CaliBenzDriver

MS, Friends and Gents:
here is something worthy of your attention.

> INTRO:
Let's try to fix another sofisticated MB defect that's plaguing my car and 99% likely yours as well.

You know @S-Prihadi and I are proven recipients of the factory ALT YOYO issue that cooks MAIN + AUX with prolonged 1300 Watts of heat (90Amps of 14V) .
This is so basic IT SHOULD THROW A HONEST CEL WARNING.​​​​​​

Today I am going to draw a parallel with the TEMP YOYO that impacts the viscosity of my precious engine oil.

Master Surya has a vibration sensor mounted in his butt and I have the same one in my right foot. Oil temp affects how well my engine can runs.

Unregulated temps affects ECU learning viscosity because operating temp swings from 80 to 120... ouch! I want 103°F (+/- 5 to 10F).

BTW, NO CODE! 👏

You wanna read the story or the conclusion...? How about we start with description then we'll connect some dots and lastly will try to chart targets.

There's a good chance we can cancel this perfect defect with another ¢10 fix... let's see where this leads us.


> W212 YOYO CHAOS:
I've come here to deal with my temp gauge asking for my attention.

Usually the cluster temp gauge shows doctored values to hide the normal sawtooth regulation + the superheat surge when M27x oil pump solenoid is de-activated. Cylinders are then cooled and the coolant circuit temperature surges.

What's happening is my gauge is basically saying my 50kMi thermostat is already wasted. That could very well be since it got heat soaked to crazy temperatures and pressure after engine stops.

We know the ECU is quick at pointing thermostat fault codes because it monitors ECT vs. heater PWM. The ECU is able to coordinate pump solenoid with Tstat heater. That feature seems working all right.

Simplicity escapes when software is used to inject intelligent processing in fan management.
We've seen the car reactions are designed to gently hide dry cylinder temp surges. Great so long we understand how things interact... it is what it is! There are things we can not change and things we can change.

The ECU controls the engine temp not with the fan but with the Tstat opening which depends in part on what the ECT reports mixed with software goals to heat or cool engine + tranny ATF.

So what collection of parts should be replaced to fix these wild temperatures swings:

• TSTAT ?
• ECT ?
• FAN ?
• ECU...?
• SAM...?
• HARNES$$,
• BELT...?
• >> NONE... this is a MB snowball

What we have in sight is this engine being normally hot. The dry cylinder superheat is only part of the abnormal heat control. Granted we can say yes, Bosch wants to run this engine super-hot BUT I have seen a much more satisfactory temp.
Ppl can stay there and enjoy their engine toasting plastics and smoking oil - I am going to bring mine back in line with standard Japanese quality. I've seen what can be done.


> THE BOTTOM LINE:
Sensors values are floating unreferenced!
I believe that for some reason part of the ECU reference voltage loop use unreferenced floating GND that's mismatched with the measurement.

What that does is create an erroneous values temp and voltage that are not referenced to the same commun GND level. This is like trying to measure a voltage with one single probe and no reference: it's a floating value!

This I believe is our opportunity to cancel a whole range of chaos from our chassis.

The question then becomes... what else shares that sensor loop shared by ECT???

The answer to this will help us scope the unknown limits of what's involved.

At the end if the day, we're aiming to reference the floating loop with at resistor or a capacitor or cawl GND screw....


> The funny thing:
It's fun for serious geeks: I noticed my 14.4V drops sinks with temperature. I used to think not much of this because 14.0V is still better than 11.x...

Me is like Surya: quickly satisfied with reliable perfection - LOL

If I drive 45mn or more my BATT voltage ends up near 14.0V (unrelated to 13.7V accelerations)

This is drains while driving if the ALT was in 12.6V float, the skewed ECU/SAM false measurement end up near 12.0V ie. solid batt killer drain by toasted ALT.

Combine that with unstable CAN-C ESP/ISM and you get the famous "X-Mas Tree" slash "limp-mode".

So now you have a better picture about this Mother Of All Bugs. Fixing this is going to help transform the ownership experience of those interested.

Engine + tranny that don't suffer from poor heat control will run with stable viscosity.


> I DUNNO:
Note this was already FACTORY DEFECTIVE as proven by early voltage yoyo at 5k.Miles.

-1- What are other floating sensors turned liars on "5V-Ref." loop ???

-2- What needs to be referenced ???

-3- Is the hidden cawl post the main GND actor ???

-4- I am aiming in the direction of a bad capacitor getting a workout with heat. Now it's getting worse.... what can this be???




Famous Floridian outpost + tropical diver's water


Entering where BS drama kills under 2mn

CaliBenzDriver

-- The temp yoyo is caused by the ECT telling stories to the ECU. This varies based on unwanted external factors.

-- This lack of proper regulation is further obstructed by the ECU firmware working the Tstat and the fan to appear smooth.

-- Regardless what fake temp says, the engine can feel abnormally hot at times for no apparent reason.

