some observations needing further testing. (More logs is in the works . . . )
Thank you for the additional data and research. I know i should be satisfied with the temp decreases when decelerating as oppose to the contrary. I am more questioning the factory reasoning behind the non-specified parts used (belt and tstat) as well as the noticeable changes to the cooler side of the dash gauge.

My post this morning about regardless of which tstat is in, once fully warm and open, the gauge should not react that fast to a few degrees of difference. It takes the spring time to open and close longer than the result on the dash. Perplexing!!!
 
#39 ·
Not necessarily, as I see it.

There is a range of coolant temperature over which the temperature gauge on the dash will not change when the coolant temperature changes.



As this diagram shows, if the coolant temperature is around 93 degrees (above the "(3)" on the bottom), a change in temperature to, say, 88 or 98, won't result in a change in the position of the pointer.

If the coolant temperature is running around 86 degrees (just where the line changes from sloped upward to horizontal, then if the coolant temperature goes up, the gauge remains the same, but if the coolant temperature goes down, the gauge pointer will move down as well.

The idea behind this is that the coolant temperature will change as driving conditions change, but as long as it remains, in this case, between roughly 85 and 100 C, the gauge doesn't change. That's considered the "normal" range. It's only if the temperature is outside that range when the gauge pointer will move away from the "normal" position.

What I believe is happening in your case is that the old thermostat was keeping the coolant temperature more-or-less in the area where the gauge response (in the diagram) is flat, whereas with the new thermostat, which we believe operates some 7 degrees F lower, the coolant temperature might be on the top end of the sloped part of the response line.

With either thermostat the ECT sensor still reacts any change in coolant temperature when the engine is in fuel cut, but the dashboard gauge reflects the change only with the new thermostat installed.

(I think we're on the same page here, and it's semantics -- how we describe the responses -- that might differ.)
 

Attachments

#40 ·
Great analysis thanks. A slight correction for the chart you have 158f it should be 185-212f.

So are we to conclude that the 030 and the 080 tstats are both ok to use? Its not that one is wrong or right just that both fit the application? I also hope that running the 080 tstat from the factory until now didnt cause premature unseen damage.
 
#42 ·
Yes, that error was behind the correction in post #4. I think they transposed the 5 and the 8. Interestingly, that diagram is from the 2005 FSM. The corresponding pages in the 2007 and 2008 FSM has the same error!

ditto what traildogck said about the tsat operating temperature. I think that if the 080 -- or whatever it was that was used at the factory -- was problematic, Subaru would have issued a notice of some sort.

I hope to have word on the two tstats soon . . .
 
#41 ·
I really would not worry about any damage. Just for fun information. My car really runs more in the 90-91*c range...I have been watching closely. My oem ECT is in the stock location...the crossover...(it has to be, or the car is not happy) My Innovate sensor is tapped into the small outlet line (to the turbo header tank) from the radiator...it is coolest, coolant in the system...

As I said before...I have seen the OEM ECT 1-2*C hotter than the Innovate, and that makes sense. The coolant hasn't gotten there yet...but they usually run hand in hand the same temp. I have also seen 2*C drop in both sensors a times after a driving event....

Like rolling onto an open highway after waiting in traffic to the on ramp.
 
#43 ·
Sorry, not much further ahead.

The two numbers we have been working with are "kits" including tstat and gasket. The dealer had only the 21200AA072, which according to post #5 in http://www.subaruoutback.org/forums...rent-set-temperature-between-4-cyl-6-cyl.html is part of the 030 kit. In the linked post, the temperature for the 072 is cited as 78 degrees (C), but my contact examined the 072 and said it's stamped 82 (?).

I also searched several on-line Subaru dealer catalogs, and there was no consistency as to which part number (of the two kits) applied to which year/model/engine etc.

Then again, there must be an awful lot of Subarus out there running quite well with either one of the replacement tstats (depending on which catalog was used!). There's a good number of posts here referring to replacing the coolant and tstat, and the concern has only been that an aftermarket part not be used (Post #8 above), but I don't recall anyone raising concern regarding which genuine Subaru part number applies.

That said, the confusion that @avk refers to in the linked post still reigns. I might be in the vicinity of a dealer in the next few days and will try to stop in to see for myself.
 
#44 · (Edited)
Sorry, not much further ahead.

The two numbers we have been working with are "kits" including tstat and gasket. The dealer had only the 21200AA072, which according to post #5 in http://www.subaruoutback.org/forums...rent-set-temperature-between-4-cyl-6-cyl.html is part of the 030 kit. In the linked post, the temperature for the 072 is cited as 78 degrees (C), but my contact examined the 072 and said it's stamped 82 (?).
.

I was in a big city today and stopped at a Dealer that I purchased the car from and they had the following:
21210AA072-- stamped 78C (is t-stat only; no gasket)
21210AA030-- stamped 78C (is t-stat with gasket)
21210AA080-- stamped 82C (is t-stat with gasket)

Using my VIN it pulled up 030 but unfortunately my own ignorance threw the original t-stat in the trash without looking at it so I wont know if it was the cooler or hotter number stamped on it from the factory.
I am going to leave as is unless new I find new information and accept the cooler running temp unless i start seeing a CEL or noticeable mileage changes.
And as stated before, within reason, a cooler running aluminum engine is probably better long term.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top