912+STi v8 Spec-C motor
-
- Posts: 384
- Joined: Tue Oct 02, 2012 5:06 pm
Re: 912+STi v8 Spec-C motor
You should be able to check resistance to ground on that sensor @ a given temp. I don't have the suby specs though. Although, if they both read the same, probably ok.
Might be worth isolating the wire going back to the ecu and checking continuity through the wire as well as to ground. If you see anything questionable, might be worth the trouble to run a temporary jumper between the sensor and ecu to see if that changes anything. Check the connectors too. A little resistance variation will throw readings.
All that might troubleshoot the discrepancy between the two sensors but doesn't explain the elevated temps if the thermo is working correctly, unless it's just that the fans are not kicking on soon enough.
Can you get a decent IR reading on the engine block @ the ecu sensor? It might not be a sensor probem at all
Hope I'm not making this worse by thinking out loud.
No plan survives contact with the enemy.
- RonW
- Posts: 544
- Joined: Tue Jan 03, 2006 8:31 am
Re: 912+STi v8 Spec-C motor
not at all, thats exactly where my suspicions are as well. last night i took the original sensor and checked resistance. i found a guy on nasioc who quoted:
20*C 2000-3000 ohms
50*C 700-1000 ohms
80*C 300-400 ohms
Result: mine fell within those ranges
Wiring: i looked up my wiring last night too and will be tracing it back to ecu (on mine is B135-18 signal and B135-19 ground)
IR@ecu sensor: its hard to get that reading cuz its barried in there but i was close to what the ecu was saying. i'm not saying i accurately got it but seemed similar. it will be worth repeating
thanks for the assistance as always
EDIT: ran out to car in AM to compare CTS signal at ecu vs IAT and they were close
55* vs 57* respectively so it seems to read ok at low temp.
20*C 2000-3000 ohms
50*C 700-1000 ohms
80*C 300-400 ohms
Result: mine fell within those ranges
Wiring: i looked up my wiring last night too and will be tracing it back to ecu (on mine is B135-18 signal and B135-19 ground)
IR@ecu sensor: its hard to get that reading cuz its barried in there but i was close to what the ecu was saying. i'm not saying i accurately got it but seemed similar. it will be worth repeating
thanks for the assistance as always
EDIT: ran out to car in AM to compare CTS signal at ecu vs IAT and they were close
55* vs 57* respectively so it seems to read ok at low temp.
912R + STi 2.0 Turbo JDM v8
- GS guy
- Posts: 909
- Joined: Wed Nov 06, 2002 12:01 am
Re: 912+STi v8 Spec-C motor
Hey Ron,
I don't recall what ECU you're running, but is it possible a scaling factor for the sensor is off? I know for MS you can adjust the readings for whatever temperatutre sensor you are running, so it can compensate for for the sensor that physically fits the engine. If the sensor seems to be operating OK (from your ohm meter readings), then it sounds like the ECU is not compensating for it correctly.
Otherwise I can't comprehend how the radiator can run hotter than the engine - especially with the Subaru sensor reading the temps after the water has passed through the block and heads, just before it heads back to the radiator!
Your sensor numbers match my records more or less. When I checked my Suby sensor (JECS 0720 A27-000-E13) here's what I got:
0C = 6030 ohms
21C = 2511 ohms
100C = 194 ohms
I don't recall what ECU you're running, but is it possible a scaling factor for the sensor is off? I know for MS you can adjust the readings for whatever temperatutre sensor you are running, so it can compensate for for the sensor that physically fits the engine. If the sensor seems to be operating OK (from your ohm meter readings), then it sounds like the ECU is not compensating for it correctly.
Otherwise I can't comprehend how the radiator can run hotter than the engine - especially with the Subaru sensor reading the temps after the water has passed through the block and heads, just before it heads back to the radiator!