​​​​​​This was made more obvious by normal oiling. With limited squirting the powerplant is like a steam engine.

With better 5w-40 oiling heat is reduced a large step to the point temp yoyo becomes exposed.
Normal sawtooth temp regulation steps are seemless. Not +/- 2O°F

Then liar gauge can't help but show yoga exercise up/down... so sensor may be involved to some extent. Perhaps crazy Temp drives the fan firmware logic nuts as well with abnormal conditions.

This is not the ECU trying to rebuild superheat when gauge shows engine is too cold.


+++ SMOOTH TEMP CONTROL +++
When this is fixed wild temps are going to be well under control of ECU + ECT + FAN + TSTAT moving heat together.

+++ SAME BLUEPRINT +++
This refined system suffers from a carefully crafted feature that tanks its performance under the radar.

S-Prihadi

Cali,

Since I removed my LIN connection to the alternator, my YoYo voltage is gone.
This is a 170 minutes logging from basically cold engine start. Jakarta city to Surabaya city my home to 1st fuel stop. Oil solenoid still un-touched.


I believe B-BUS is by OBD port and that means this is from Rear SAM.



EPS means : Me backing out of my home, in reverse, I will need to do FULL LEFT LOCK steering and then from there FULL RIGHT LOCK steering and the straightened the steering while moving forward.
This is the heaviest sudden burden on alternator and battery by Electric Power Steering, as you recall typical alternator response as per their data sheet is 2 seconds.






If I zoom around the EPS activity region, this is what is actually happening.
.



.
Prior to un-plugging alternator LIN, the voltage drop for the same EPS activity is rather extended low voltage.
Battery is providing the deficit of -59 amps for EPS, no choice, alternator power/current rise delay is a fact we can't do anything about it, short of keeping our battery healthy.






.
The ECM YoYo charging strategy via the LIN command is so awful, see below......I look back and I smiled today ear to ear , for disconnecting the LIN.
Below is a 15 July 2021 archive






=========================



THERMAL....... I still am looking for data logger which can log duty cycle of my radiator big fan M4/7 and R48 thermostat heater.
Banks does not make such logger, it can do frequency only.


I do not have the luxury of surplus good cool 20C / 68F or cooler ambient temperature like you guys.
I said surplus is because I am at 30-32C / 86-89F most of the time in daytime, I loose 12C delta to begin with compared to you guys.


The same 170 minutes run for voltages, now we see the thermal.





Cold start and speed at paid highway heading out of town is decent, often 80KM/H / 50MPH or higher I can get, the engine is running cool.




Let's zoom on the coolant temp and oil temp. For me I am seeing 87C / 188F as the ECM lowest temperature target starting from data point 20,000 onwards.
Ignore the first 25% of the run at 15,000 data points, where coolant temp can be as low as 84C.





.
When zoomed out, coolant temp look stable... that is why I like to zoom-in for better data viewing.




All is good when car velocity is 80KM/H / 50MPH or higher when and if the driving style is not too sporty, as in this 170 minute run I was not in the mood to drive too aggressive.
I did not even exceed 95MPH ..... and I did not do 6,000+ RPM often, only twice.

===========

Coming home into Jakarta city..........

Now come the deficit of cooling, where car velocity is too slow..............
Back into Jakarta and entering the traffic jam of the city paid internal highway. I take a bit of the log only.

Ambient actually is 32.xC and not 34C. Once the car slow down under 50MPH, that bumper located ambient air temp sensor is effected by engine bay turbulence hot air.



.
ZOOM-IN

SIDE NOTE : See how the under-hood temperature gets hotter as car move faster at data point 324, while engine coolant temp comes down.
That is heat transfer we are seeing from radiator core to the ambient air, and too bad it blows that hot air back into the engine bay/under hood .... hahahah.
You know some sports cars have extra radiator/cooler which its exhaust hot air does not get pushed back to engine bay.


The ECM target for coolant temp when engine is already hot , is 95C / 203F as per Xentry , but never it is possible in my city traffic jam and in my hot humid climate.
100C / 212F is more common and realistic actual result. But that 170 minutes log at 50MPH or faster, you guys saw 91C is the average I got.
So indeed my cooling system is in mild deficit if slow creeping traffic. Sure, no overheating ever.

Here is why I wanted to know, how much power or % is allocated for the radiator fan by the ECM and if the thermostat heater is at full power to maximum open the mechanical thermostat or not .. ?
This I can't answer yet, until I can log these 2 parameters/devices.

MB fan being shared by engine radiator and HVAC condenser is blessing and a curse at the same time.
I can't log using OBD2 ( but possible with Xentry ), the fan speed percentage requested by HVAC.


Coolant water pump, the mechanical one.
This coolant pump at 650 RPM idling is actually not good enough for its flow for my engine in my climate, wish it can do a tiny bit better.
Radiator fan speed is one thing, but coolant exchange speed is based on coolant pump itself and is super important.

