When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
Here's something cool for your excellent M276-NA engines out there.
The stock coolant circulation through bank2 is extremely poor. It makes use of an electric pump that is not activated.
AUX pump is disabled programmatically by being set at 302°F.
Here is my clean free 5mn experimental solution to enable effective heat management.
I have been researching the wide topic of what's impacting my engine heat since new. I have made great strides to normalize heat with MOD-X effective oiling.
From that stand point of near-normal I could tell REGULATION was sub-optimal. A little bit of up/down heat, not smooth control... I could get good heat and other times not.
I wondered how can a modern car with so much sofistication cool so ineffectively??
Bank2 proper heat removal is factory disabled !!!
This explains why liar-gauge shows a drop when tstat opens up after getting on the Hwy.
The single ECU coolant sensor does not sense circulated coolant - Sensor lies to computer that mismanages t-stat opening.
Circulating coolant makes the temp sensor measure TRUE HEAT level.
Sounds good... then what ?
I can write this because I've experimented around to figure this out.
The key is to run AUX on the NA engines (on TT AUX is plumbed to run charge air circuit).
I was going to run AUX using a bypass relay and temp sensor - I recalled seeing computer settings related to AUX and had a closer look.
Everything is there available but conveniently disabled. Engineers left this system disabled for us to discover this Easter Egg.
Danke!
The AUX pump menu exposes flexible settings to be used in different ways for NA vs. TT.
At this stage for preliminary testing, what I did was to simply activate AUX=ON.
Next step will be to make it regulate temps knowing that it affects the very same temp measure it's based upon.
We'll activate AUX knowingly with derated temps.
Say good by to uncontrolled engine/tranny heat. We know what heat does to viscosity and basic performance.
Last edited by CaliBenzDriver; Jun 23, 2025 at 03:54 PM.
> WHERE IS THIS....
Here are some F-SAM AUX settings...
In the Front SAM Module
under VARIANT CODING menu option
near the bottom: AUX WATER PUMP
Before you modify anything, remember to save yourself a data backup. Each car has different options and different data backups... not really interchangeable.
POWER USER... WRITE / BKP / RECOV
> ENABLING LOGIC...
The coolant pump can be actuated in different ways based on these settings.
For testing purposes you can play around a bit.
STOCK ACTUATION SETTINGS
We've got to figure what the interaction is between all these settings.
01: really activates pump: ON!
02: Disables automatic regulation
03- for off when chassis sleeps.
04- Aux pump: Active **
The interactions can be confusing and the bottom line is pump was kept off.
> The SOB Setting:
*** AAC module has an option to disable AUX pumping through the unrestricted heater-core on hot days to help cool driver by overheating engine.
Here we disable all the stock chaos...
> STOCK REF. SETTINGS...
We are gonna play with this enough to wonder what defaults were... here what my M276-NA defaults were. For experimental purposes, your chassis may react differently with these options.
options 1 to 4
01: pump power enabled yes/no
02: normal dumbo or commutated intelligent
03: "Chassis sleep" disables pump power
04: the NA engine option is installed.
options 5 to 7
05...07: emergency pumping level in 0/90/600 seconds.
I don't know what triggers these distinct levels and what this does for x-many seconds???
options 8 to 10
09...10: Pretty amazin' 150°C (300F) numbers.
options 11 to 13
11...12: monitor flooded pump: then new pump needed.
13: drain battery down to 10V... then tow truck!
options 14 to 16: bottom.
14: over voltage protection. No use.
15...16: AUX pump is also used to push hot water for stationary winter heat given IGN active.
> Bottom line :
Settings of interest are the actuation mode and temperature range. I aimed for effectiveness.
Last edited by CaliBenzDriver; Jun 23, 2025 at 06:30 PM.
We can try to settle on a set of best options for M276-NA.
My first step was to run the sleepy pump to celebrate the second day of SUMMER.
Engine:ON Bk2 pump ENABLED unconditionally for now
I have not yet tested the COMMUTATED settings: options for self regulation... later!