Your sensor numbers match my records more or less. When I checked my Suby sensor (JECS 0720 A27-000-E13) here's what I got:
0C = 6030 ohms
21C = 2511 ohms
100C = 194 ohms
-
- Posts: 384
- Joined: Tue Oct 02, 2012 5:06 pm
Re: 912+STi v8 Spec-C motor
That's the part I can't get my head around. Seemingly confirmed with the hand-held IR too. I use the IRs alot and trust them, they are pretty acurate and reliable. The only thing that I have seen mess them up consistantly is trying to take a measurement on a reflective surface. A shot of flat black paint usually cures it, but it doesn't look like you were reading from any shiny spots.GS guy wrote:Otherwise I can't comprehend how the radiator can run hotter than the engine
I'm not familiar with the turbo cooling path. From the bleed port off of the radiator, what is the rest of the circuit? Any possibility of a backflow situation from a hot spot?
baffled.
No plan survives contact with the enemy.
- RonW
- Posts: 544
- Joined: Tue Jan 03, 2006 8:31 am
Re: 912+STi v8 Spec-C motor
SUMMARY ....EDIT: skip to end of this post
Problem:
Car has suddenly started having cooling issues.
Background:
running same radiator set up, motor config, tune and sensors for several months without issue. On Saturday autocrossed and on Monday drove to work and noted the car starting to overheat per the radiator coolant sensor (elevating to 225*F) and the fans which are normally triggered by the ecu not turning on which they have been since last year. ECU fan threshold was set to 195* but no fan. for the drive home, i just jumpered the fan relay and manually turned it on and it runs fine that way.
Testing Behavior:
At home I basically bring it to operating temp and monitor ECU/motor and watch it rise, see coolant temp sensor at radiator also rise. ECU however would stop around 190* sub fan triggering threshold while EVERYWHERE else would keep climbing (ie, radiator cts, IR readings on tubing leaving crossover tube, IR tube leaving radiator and, even right before entering T stat housing!!! all being 200-210* and just gets worse if i drive it.
Hypotheses:
1) air pocket in system
2) stuck or malfunctioning T stat
3) erroneous radiator cts reading
4) ecu tune corrupted
5) erroneous motor cts reading
Methodology/Results:
1) air pocket in system: purged multiple times with front end up 25-30* and run and then the rear end up 25-30* and run. same behavior
2) stuck T stat: pulled and boiled but its spot on 190*. trial running without Tstat, same behavior.
3) erroneous radiator CTS reading: IR reading at multiple locations confirm similar readings of the CTS sensor to the digital autometer gauge.
4) ecu tune corrupted: easy to test, reflash it. same outcome, reflash with much lower thresholds (180* from 195*) and it runs as it use to before this issue ie fans turn one when radiator cts at 195*. as if the ecu were reading the incorrectly. no where in ecu to change/check resistance values for calibration.
5) erroneous motor CTS reading: on morning reading (bone cold) per ecu, ecu CTS 55 while IAT 57 so its reading correctly at low temps. swapped in a spare CTS and same behaviors at low temp and as described above. check resistances with my Fluke and per nasiac 20*C 2000-3000 ohms, 50*C 700-1000 ohms, 80*C 300-400 ohms and it reads correctly. INTERESTING TO NOTE: to the best of my ability, the IR reading at the motor CTS IS accurate!!!
Discussion:
the readings each CTS is giving seems to be accurate, water appears to be circulating as is suppose to. motor operates as it previously did if i just significantly lower the fan triggering thresholds. How is it possible to have the output of coolant from the crossover tube be high and remain high thru the cooling system (no fan on) and return to the motor but then the CTS at the motor is LOWER???
LIGHTBULB!!!!!
The only thing that has changed is that I mounted the WAIC radiator above the alternator. If has an 8" SPAL fan that draws thru the rad. It outputs the relatively cool air since its not on boost onto the alternator. whats behind the alternator... the crossover tube with the motor CTS. I'll bet the fan is cooling it and it never reads high enough to trigger the fans. and when did i install this? days before the recent autocross.
i dont want to get my hopes up but the IR reading seem to be match what the ecu is reporting and that is the ONLY thing i can think of but it makes sense.