.

CaliBenzDriver

Thank you Surya for this awesome live data logs. I really like looking at reality to see how the software tracks.
We see the ECT vs. speed regulated coarsely.

..../....
Let me look at this in more details.



> Analog vs. digital:
Digital data are born digitally computed by a logic and Analog data are physically acquired from sensor measurement such as voltage, temperature, pressure.

ANALOG:
Voltage varies from 12.5 to 14.9V above GND.
ECT varies from ambiant to around 103°C

DIGITAL:
CPS or CKP are like ON/OFF tics of 5V, 12V or bus logic levels.

When the ECU wants to read the ECT, it reads a logical register linked to an I/O input. Between the chip and the physical sensor there is an interface circuit.

​​​​​​Dedicated circuits are used to prevent capturing unrelated variations. Dedicated line for power reference are both used strictly with ultra low current load. This prevents creating drop-voltage mixed with the input data.
In other words the ECU inputs should be acurate.

What if ECT is mixed with voltage errors? It's hard to tell because there is only one ECT circuit to read a single value.
However I can make an educated guess about what I think the reference voltage is doing... "Houston: we have a GND problem."
​​​​​​

> LIAR FAMILY :
We know the GAUGE is doctored to hide the superheat ups and downs when dry cylinders are sprayed with cool slippery engine oil.
This is a sort of data translation done by
the ECU software logic operates a translation such that when temp stays within a range, the gauge does not move one bit.
The Gauge only moves up/down when the true Temp is outside the regulation min/max window.
ECU tries to keep the engine coolant within the control min/max range using Tstat and fan over engine radiator.
Thing is... the ECT CIRCUIT LIES TO ECU logic. We are working with skewed Temperature data.
The outcome is car ends up being too hot because it's not being cooled enough. ECU is working with temps lower than reality.
I don't know the exact difference 10 to 20°F are misrepresented causing the engine to heat that much more. (I understand MB wants to run these engines hotter... this issue is unrelated).

​​​​​​
> BENEFITS:
Canceling the yoyo fixes everything involved with that sensor loop.
Here I am dealing with the ECT input.
Fixing this issue will deliver cooler ATF + Eng oil temps.

That means more rock solid performance from hydraulic VVT + Mixture adaptation.

The ECU will be able to regulate true temp levels on target instead of roasting engine plastics and seals.


> YOYO FLOATING Voltage + Temp data
-- We only know we are dealing with a voltage yoyo because we have a secondary voltage source of measure that shows us the twisted reality.

-- With the ECT, we can hardly tell temperature is wrong because we have no other source to measure the same coolant temperature. ECU can pull a code on the sensor when it's open/shorted but in between ECU has to trust the single ECT.
(Japanese have two temp sensors to cross check each other. I don't think this chassis features a radiator temp sensor).

.../....

S-Prihadi

I have some sort of back up sensor for Engine coolant temp, but for turbocharger return coolant line.
But the sensor is a thermocouple, so I read ALU pipe outside temp and can't read actual liquid coolant temp. Still it is a good back up.
The title of this sensor I choose what Banks Gauge has in the menu, so I chosen Coolant Temp - Engine Out

The turbocharger gets a cooler coolant, which is not yet cooling cylinder walls. It is sourced from very low side of the engine block.
The ECT by MB surely reads the spent coolant after it cooled the cylinder walls and then the cylinder heads.



TL = Turbo coolant source Left, Bank 2
TR = Turbo coolant souce Right, Bank 1
OUT is to HVAC heater core
ECT = Engine Coolant Temp
CKP = Crankshaft position sensor



.




.
My thermocouple sensing location for the Coolant Temp - Engine Out ( turbocharger spent coolant ) is below :






.......




.
I choose RIGHT side turbo Bank 1, coz it is shortest distance for me to work with. My Banks Gauge sensor modules are at Front SAM and my car is a RHD, so the Front SAM is at the right side of the engine bay.





The data :

Out of Jakarta, 1st leg, that same 170 minutes.




.




========================


.
Back into Jakarta last leg, the whole 248 minutes. See how I suffer at the traffic jam






----------





-------




This way I can track if my ECT by MB goes banana.
So far OK.

The K thermocouple I am using is immune to noise, sort of.......... and since it is basically a stand alone mini galvanic battery, voltage YoYo can't effect it at all.
So it can be a good back up.

Actually, the best back up for ECT, is the metal body of the coolant pressure transducer I installed at the engine block to HVAC heater core coolant hose.
Still I can only read skin temperature only, but it would be more accurate back up to MB installed ECT.
My extra ( 3rd) unit of Bank 4 input thermocouple module will be arriving mid 2024, it has 1 spare channel I can dedicate for ECT back up. 3 more channels are for HVAC components temperature tracking.





================



ECT wiring, 2 wires. It is using ECM internal ground, not sharing to cylinder head ground. This should be "cleaner" than the single wire ECT don't you think?