I know what the options sounds like but I don't care to debug untested menu options just yet.
better number settings ?
09...10: Emcy pump... these are not different speed settings (2wires DC-Pump) ...
-- What do these options control if applicable?
-- We could guess these are the high/low trigger on/off commutation temps ??
The word translation are not really helpful.
don't kill my battery today (default:10.1V) lol
Plenty of F-SAM settings some of them obvious others less. My primary interest is only cooling the exact menu settings are secondary.
Right now I want unconditional pumping to understand what "cool Bank2" feels like.
I got that: ENGINE/TRANNY RESPONDS WITH FURTHER IMPROVEMENTS to regulated Bank2 heat.
Everytime the engine runs better I wonder where the limits are and how slow-poke heavy it was running before.
I am simply providing engine with an honest setup: normal oiling, normal heat... not nitrox boost.
The difference is spectacular based on vibration free engine pressure sensitive throttle! By that I mean throttle acts like an electric car with instantaneous torque response.
This is due to GDI timings being so touchy. The ECU has precise controls under narrow conditions.
When heat runs away uncontrolled performance is down tuned. ECU measures the exhaust gas temperature with its Lambdas.
Uncontrolled extreme heat is the enemy, forget LSPI: man-made nightmares !!
+++ As mentioned: I have not tested "self-regulation". I am running AUX:ON to sense exact difference with cooled bank2.
+++ Some of these settings can be transfered to TT AUX control of charge intercooler.
> FAIR T-STAT OPENINGS ...
With heat circulated around the temp sensor, ECU has a chance to manage T-STAT effectively. Circulating coolant out from the engine dumps heat into radiator and should ALLOW FAN TO STAY OFF reliably.
By circulating coolant, we should get zero fan in cool weather: duh!
Engine will regulate heat normally.
+++ FINDING A CORELATION... -- Emcy T:1,2,3,4 is about 4x actuation duration in seconds
-- Emcy Temp: 1,2 is about 2x trigger temperatures
-- Two of the durations are missing a trigger, so I don't really get this mapping ???
Some kind of on/off cycle based on startup temps, like:
-- If coolant temp crosses temp1 turn on 60sec the turn off: 90 sec
-- If coolant temp crosses temp2 longer timings may be applied .
Last edited by CaliBenzDriver; Jun 23, 2025 at 07:20 PM.
I bumped into a lille snake....time for some dog-food - It comes with untested research territory.
Already I found out the Launch "A/C clutch activation test" is freaking reversed : The Clutch On is Off.
AUX-PMP feature is pretty nice but it comes with a serving of compatibility madness.
I've tried to reset this feature back to vanilla-stock without my A/C restarting.
The bug disabling my compressor may be related to what's interfering SAM-CGW performance.
There is something unstable in there....
A/C COMPRESSOR CTL SOLENOID DISABLED
I don't know exactly what is doing this.
The F-SAM that switches the compressor solenoid or the AAC Module that computes values.
I don't know if this is a normal behavior based on options I have toggled. I know the eggs I am walking on...
I have massaged variant settings in both modules. I would not be surprised if this was a Launch scanner incompatibility bug with FL W212 F-SAM...
Sure enough I did "backup" some of the settings, just not all of them to restore 100%.
I went ahead and gave chassis a reboot for good measure - I want my cool A/C back!!
Actuation test... Solenoid=OK.
Debugging blindly will take some time to optimize the combinations of this can of worms.
Something to be learned about chassis unstabilities.
Last edited by CaliBenzDriver; Jun 28, 2025 at 02:04 AM.
Don't waste your time on this small coolant pump , which is only to boost hot coolant circulation to heater core for passenger.
Unless you dump the heater core heat to cabin, you won't benefit by running this pump hoping for a better radiator cooling performance.
Draw your cooling system schematic first, share with me and I will compare to 3.0 TT version.......... I won't be suprised if you missed a few things
Don't waste your time on this small coolant pump , which is only to boost hot coolant circulation to heater core for passenger.