Problem:
Car has suddenly started having cooling issues.
Background:
running same radiator set up, motor config, tune and sensors for several months without issue. On Saturday autocrossed and on Monday drove to work and noted the car starting to overheat per the radiator coolant sensor (elevating to 225*F) and the fans which are normally triggered by the ecu not turning on which they have been since last year. ECU fan threshold was set to 195* but no fan. for the drive home, i just jumpered the fan relay and manually turned it on and it runs fine that way.
Testing Behavior:
At home I basically bring it to operating temp and monitor ECU/motor and watch it rise, see coolant temp sensor at radiator also rise. ECU however would stop around 190* sub fan triggering threshold while EVERYWHERE else would keep climbing (ie, radiator cts, IR readings on tubing leaving crossover tube, IR tube leaving radiator and, even right before entering T stat housing!!! all being 200-210* and just gets worse if i drive it.
Hypotheses:
1) air pocket in system
2) stuck or malfunctioning T stat
3) erroneous radiator cts reading
4) ecu tune corrupted
5) erroneous motor cts reading
Methodology/Results:
1) air pocket in system: purged multiple times with front end up 25-30* and run and then the rear end up 25-30* and run. same behavior
2) stuck T stat: pulled and boiled but its spot on 190*. trial running without Tstat, same behavior.
3) erroneous radiator CTS reading: IR reading at multiple locations confirm similar readings of the CTS sensor to the digital autometer gauge.
4) ecu tune corrupted: easy to test, reflash it. same outcome, reflash with much lower thresholds (180* from 195*) and it runs as it use to before this issue ie fans turn one when radiator cts at 195*. as if the ecu were reading the incorrectly. no where in ecu to change/check resistance values for calibration.
5) erroneous motor CTS reading: on morning reading (bone cold) per ecu, ecu CTS 55 while IAT 57 so its reading correctly at low temps. swapped in a spare CTS and same behaviors at low temp and as described above. check resistances with my Fluke and per nasiac 20*C 2000-3000 ohms, 50*C 700-1000 ohms, 80*C 300-400 ohms and it reads correctly. INTERESTING TO NOTE: to the best of my ability, the IR reading at the motor CTS IS accurate!!!
Discussion:
the readings each CTS is giving seems to be accurate, water appears to be circulating as is suppose to. motor operates as it previously did if i just significantly lower the fan triggering thresholds. How is it possible to have the output of coolant from the crossover tube be high and remain high thru the cooling system (no fan on) and return to the motor but then the CTS at the motor is LOWER???
LIGHTBULB!!!!!
The only thing that has changed is that I mounted the WAIC radiator above the alternator. If has an 8" SPAL fan that draws thru the rad. It outputs the relatively cool air since its not on boost onto the alternator. whats behind the alternator... the crossover tube with the motor CTS. I'll bet the fan is cooling it and it never reads high enough to trigger the fans. and when did i install this? days before the recent autocross.
i dont want to get my hopes up but the IR reading seem to be match what the ecu is reporting and that is the ONLY thing i can think of but it makes sense.
912R + STi 2.0 Turbo JDM v8
- Piledriver
- Moderator
- Posts: 22518
- Joined: Sat Feb 16, 2002 12:01 am
Re: 912+STi v8 Spec-C motor
Move your CLT sensor to the hottest point in the system, rather than return side?
Addendum to Newtons first law:
zero vehicles on jackstands, square gets a fresh 090 and 1911, cabby gets a blower.
EZ3.6 Vanagon after that.(mounted, needs everything finished) then Creamsicle.
zero vehicles on jackstands, square gets a fresh 090 and 1911, cabby gets a blower.
EZ3.6 Vanagon after that.(mounted, needs everything finished) then Creamsicle.
- RonW
- Posts: 544
- Joined: Tue Jan 03, 2006 8:31 am
Re: 912+STi v8 Spec-C motor
I think i'll first test with that waic fan off heh. then if that is it, shroud it from the fans output.