M276.9 3.5NA




.
M276.8 3.0 Turbo




.

CaliBenzDriver

MS... how about we use the "liar family" for your advantage. Would that be cool in tropical heat?

Currently I suspected the floating ECT circuit is adding heat to our engine. Let's reverse that setup...

How about we use the same trick to tell the ECU engine is 10°C hotter than reality. - This will force greater cooling instead of vaporizing oil.

This will trick the ECU to open Tstat + fan to cool harder.... Possible ?

S-Prihadi

What do you propose ?

CaliBenzDriver

The word is Thank You Very Kindly Master Surya for the schematics with ECU internal GND. This is exactly the heart of this topic.

We need to understand more about this sensor string - I think this is screwing up ALT + ECT +....


Here is what I see our deck Captain:

-- Investigate floating reference to cancel yoyo on whole string of connected sensors. Need to identify the sensors sharing the floating ECU GND.


-- Fix skewed coolant system temps. This should make ECT stop telling wrong data to ECU. Decrease engine heat (15°) back to a true normal range of sawtooth regulation.


-- Artificially Drop Temps a bit : shift ECT measure with serial or parallel resistor according to PTC or NTC. (SW Temp validation?).

The outcome is this chassis temperature is going to get back under stable control, minus yoyo drift.
I could be mistaken... that's exactly what testing is for. A ¢10 detail screwing things on large scale.

What do you think?

++++ I kind of thinking the shared splice is a common +5VRef or 12VRef and is not the ECU floating GND itself. ???

WHERE is this ECU INTERNAL GND available so I can measure then tie it to... a clean single GND.

S-Prihadi

Here is the list for my engine : From my own "W212.065 Booklet"

.
Sensors which are directly powered by ECM 5 volts power supply and not using direct fuse from Front SAM :
The grounds for these sensors are direct ECM and to two different ECM pins.





For me now I know for sure that my cooling system can not do the programmed or target 95C coolant temp at traffic jam.....or it is indeed willing to allow 100C , after all sometimes when
ambient air is cold, it can give me 105C hahaha. In fact the only time I see 110C is when ambient air temp was 20C or so.

So I do not think tricking my ECT to be seemingly hotter will do any good.
The only thing I can think of is , how do I command more fan speed ( assuming there is reserve fan speed available ) during traffic jam ?

I would not call the pin 70 and pin 44 the 2 sensor ground at ECM as a floating ground, it is more of a filtered ground because when I scoped it, it is much cleaner of noise compared
to me grounding my scope at suspension main ground stud.

Since we do not know what is the algo written by the engineers for its MAP controlled thermostat or actual dynamic target seconds by seconds ( not the Xentry temperature target ), we can only guess.


I been showing this quite a few times on this forum, but since 99.9% members do not log their OBD output and rely on the liar instrument cluster gauge,
no one realized or knew that their ECM is an opportunistic bad-azz when it comes to do 105C coolant temp LOL.

One would freak out seeing this, the moment the rain kicks in, coolant temp goes up from 90C to 105C....




.

CaliBenzDriver

thank you Surya for this precise wiring information.

I need to study it to understand the next step: measuring the voltage difference between the floating GND vs. chassis/engine GND.
It should be near zero, yet I believe it varies and that that is trouble.


I don't understand why you don't think tweaking ECT can change your cooling temperature ??


I on the other end don't think bumping the fan speed can help cool the engine out of range.

Here is my reasoning for that...:
if the ECU see temperature drop out of the regulation range, ECU simply close thermostat a bit more to prevent heat from escaping into cool radiator.
So event if you run the fan full speed, it can only cool the radiator alone and can not cool the engine behind a partially shut Tstat.

That's why fan speed is only useful to remove heat above temp max, right ? Your excellent data chart shows 100C is well controlled. (except when you go racing ... then more fan speed: yes because heat is above range. I think an engine oil radiator would help you dump heat out directly bypassing the MB heat exchanger.


The questions around ECT are multiples.
- Right now I believe the coolant measure is based on floating reference. My primary target here is to obtain trusted values without sensor yoyo.

- Even if we trick ECT sensor, there's a chance ECU will sense the error we introduce using another temp sensor (IAT). If error is 10C this may fall within ECU acceptable margins.

Personally I understand the ECT is the basis for the whole engine temp regulation. It drives the ECU target map used to open Tstat and spin radiator Fan to control smooth temps.
If we tweak ECT to report coolant is at 110C, the ECU will manage Tstat + Fan to bring engine coolant temp back down to 100C... in fact 90C for real. This is MAP based so we need to understand what tricks the SW Logic does that's necessary - Definitely using vehicle speed, engine load ,... to adjust fan. This can be useful to fake ECT when it fails as seen recently with Diagnose Dan. I don't think it will prevent us from tweaking.


This whole ECT topic must have neen covered by many others ppl before us on different cars. Engine temp and resistor ECT is now a century old study.