Unless you dump the heater core heat to cabin, you won't benefit by running this pump hoping for a better radiator cooling performance.
Draw your cooling system schematic first, share with me and I will compare to 3.0 TT version.......... I won't be suprised if you missed a few things
I am testing circulating heat out of bank2 circuit. I like both of my 60° engine banks to be close to equal.
On NA engine AUX-PMP when stopped restricts Bk2 coolant flow. It's the bank that has ECU's temp sensor liar.
I believe engine benefits from heat removal so I am test driving changes.
The tiny main pump impeller is slow at removing heat to "save gas". Personally I do not race my engine so heat removal should be manageable if tstat opening would cooperate.
Eventually it may just be better with a straight-pipe to replace tiny AUX restricting the flow.
The key is not accumulating extreme heat by removing as much heat as gets produced.
I am underwelmed with the stock heat mgt. My target is regulated constant heat without wild artificial temperature swings.
preliminary results were interesting until I had to restore my A/C functions. My tiny AUX-PMP moves heat SIGNIFICANTLY better when running than not.
Thank god you don't have heat issues cooling your turbo engine.
Last edited by CaliBenzDriver; Jun 28, 2025 at 07:04 AM.
I never paid attention to the full design coolant flow of our thermostat assy, until few days back when I have to modify it as a "plug".
A (hot) coolant flow to radiator will only happen when thermostat happen.This hot coolant is from CYLINDER HEAD and not ENGINE BLOCK.
A is the twin shared output port of hot coolant of BOTH cylinder heads, the one with the small centering-sharing hidden plastic pipe.
B is the by-pass path from combo Bank 1 and Bank 2 coolant output, back to mechanical coolant pump.....SUCTION side.
B is located at Bank 2 mechanical coolant pump SUCTION side.
B path will stop* flowing ( *to mechanical coolant pump SUCTION/Input side ) when thermostat is opening full and only A path will flow to radiator INPUT side.
A path is diverted to B path when thermostat is not yet open... aka all goes to SUCTION of mechanical coolant pump.
----------
See below : S = SUCTION.
Two suction ports for mechanical coolant pump. Top one is that B path and not a full time suction zone.
The bottom suction port is full time suction from radiator OUTPUT and the Bank 2 rear azz output via heater core and back to mechanical pump suction port using Tee at hose.
O = Output
Output of mechanical coolant pump is a radiator cooled down coolant, injected into those two holes I am now using as recirculation port.
From mechanical coolant pump point of view, those two are ports for OUTPUTs of coolant pump into engine block, not cylinder head.
So this cooling system is UPWARD flow, I guess it is good to remove air bubble and afterall cylinder/piston liner which is at engine block is the hottest zone anyway.
From engine block to cylinder head is the coolant flow.
Green zone belongs to cylinder head, red zone belongs to engine block
.
.
I believe, and you can test it , that the small baby 12v DC coolant pump for heater core, is only activated during start-stop in engine OFF mode.
During normal engine running, the mechanical coolant pump suction is the one doing all the circulation, using its own suction power.
Before thermostat open, and to prevent mechanical coolant pump from liquid starvation, mechanical coolant pump does 2 region suction.
The azz of Bank 2 heater port via hvac heater core in your case and that by-pass B channel I marked dark orange....if for 3.5 NA
If for 3.0 Turbo, the return or the hot side of turbo coolant circuit also join by using a Tee, the heater core return path to mechanical coolant pump suction side.
The mechanical coolant pump power is huge, compared to the 12V DC cute small baby coolant circulation assist pump for HVAC heater core,
therefore is of no benefit to keep the 12V DC cute small baby coolant circulation pump running WHEN ENGINE IS RUNNING.
I never paid attention to the full design coolant flow of our thermostat assy, until few days back when I have to modify it as a "plug".
A (hot) coolant flow to radiator will only happen when thermostat happen.This hot coolant is from CYLINDER HEAD and not ENGINE BLOCK.
A is the twin shared output port of hot coolant of BOTH cylinder heads, the one with the small centering-sharing hidden plastic pipe.