912R + STi 2.0 Turbo JDM v8
-
- Posts: 384
- Joined: Tue Oct 02, 2012 5:06 pm
Re: 912+STi v8 Spec-C motor
That sounds like a BINGO moment. Explains everything very nicely.
I was reading through your post to see if anything else was changed about the time the heat issue popped up. Was under the impression that the waic rad was elsewhere (bag of ice between ax runs).
Hard to imagine the fan would be able do that, until you think about it... In the stock suby config, the rad fans exhaust pretty much in that direction...BUT...it is always warm/hot air, and there is a little distance. The ic rad fan running continuously, under little or no boost(cold-ish air) and blowing "right there" sounds like it could very well be the culprit.
Any thoughts of thermostaticaly controlling the ic rad fan?
Interested to see your test results.
I was reading through your post to see if anything else was changed about the time the heat issue popped up. Was under the impression that the waic rad was elsewhere (bag of ice between ax runs).
Hard to imagine the fan would be able do that, until you think about it... In the stock suby config, the rad fans exhaust pretty much in that direction...BUT...it is always warm/hot air, and there is a little distance. The ic rad fan running continuously, under little or no boost(cold-ish air) and blowing "right there" sounds like it could very well be the culprit.
Any thoughts of thermostaticaly controlling the ic rad fan?
Interested to see your test results.
No plan survives contact with the enemy.
- RonW
- Posts: 544
- Joined: Tue Jan 03, 2006 8:31 am
Re: 912+STi v8 Spec-C motor
GSguy,
thanks for checking values for me, much appreciated. i am running stock jdm sti ecu which is programmable but no where to change the sensor values.
883
ya, it was baffling how it could be so. still may be wrong but im hopeful.
forgot, i snapped a few quick pics of it the night i was installing it and had posted it in roach thread per his request but never linked it here. this is its position, dumping cold air right there onto the crossover tube.
i dont know if i'll thermostatically trigger it, i like being able to do it manually when the car is off and i can keep it running with a bag of ice resting on the top of it between runs. if this is truely the issue, i'll shroud it from the fans air flow.
thanks for checking values for me, much appreciated. i am running stock jdm sti ecu which is programmable but no where to change the sensor values.
883
ya, it was baffling how it could be so. still may be wrong but im hopeful.
forgot, i snapped a few quick pics of it the night i was installing it and had posted it in roach thread per his request but never linked it here. this is its position, dumping cold air right there onto the crossover tube.
i dont know if i'll thermostatically trigger it, i like being able to do it manually when the car is off and i can keep it running with a bag of ice resting on the top of it between runs. if this is truely the issue, i'll shroud it from the fans air flow.
912R + STi 2.0 Turbo JDM v8
- RonW
- Posts: 544
- Joined: Tue Jan 03, 2006 8:31 am
Re: 912+STi v8 Spec-C motor
update:
not it
got a brand new sensor and tested to compare to the old ones and they are spot on.
ecu testing next:
background: ecu takes cts resistance and converts to temp via scaling.
Methods: if we apply a known fixed resistance, it should read the corresponding temp
(copied from a post on another forum)
Haynes:
Temp in °F Resistance (ohms)
212 ----------176
194 ----------240
176 ----------332
158 ----------458
140 ----------668
112 ----------972
122 ----------1182
104 ----------1458
95 -----------1800
86 -----------2238
76 -----------2795
68 -----------3520
58 -----------4450
50 -----------5670
40 -----------7280
32 -----------9420
i'll get back w results
not it
got a brand new sensor and tested to compare to the old ones and they are spot on.
ecu testing next:
background: ecu takes cts resistance and converts to temp via scaling.
Methods: if we apply a known fixed resistance, it should read the corresponding temp
(copied from a post on another forum)
Haynes:
Temp in °F Resistance (ohms)
212 ----------176
194 ----------240
176 ----------332
158 ----------458
140 ----------668
112 ----------972
122 ----------1182
104 ----------1458
95 -----------1800
86 -----------2238
76 -----------2795
68 -----------3520
58 -----------4450
50 -----------5670
40 -----------7280
32 -----------9420
i'll get back w results
912R + STi 2.0 Turbo JDM v8
- RonW
- Posts: 544
- Joined: Tue Jan 03, 2006 8:31 am
Re: 912+STi v8 Spec-C motor
Nothing wrong with ecu/wiring either, which is good news.