​​​​​+++
I hear the ECT thermistor decrease as heat climbs, so this is an NTC thermistor. To be confirmed....
An I think engine heat increases as GND floats away from true 0VDC.

S-Prihadi

When I have time I will dig deeper too.

If that ECT live at an easy location, I would have taken it out.
Never mind, I will buy a spare one and spare CKP sensor too.... , since the coolant plastic pipe at end of Banks 2 for HVAC heater is so close to the ECT, I might as well replace ECT while I am there
and CKP too. Keep the old ones as spare.

What I saw during my stationary test for the MAP thermostat, at 95C .....the R48 heater is already at maximum power aka ECM wants thermostat to open maximum already.
But I know from fans sound, FAN is not maximum speed yet.
That is why I am keen to mess with the FAN more than the thermostat R48 heater or ECT sensor.

Still the "dynamic-ness" of how the ECM wants the actual coolant temperature is still unknown.
That rain and coolant temp on purpose ECM allows to hit 105C shows unique opportunistic tactic of the ECM,I wonder if Xentry is online at that time, could I see some data changes for temperature target.
So far Xentry does this only :
01 - Cold egine start. Target Temp set at 105C by Xentry.
02- By the time coolant temp hit 80C ish, target changes to 95C as see on Xentry. That is all I seen and I was stationary.

If only such data exist like Camshaft VVT , commanded and actual degree....it will be awesome.

CaliBenzDriver

Okay thank you for the logic. I totally agree we could use better heat control.
It's interesting to read about ECU/PCM dynamic "Temp targets" to manage Tstat opening.

The whole cooling system is pretty fancy, yet the results can be disappointing - SEE AT BOTTOM -

I remember you said your engine got too hot while racing on circuit track. You had to go easy to let it cool down. I don't race but can directly relate to that runaway sort of heat control. Its like when the engine heats up the ECU is unable to keep true heat down.

The Temp numbers look ok but the reality says otherwise. The heat level of my engine varies above what it should be.

​​​​​​We agree that when the radiator has too much heat the fan needs to ramp up to bring Temp back into range. I think the ECU fan logic is biased to be quiet and not speed too fast.

> Thermal runaway:
What if the ECT circuitry makes it report less hot numbers the further above 100C.... that is the excellent recipe to... OVERHEAT!

I think something's wrong when an engine can run hot in mild California's winter when driven easy. Answer found!!!

> NOTE MOD2.1:
-- the engine core heat is actually deregulated by the limited oil squirting.

-- With adequate viscosity and experimental oil pump setup, it is possible to obtain normal temperatures.


..... still need to test the floating GND reference

S-Prihadi

The sad thing about my engine and its cooling system is when driven hard at the track, it will overheat for sure by lap 4, if drive hard enough because by end of lap 3, 120C is guaranteed.
Simply the cooling capacity is not designed for 32-34C ambient temp, driven hard. Which I can accept, it being a family sedan.

My very first own car in 1983 ( courtesy of Daddy ), Supra MK1 was using the oldie viscous clutch type cooling fan as main cooling fan.
This engine was lightly modified 1/4 race camshaft + porting + full street racinge exhaust if for engine section, and I do top speed all the time on it, at a super straight road parallel to Changi Airport runway ( Singapore ), which is 4.9 KM long or 3 miles. The engine never overheated or run say at 95C or higher at all. I guess its cooling is more than generous, being a Japanese car.

Later in life, back in Jakarta, I have a small 1.0 liter Toyota Starlet which I modified a tiny bit too, but not the camshaft, but I did made special camshaft advance sprocket.
This baby can only do 100MPH at redline 4th gear. Since 100MPH is so slow and the paid highway back then in the 90s were so un-crowded
This baby Toyota will overheat when I maintain top speed for 5 minutes, so me got to slow down to 80MPH and when engine get cooler, I do the top speed again.

The 2 above cars are cars I really really abused its engine to red line and sustain the RPM red line.
So I know their cooling system thermal capabilities well.

This MB is the only personal car of mine ( my Ford Fiesta 1.6L too ) I driven at the track for fun but hard, and I would say compared to how I driven my MK1 Supra,
the E400 cooling system is not good enough for my climate when driven hard.
I guess if at a country/city with 20C ambient temp , my 120C would be 110C only and that is acceptable for modern run-hot engine.

There was a member here,also from Jakarta with a CLS400 and same exact engine as mine, I recalled he modified the car using ECM tuning and sure enough, he told me the engine coolant temp is way too hot too fast
when driven hard and NOT AT THE TRACK mind you, while in the city. So me M276.8 owner in my climate, I must swallow the bitter pill.

Of course being older now, I do not drive like a maniac on public road anymore, I only do civilized high speed at paid highway and I do not use emergency stop lane or
enter by brute force into the middle of two cars space ahead of me. Too shameful to do that now.