B is the by-pass path from combo Bank 1 and Bank 2 coolant output, back to mechanical coolant pump.....SUCTION side.
B is located at Bank 2 mechanical coolant pump SUCTION side.
B path will stop* flowing ( *to mechanical coolant pump SUCTION/Input side ) when thermostat is opening full and only A path will flow to radiator INPUT side.
A path is diverted to B path when thermostat is not yet open... aka all goes to SUCTION of mechanical coolant pump.
----------
See below : S = SUCTION.
Two suction ports for mechanical coolant pump. Top one is that B path and not a full time suction zone.
The bottom suction port is full time suction from radiator OUTPUT and the Bank 2 rear azz output via heater core and back to mechanical pump suction port using Tee at hose.
O = Output
Output of mechanical coolant pump is a radiator cooled down coolant, injected into those two holes I am now using as recirculation port.
From mechanical coolant pump point of view, those two are ports for OUTPUTs of coolant pump into engine block, not cylinder head.
So this cooling system is UPWARD flow, I guess it is good to remove air bubble and afterall cylinder/piston liner which is at engine block is the hottest zone anyway.
From engine block to cylinder head is the coolant flow.
Green zone belongs to cylinder head, red zone belongs to engine block
I believe, and you can test it , that the small baby 12v DC coolant pump for heater core, is only activated during start-stop in engine OFF mode.
During normal engine running, the mechanical coolant pump suction is the one doing all the circulation, using its own suction power.
Before thermostat open, and to prevent mechanical coolant pump from liquid starvation, mechanical coolant pump does 2 region suction.
The azz of Bank 2 heater port via hvac heater core in your case and that by-pass B channel I marked dark orange....if for 3.5 NA
If for 3.0 Turbo, the return or the hot side of turbo coolant circuit also join by using a Tee, the heater core return path to mechanical coolant pump suction side.
The mechanical coolant pump power is huge, compared to the 12V DC cute small baby coolant circulation assist pump for HVAC heater core,
therefore is of no benefit to keep the 12V DC cute small baby coolant circulation pump running WHEN ENGINE IS RUNNING.
Yes that makes sense, coolant must be circulated to move engine heat to radiator.
main coolant pump pulley ratio (we can see oil solenoid at 8 o'clock position)
M276 main pump on slow pulley ... economy sized pump
We have seen many engine water pumps to say this 276 unit is HALF-SIZED.
Porsche pump... not tiny volume
On low Rpm the main pump has very limited effectiveness to speed up water flow throughout the circuit.
-- Around city stop and go traffic, coolant flow is extra poor given low Rpm.
-- With Hwy constant speed heat is moved around more effectively.
-- HOW DO I KNOW... : this explains how the Tstat opens wide during city driving then over-cools once engine gets up to Hwy speeds.
This witnesses heat is not being removed in stop-N-go Rpm - That means heat is accumulated by ineffective pumping.
-- Idle engine hardly produces any heat.
-- Our best bet is not to get to traffic stop engine idle with already pre-accumulated heat. ... MOD-X gets that.
> RPM RANGE MATTERS...
-- I am only concerned about "normal driving Rpm" from idle to 2500Rpm where engine spends most if its time.
-- I don't mind what volume coolant/oil pumps can be moved when spun at 4 or 6k Rpm
-- I particularly care what pumps move at 1750.Rpm.
The electric AUX-PMP is limited to half engine on bank-2. I know it's intended purpose is NOT to cool engine...
It's actually great because its single speed is effective exactly when main pump is not. Its a perfect heatsoak assist for 276-NA.
> SIGNIFICANT HEAT YOYO...
I am dealing with this topic because extreme heat significantly de-rates viscosity of both engine + tranny.
Regulating heat helps ECU run its GDI timings.
There is a BAD HEAT rollercoaster EFFECT, it's not mild at all. How do I know... I partly smoothed it experimentally. My MOD-X setup does NOT have large Qty of heat. I don't need to move a lot, only to be effective at NOT ACCUMULATING A BALANCE of heat.