Got a 0-5K ohm potentiometer. set it to 176ohms, hooked it to plug, looked at ecu and low and behold 212, just like the manual says it should be to the DEGREE!
going thru the full sweep, the ecu converts just fine and triggers the fan at the appropriate thresh hold temperature
So, sensor is operating properly, ecu is reading and functioning properly. It must be an air pocket in the crossover tube or something. pita!
Got a 0-5K ohm potentiometer. set it to 176ohms, hooked it to plug, looked at ecu and low and behold 212, just like the manual says it should be to the DEGREE!
going thru the full sweep, the ecu converts just fine and triggers the fan at the appropriate thresh hold temperature
So, sensor is operating properly, ecu is reading and functioning properly. It must be an air pocket in the crossover tube or something. pita!
912R + STi 2.0 Turbo JDM v8
- Piledriver
- Moderator
- Posts: 22518
- Joined: Sat Feb 16, 2002 12:01 am
Re: 912+STi v8 Spec-C motor
Move the sensor.
Addendum to Newtons first law:
zero vehicles on jackstands, square gets a fresh 090 and 1911, cabby gets a blower.
EZ3.6 Vanagon after that.(mounted, needs everything finished) then Creamsicle.
zero vehicles on jackstands, square gets a fresh 090 and 1911, cabby gets a blower.
EZ3.6 Vanagon after that.(mounted, needs everything finished) then Creamsicle.
- RonW
- Posts: 544
- Joined: Tue Jan 03, 2006 8:31 am
Re: 912+STi v8 Spec-C motor
pile, you are right, for my set up, this cts position is not optimal, it is basically the high spot so if there is any air, it is essentially trapped at the crossover pipe.
i was looking at the motor a moment ago and was trying to figure out what changed. then i started thinking, i think i rotated the 90* hose coming out of the crossover tube to point down so it was out of the way when i hooked the ground wire up for the waic rad fan and the pump. if there was any air in there it just went up to the crossover pipe,
prior to that i had that hose higher than the crossover tube, so the crossover had to be full for coolant to exit.
and there ya go, operating as she id in the past and all it took was a turn of the screwdriver on the hose clamp, rotate and re tighten. man do i want to just delete the above posts and go hide in a corner but hell, maybe there is someone who will learn from my embarrassing oversite.
i was looking at the motor a moment ago and was trying to figure out what changed. then i started thinking, i think i rotated the 90* hose coming out of the crossover tube to point down so it was out of the way when i hooked the ground wire up for the waic rad fan and the pump. if there was any air in there it just went up to the crossover pipe,
prior to that i had that hose higher than the crossover tube, so the crossover had to be full for coolant to exit.
and there ya go, operating as she id in the past and all it took was a turn of the screwdriver on the hose clamp, rotate and re tighten. man do i want to just delete the above posts and go hide in a corner but hell, maybe there is someone who will learn from my embarrassing oversite.
912R + STi 2.0 Turbo JDM v8
-
- Posts: 384
- Joined: Tue Oct 02, 2012 5:06 pm
Re: 912+STi v8 Spec-C motor
Always the simple things that make you pull your hair out...who'd a thought?
Good find, glad you're back to normal.
But dangit, I really liked the fan idea.
Good find, glad you're back to normal.
But dangit, I really liked the fan idea.
No plan survives contact with the enemy.
-
- Posts: 19
- Joined: Sun Mar 19, 2006 9:34 am
Re: 912+STi v8 Spec-C motor
Don't feel bad you taking One For the team. . I enjoy Reading all of your post .
Last edited by jokswgn on Sat Mar 30, 2013 9:48 am, edited 1 time in total.