But I will take corner very hard on curvy back roads, if there is no one around or no houses where when uncle Murphy decided to throw some sand or oil leak on the asphalt from
other cars and make my car loose suddenly its traction and azz skid into some plantation.. LOL.

WIll do some ECM vs Chassis ground test when I have more time yah Cali.

ASK:
How hot does your engine get ( using OBD ) ?
What ambient temp were you to get to that "X" temperature ?

CaliBenzDriver

Oh yeah, the Supra was a ton of fun. Great technology. A friend of mine had one. Strong and smooth six-in-line with stick shift clutch. Engine was gently controlled.

I had another friend with a Ferrari 308-GT: light V8 with double barrel carbs and double ignition points too... wild horses always ready to rock. The best were the suspensions with 4 control-arms instead of 2 winsbones. A huge wheel bearing and tiny tight bushings.

Cars are fun but they can't rival with the raw power of JPN motorcycles. The open roads are fairly dangerous with all sorts of ways to alter life. So now I've learned preparedness of defensive driving.


>>>HEAT MGT:
Surya, you coined the perfect term of "Liar Gauge"... actually it's based on the "Liar ECT" -
​​​​​​
The large amount of heat that is vaporizing oil is stuck inside the engine core, never measured per say.

By the time the ECU is done doctoring everything the result is USELESS TEMP. DATA !!

I would say the "pump solenoid" is part of the engine cooling. The ECT CAN NOT BE ACCURATE when the coolant is not heated by the oil that's not circulated... LOL


core temperature around 400F

- The coolant temp is meaningless. It's can be at a nice 95C while the engine core is over 400F.

Normal engines don't have oil coolers and they keep the engine cool. Ours is the opposite: it has all the features including limited cooling to save-gas.


Knowing the engine oil temp is better than ECT but if oil is not picking up heat from pistons... oil temp is hardly related to core temp either.

I have looked at my engine heat through scanner but it did not help me understand the massive heat that was stored. I've just realized why, yesterday.

ASK: could it be the oil seals on the timing cover droping significant oil to make pressure worse??

S-Prihadi

You know while discussing oil temperature.

The M157 oil temperature + oil level sensor is a nice one, it surely reads "spent" hot oil temperature, before being cooled down by the oil cooler, because it is located at the sump/pan.

On my 1.0L Toyota, I use VDO dipstick style oil temperature sensor and since that cheapo small engine has no oil cooler, and I am reading "spent" oil temperature, it is hotter than coolant temp by
a decent margin.

This VDO. My God, they still make it today ? Wow.
https://www.ebay.com/itm/171274182153

CaliBenzDriver

How about inserting a thermocouple probe wire down the dipstick tube? This will give you direct measures of oil activities to some degree.

Using the simple "MOD1.0", the temperature control you're missing is Piston cooling below 1850.RPM ie. the "MOD2.1". Normal driving RPM benefits from being cooled.

S-Prihadi

Yes, that is possible with some good DIY plug for the oil dipstick tube, so that no external air enter crankcase.
Must custom cut the thermocouple wire length

CaliBenzDriver

Master Surya, this is going to be

I cant wait to see your oil temp curve in relation with coolant temp, (thermostat opening + fan speed).

This will let you survey the whole cooling system easily.

> IDEA: how about you use a $5 bargain dipstick, drill a hole through top plastic to attach wire without air leak.
Easy to use custom oil-temp probe.


> ENGINE TEMP:
This is going to give you an oil temp directly in relation with coolant., right?

The main heat source for the oil are the.... sizzling pistons.

How do you think we can measure or see at what RPM the individual piston squirters open up ??


> Oil Pressure LEAK...?
Remember the timing cover missing seals we looked at 6Mo ago... I wonder if this is dumping a significant pressure to push up RPM necessary to spray cool dry pistons.


> VISCOSITY'S cool:
I am pretty happy we found an simple way to normalize extreme heat by cheating 5w40 viscosity.

S-Prihadi

There is no chinese made oil dipstick in my market for M276, as no one needs dipstick yet....hahahaah.
I seen W124 oldie dipstick on the market but I do not know its plug OD and actual length.
Also I can no more import from AliExpress China, as my government has now banned them since early last year...DUGGHH !!!!

Cali wrote :
I cant wait to see your oil temp curve in relation with coolant temp, (thermostat opening + fan speed).
These two above are not possible to be acquired with Banks gauge, as above is Xentry territory and I can't run both Banks and Xentry on the same OBD2 port, it will crash.
I must video record both Xentry screen (easy) and video record the inside oil sump K thermocouple with its stand alone display, do able.
The problem with video record is I can't graph the data in CSV/Excel, so it is very time consuming to study them.

I must do hard-wire for duty cycle sensing on the fan PWM signal and probing its wire while doing long test drive is not easy, the probing pins will fall off.
I have spare fan connector, but I do not have the female version of it, I only have the male, so I can't do robust test connection like doing/testing other sensors on engine
where I have both the male and female connectors as test kit wire set for the 2 or 3 pins Hirschmann connector with extra 4mm safety banana jack sensing connection I prepared.