> HEAT LEVEL MUST BE **REGULATED**... duh!
Stock is unable to regulate normal coolant temperature, it's swamped with heat it can not mobilize.
This is in relation to reliably holding kick-*** GDI timings at normal driving Rpm to glide heavy chassis.
> HEAT TRANSFER IS A SLOW PROCESS
Temperature is the difference between increase minus removal. It's really a balance.
Right now we need to help circulate more heat so Tstat can begin to regulate around normal level. smaller pump pulleys
The real fix would be a smaller pump pulley to spin faster... I'd be game for this ultimate fix. Heat is a killer of performance + plastics.
+++ Testing Bk2 AUX-PMP : ON...
There is a noticeable difference in cabin heat between pump ON vs. OFF. Heater-core has no heater valve, always max heat.
The difference is the heat that's circulated out instead of accumulated in Bk2.
Let me do more test-driving this weekend now that my A/C is back up running.
cooling pump upgrade... a proven concept !! 👍
Last edited by CaliBenzDriver; Jun 28, 2025 at 08:14 PM.
Good, keep on experimenting and show me your data/result.
Surya, you know if M276 heat removal was under control you wouldn't need a collection of Banks gauges nor deal with heat issues.
Heat issues are optional, not mandatory. This car has a HUGE RADIATOR, the heat just need to be transfered there from where it's produced.
-- Last year I drove in 100F weather with Fan: off, no problem. I was really surprised!
-- Fan: ON in Winter was my original call for action... No numbers necessary with heatsoak overboil and vaporized oil.
What's the difference between both situations ?? Tstat working based on false sensor data.
Sick systems are fun to debug. It's best to recognize marginal performances before they mature into pricey failures.
W212 can run so hot it makes the struts bouncy. Thats beyond my comfort range.
Last edited by CaliBenzDriver; Jun 29, 2025 at 02:22 AM.
What a joke .....U think I install Banks Gauge because of thermal issue ?
I am a person who DO NOT like to be lied to by a white liar gauge and I do not like when not having oil pressure and oil temperature information, including for my tranny and DIFF when and if I can install it DIY.
I love instrumentation. Please don't assume anything Cali.
You talk of thermal yada yada and you don't have instruments that logged your drives for thermal matters... what a waste of time...that is my personal point of view.
FYI even at 50% radiator fan PMW speed ( 10% is OFF ) , you can not hear in inside your car, more so when you are at cruising speed.
Go and do your experiment, but share the data and not butt dyno or sound of fan speed you heard.
So that we can relate with real data.
When you drive pass 80KM/H or 50 MPH, the fan would likely be OFF anyway, because the ECM knows the installed fan velocity is much
less than what 80KM/H real wind velocity can do in terms of cooling. The only extra allowance if any for radiator fan , is by the HVAC demand, usually when during hot engine re-start.
Here........ remember 10% fan actuation means OFF.
.
I recalled you have Xentry passthru, albeit it is slow, it should be good enough to log similar data on the move.
You can reduce the PIDs, to only what you need to make it swifter logging. Get screen recorder software or stand alone unit like mine, up to you.
I admire your patience and inquisitive mind set. I don't get your purpose with logging data non-stop through busy gateway module.
For me what maters most are practical results. I avoid mixing desktop computing and motoring.
Sometimes I get more results than what I expected like with oil viscosity.
Strangely enough, the unregulated heat also seems tied to performance like so many other things.
I have limited interest to gather temps all over my engine - You do that FAR BETTER than I ever could. It's difficult to get unbiased quality data.
Instead I focus on actions... :
> The only thing to do with high temps is lower them, so I do just that!
Similarly the only thing to do with thin oil is go up viscosity...
I am delighted that my oil is no longer vaporized
engine not longer smells coolant during heatsoaks
---> WAIT THERE'S MORE...
Instead of collecting boat loads of numbers about MOD-X, I go ahead and test M276-NA Bk2 coolant pump at my "normal driving range".