Hence Xentry screen is the best for thermostat opening + fan speed)
Even then we only can assume that when R48 heater is activated at full power, thermostat fully open....assume only....... but we can't confirm it as there is
no flow meter between thermostat and radiator

Will figure it out, no worry. wont be so soon though.

BTW, the LQ Molygen is available now in Indonesia, but the 5W-40 version is not, only 5W-30. It still carry API-SN only.
When I have test enough the new Mobil 1 API-SP and when Molygen get its API-SP and the 5W-40 available, I will get it and do some test too.

Green color oil ? Dang....

CaliBenzDriver

Everything's amazing right now...
COOLANT is yellow, turquoise or pink
OIL is green, black or amber
ATF is blue

It's going to be harder to forecast the mayo mixed color ....

S-Prihadi

Stationary test only. K thermocouple inside oil pan. I extended the length of the thermocouple insertion, hoping it will stay submerged in oil at all times.
Oil solenoid is defeated.

Set up................



.

That white thingy is 3M electrical tape, to prevent air leak into oil dipstick tube.




Time code zero start is the same as engine start.
8 minutes I needed to get 88.5C ( stationary car ) with the last 4 minutes I rev higher and higher and keep at 4,200 RPM to heat up coolant faster.
DO NOTE : R48 heater is opposite value of actual, so 100% means R48 heater is NEVER powered at all by ECM


ABOVE : ECM set coolant target at 105C, when and if coolant temp has not exceed or touch 90C.


BELOW : ECM set coolant target now at 90C, when coolant temp is already at 90C. R48 heater is still OFF.
HVAC is running on my car ( too hot without it ) , also I wanted to load the engine more.
Fan power set at 55% by ECM, when fan power this high, HVAC request for fan would be like nil, only 1%


.


.
Albeit ECM coolant target is 90C, the mechanical thermostat needs 95C to start opening.
R48 thermostat heater is now at 22% power as coolant temp climbing to 93C.
Fan is now reduced to 43%, from 55%
Here you can see that ECM traded fan speed for what it calculated as slightly more coolant flow/exchange volume to radiator with heater at 22%.





.I never knew 80C could be coolant temp target by ECU ,. Now R48 thermostat heater is full power while coolant climbed up to 93.75C.
ECM is still slowing down the fan now still at 43%,





.
As coolant climbed up to 100.5C, fan speed increased to 50% and R48 heater is full power meaning full open thermostat assist.




.
ECM here realized how dumb azz it is to slow fan down even though full heater power to thermostat ....to make thermostat open fully.
Coolant now at 105C and ECM has no choice but to boost fan speed to 74% as coolant temp is too hot now.
Here lies the weakness of water flow at idle RPM for coolant circulation pump. Under 1,000 RPM is too slow.
ECM engine coolant target still remained at 80C.





Here I am trying to use 1,500 to 2,000 RPM to speed up coolant circulation pump.
Below image is 100 seconds later than above image.






This could be a glitch, or the thermocouple end tip was in air and later in oil ( hence hotter), due to me revving the engine.



.




Video for this sudden rise to 95C, but the cooling down is actually nice and slow. So likely the thermocouple end tip was moved into air and back into oil again.
.


.
End of test, 22ish minutes


Holy cow, my ECM is at 72C internally. Bottom data field.




WILL CONTINUE...................

S-Prihadi

This is the Omega/Banks oil temperature K thermocouple at the skin of the Banks oil pressure sensor.
Hence I can log it with Bank iDash gauge system.

ITALIAN TUNE UP
Traffic quite high at the highway, hence I can't hit 100MPH+ / 160KM+


.




The details
Zone 1, Shell fuel station to paid highway internal to city.
Zone 2, paid highway leg 1
Zone 3 out of highway and back into highway leg 2, heading home.
Zone 4, out of highway and to home



Zone 4 - Slowing down after Italian tune up into slow traffic will always cause coolant temp to go up.
There is not enough wind velocity to cool the engine
That is why engine oil temperature also goes up as it is using the "hotter" coolant circuit to cool oil down.


==========================



Super civilized driving with Wife, RPM highest 3,500.
From Mall to my home. See how slow is my best speed , 40MPH max ? Not using paid highway, normal freebie road.





I think overall the oil after being cooled at oil cooler vs spent oil in the pan, the temperature difference is not that great for spent oil in pan.
Maybe 5-7C extra hotter only when I drive it Italian Tune Up style.

If anyone having log their M157 engine oil temp vs coolant temp when driven hard, it would be a great database to share with us.

CaliBenzDriver

Thank you MS for capturing new oil temp input combined with Xentry controls.

We see the Bosch firmware uses "target strategies"... not a basic ON/OFF relay with multiple temp sensors.

ECU has one direct coolant sensor. Our engines are missing oil pressure and oil temps and I think oil level as well.
The programmed strategy makes up for what ECU does not know, right...?!
The coing system is resting on a slow water pump that saves a lot of energy by lack of efficiency.
Idea: replace water pump drive pulley with a smaller diameter to spin it faster - Every savy tuner wants to under-revs all belt accessories.

Our engines have reached that limit where minimal circulation impacts control efficiency :

• Minimal oil pressure circulation
• Minimal coolant pump circulation
• Mimimal electric load for smart fan

We see the cooling system appears fairly accurate(*) at moving heat around to keeping temps under control. We are familiar with hysteresis used to keep control loop captured.
(*): besides possible yoyo floating voltage reference - TBD.

Here is some good news... your data is "all perfectly good"

It's hard go be excited about "hunk of steel" engine systems but we are always happy to swap bugs for better performance!!

Here:

reality is not glitched sensor

You clearly show that by reving-up your engine RPM high enough, the oil spray open up to cool hot pistons.

Below that magical RPM, the pistons are still kept dry regardless of active MOD1. Removing piston heat requires MOD2.1!


-- Engine pistons without squirting at driving speed are smoking HOT & DRY by design... it works!
-- Special approved oil package is required to allow operations with limited cylinders oiling.
-- Limited cooling, smoking oil and high blow-by are associated.

We didn't mention "VVT + Tensioner" because they are not related to cooling yet benefit from normal MOD2.1 oiling.

​​​​​​Now onto Post II...

CaliBenzDriver

MS it's always a pleasure to dive with you for the deep intelligence you invariably supply.

Fresh live data is invaluable like fresh live seafood on the beach!

thank you very kindly



With everything afront of our eyes, many things to see... (following PIC A + B + C )


PIC-A : complete cooling system
This let's us appreciate how speed, RPM and pressure are linked to both oil/Coolt temperatures.

The interesting parts are the interactions of coolant with oil and RPM.


PIC-B: Test drive explains
Within clear context we can understand what test data represent.

> The MEAT:

PIC-B: multiple sections 1234
This is Italian spirited turbo driving with plenty of RPM to mobilize oil through squirters + exchanger.


warm up phase
the oil warms up slowly using
it's not able to bring any heat out


hard drive condition
Given enough RPM
Oil + Coolant track each other well
Working squirters are spraying working pistons
the engine oil heat-exchanger cools the oil well


traffic-jam crawl
-- Suddenly RPM drops near idle after Italian style
-- Now we see oil temp dropping
-- Coolant temp shoots up by itself
-- Coolant is bringing more heat than radiator can handle at once.
-- This section has the same heat load than prior but without higher RPM. Heat shoots up!!



PIC-C: normal driving with Mama on-board... not Italian
-- We clearly see coolant keeps warming up oil below it's temperature: "oil warmer"!
-- Meaning oil never brings any heat into coolant.
Coolant with slow pump is forced to move 100% of the engine heat to hot radiator.
-- Not very promising limited cooling.
-- We see coolant Temp hits a hard limit at 100C with fan speed helping well.
-- The ECU does not have ALL data sensors to measure everything so we use substitutes.
-- There's a big "ELEPHANT IN THE ROOM" we can not see but have come to appreciate: high TRAPPED HEAT.
The graphs shows how everything is working well right?
By design the engine uses an advanced Euro oil formula for cylinders coated with new metals that withstand extreme conditions. This is proven to work besides undesirable heat, lag, blow-by.





Overall, here we are looking at a PARADOX - This thread topic. In prior post we saw how ECU has plenty of ways to cap temperatures and here we see Temps keep climbing up as if out of hand.
We would expect cold temps to climb to a regulated plateau around 90C, between 85 to 100C.

This is what forced us into deeper diving. With data there's no if and buts scenarios: it's the true really.
..../....

> Part III - Simple With vs. Without :
Now we have excellent baseline, let's
Test drive MOD2.1 solution to get heat + tork....

1. ​​​​Engine esponsive heat removal by oil
2. ​​​​​​The ultimate "TT-TTT" : TwinTurbo - Tiger Tank Tork ".

The MOD2.1 overall graph will show :
much quicker oil warming
followed by both temps tracking around 95C
minor increase in traffic jam without spike
Engine will not pack hidden trapped heat.


overall will be smooth across with limited spikes

Shortcut to skip from MOD1 to MOD2.1: I have a tweaked experimental formula when you're ready.
Engine needs sufficient oil viscosity to build 20PSI required to open squirters at driving RPM below 1500.RPM. Squirting above 2000.RPM traps extreme pistons heat.

There is nothing we can unplug after MOD1. to squirt below 2250.RPM... viscosity booster-shot takes care of that.

Now I have to go bang my new lower balljoint.

S-Prihadi

Too bad no similar engine to compare with, if possible one at 32C ambient temperature.

I think you should start installing Banks Gauge Cali , oil pressure + oil temperature at sensor skin, so we can compare NA to Turbo