On TT there is no BK2 pump - The only thing to move more coolant is spin the pump faster: either Rpm or pulley ratio.
Jakarta is not particularly hot from 77 to 90°F. Plenty of cooling gradient.
Here in California we go from 90 to 105°F - I can run with fan:OFF at 100°F, that's my temp number.
In search of better more stable temp regulation that help ECU not derate performance. it's already great the way GDI runs...
I am researching superior driveability parameters of Bosch ECU.
Last edited by CaliBenzDriver; Jun 29, 2025 at 05:15 AM.
100F is no issue Cali, as long as you car is moving minimum 15 MPH.
You have not experience Jakarta snail-stop-snail at 90F, which is worse than your 100F car is moving.
I tracked my car in 2020 at 35C/95F , it hit 120C /248F only after 3rd lap.... decent for a family car, albeit I wished it stayed at no higher than 110C / 230F.
Well, I wish you good luck on your experiment.
I am just telling you that the 12V DC pump is not worth as cooling circulation assist.
I have coolant pressure gauge at the bank 2 azz pipe to heater core, I know the flow-pressure value at all RPM and at different coolant temperature/s.
M276 mechanical water pump is indeed too small capacity when at idle of 550 RPM in D under braking.... if for my climate, but it won't overheat the car ... BUT can't cool down fast too,
if compared to 900 RPM.
I don't have oily air intake manifold like you used to have, so 1 less issue for me which for your case you relate that to heat.
But my HVAC contaminated by heater core is a big deal when I loose 3C at the vent.
Thus heater core get defeated and is what I was busy about,while you are busy on your oil mix experiment.
One thing we share in common is our dislike for 200F hot heater core. What a glorious F -up!
In the name of lower humidity the whole dashboard is kept boiling hot year round... brilliant design !
When my A/C was disabled... I got to sense the difference between circulated and not circulated heater core. There is a lot of heat not circulated out of Bank2.
That means... heat is proven to accumulate in coolant: ouch!
The AUX pump moves significant heat that allows the ECU temp sensor to be more realistic.
The parasitic heat from heater core is significant. It's very noticeable despite flaps being shut with "LO" setting. The AUX PMP does tilt the scale.
What NA engines need is a normal Bank2 coolant flow unlike hot-stock.
+++ Testing... new "EXPRESS COLD START".
I haven't gone far yet... upon starting I noticed the cold start sequence is shortened by like 25% faster to step down to idle. The engine runs really evenly even though its guaranteed in "open-loop" mode - I think GDI timings compute mixture on different bases.
Theres nothing random in perfect timings. I bet ECU closely watches CKP for cylinders effectiveness.
I don't yet understand what role AUX-PMP plays in enabling that ?
I need to let ECU adapt to see what this enables.
There are engine requirements to gain from this GDI thingy.
At the end of the day engine is gaining power and I hope to drop heat under control - This is normally exclusive: power comes with extra heat.
I drop heat and get rewarded with power: arigatō !
Last edited by CaliBenzDriver; Jun 29, 2025 at 02:29 PM.
This doc explains how our ECU fine tunes itself and learns fine-tunes by storing "best runtime" settings.
This explains new advanced cold-start improvements based on prior saved data while still in open-loop.
> REVEALING...
Pay attention to what your cold start is like to understand how engine timings are being run by ECU.
The ECU is only responding to engine hability to run sealed cyls..
> * * PREQUISITES * *...
You guessed it: the Bosch ECU is not a magician.
To gain smooth running, you need to help balance cyl. contributions.
I've been saying this all along: results are limited by oil selection.
upgrade the stock "5W-40 full-syn".
Last edited by CaliBenzDriver; Jun 30, 2025 at 12:25 AM.
Mercedes SLR McLaren 722 S Is Extremely Rare Example Modified by McLaren
Slideshow: A one-of-one U.S.-spec Mercedes-Benz SLR McLaren Roadster became even rarer after a factory-backed transformation at McLaren's headquarters.
