W124 230E - Modification to Further Protect from Overheating

Post by **Christo C** » Fri 24 Jun, 2016 12:31 pm

Mod carried out to 1986 230E Sedan W124 M102.982 engine.

A while ago, just at the end of Summer, my new girlfriend started overheating; luckily I was paying attention to the temperature gauge, and later ascertained the cause fairly quickly: the main COOLANT TEMPERATURE SWITCH S25/4 had failed. (3-PIN with red top, 100ºC/110ºC, Switches Fans – located at top front of Cylinder Head, just behind the Thermostat Housing, Part # 006 545 42 24).

3-PIN_COMP.png

main Coolant Temperature Switch S25/4.

This device contains 2 internal switches, so has two modes:
– When the Coolant reaches 100ºC, switches the Engine Fan on (the fan on the engine, behind the radiator, controlled by the Electro-magnetic Clutch). This is the normal cooling mode of the engine for long idling or heavy traffic when there is not much air movement through the radiator due to low speed, hot day etc etc.
– If the Coolant reaches 110ºC, switches the Auxiliary Fan on (the fan in front of the radiator) to Full Speed. This mode is to protect the engine from overheating damage e.g. if the Electro-magnetic Clutch engine fan fails, or Stirling Moss is driving the car hard, fast & furious on a hot day….

When I removed the device and tested it in hot oil I found that BOTH internal switches had failed; therefore, had I not caught the overheating in time and continued to crawl along on a hot day in very heavy traffic, I could have boiled the engine dry, damaged the head gasket, warped the head or eventually seized the engine…. Disaster!

I was interested to find out why the device had failed, so decided to carefully file it open and investigate.

Inside the lower brass part (with 12mm threads which screw into the cylinder head) is a small sealed reservoir of a substance (I read somewhere this is probably some type of wax) which pushes a small piston part upwards as the substance melts and suddenly expands at around 100ºC; as the piston rises it flicks a small spring-loaded switch On. If the temperature increases above 100ºC to 110ºC the piston continues to rise which triggers a second spring-loaded switch.

3-PIN_BASE.png

Lower half showing small piston.

3-PIN_TOP.png

Upper half showing two switches.

I had expected to find failed contacts on one or both switches due to age, but instead I found that the tiny piston part had stuck (in the down position), therefore neither switch would operate, leading me to conclude that there was no possibility the device could protect the engine because of the way it had failed. Later I was able to free the piston and confirm that the device could still work at correct temperatures, however by now I had destroyed the seal and the crimp flange which keeps it assembled, so a new part was ordered.

The M102 engine (with Air-conditioning / Climate Control) also has a second COOLANT AUTOMATIC TEMPERATURE SWITCH S25/3 (1-PIN, 110ºC, located on the passenger side of the Coolant Pump housing, Part # 006 545 90 24), which is a sensor for the Compressor Shut-off Unit N6 (KLIMA Relay). This operates if the Coolant temperature rises above 110ºC and sensibly disables the Air-conditioner Compressor, thereby reducing the load on the engine when it is overheated.

1-PIN_COMP.png

Coolant Automatic Temperature Switch S25/3

This second switch on the side of the Coolant Pump is designed to detect an overheating engine condition and protect the engine to some degree too, but all it really does is disable the Air-conditioner compressor!

So I had a bright idea: What if this second switch could be utilized to do it's job plus turn on all the fans as extra insurance against the failure (open circuit) of the main Coolant Temperature Switch on the Cylinder Head?

The likelihood of simultaneous failure of two such switch devices is extremely remote!

Note also the possibility that the main Coolant Temperature Switch on the Cylinder Head could fail closed-circuit, but that is less of a problem as the cooling fan(s) would be left operating continuously – a cold engine is infinitely more desirable than an overheated engine!

I therefore designed the following modification which is easily retro-fitted as all the original connections are the same as 4mm “Banana” plugs and sockets.

TEMP_MOD_CIRCUIT.png

The top 3 wires can be simply spliced into the existing loom just behind the Coolant Switch connectors. Alternatively, additional 4mm “Banana” Plugs & Sockets can be fitted to existing connections. Diodes D1 & D2 at the “output” of the Relay prevent the Aux Fan from running under “normal” conditions i.e. when S25/4 (the 3-Pin switch) is operating the main Electro-magnetic Clutch Fan correctly. (Strictly, D1 is not required, but the diode is cheap, and provides the same operation & protection if connections to the switches are inadvertently reversed.)

The diode D3 in series with S25/3 (the 1-Pin switch) can be inserted between existing connections by using 4mm “Banana” Plug & Socket; this diode protects the KLIMA relay from possible over-voltage (thru the relay coil) when S25/3 is Open.

I ran all wires across to the side of the engine bay and mounted the Relay & Diodes on the panel in front of the Battery well.

INSTALLED.png

Download a PDF of above at

MERMOD01 COOLANT TEMPERATURE SWITCH 1986 230E Sedan W124 M102.pdf

** Edited 06/09/2017 ..... The above PDF is more useful!

Post by **Christo C** » Fri 24 Jun, 2016 12:54 pm

Hi
...seems the Moderator of the Planet has moved this topic - thanks!

Post by **Tony From West Oz** » Fri 24 Jun, 2016 4:12 pm

Well thought out.
You have covered the most likely scenarios.
How would you know if one of the switches has failed,(so you can restore the temperature switch redundancy)?

Regards,
Tony

Post by **ChristoC** » Fri 24 Jun, 2016 5:41 pm

Good question Tony [emoji3]

Should the 3-pin Coolant Temp Switch fail with this mod fitted you would notice two symptoms, which should alert:
- temp gauge would raise well over "normal" max high of around 100 deg, to around 110 deg when travelling slowly in hot weather before the relay kicks in....
- higher than normal fan noise under same conditions once the relay kicks in. .. Two fans is much noisier, and the front fan running at full speed really roars (compared with half speed thru resistor when switched on by the air-conditioner / climate control system).

Those symptoms may be subtle but I would argue less "trouble" than a overheated engine!

I suppose a LED (+ buzzer?) could be added somewhere on the dashboard whenever this relay is energized, which in essence would be a "3-pin Coolant Temp Switch Fail" indicator; which then begs another question - how do we know if the second 1-pin switch on the coolant pump is also functional?? Same as when this mod has not installed - drain the Coolant, remove switch & test in hot oil.

~ Christo

Post by **MMWA** » Fri 24 Jun, 2016 7:56 pm

out of interest,

when the system is operating as designed, ie, no airflow what so ever below 100c coolant temp while stationary, where does the temp gauge normally sit while moving ?

Do these electronic fan clutches have some sort of speed limiter? how does the fan not explode if locked 1:1 with the engine at high revs

Post by **Tony From West Oz** » Fri 24 Jun, 2016 10:19 pm

While I have not shown the current limiting resistors (they are often within the LED housing for 12V applications) in the attached diagram, the addition of 2 more diodes and 3 LEDs will tell you what is happening with the 2 temperature switches.
The changes are shown in green.
Tony

Overheat mod.png

Regards,
Tony

Post by **Tony From West Oz** » Fri 24 Jun, 2016 10:22 pm

MercMad West Australian wrote:out of interest,

when the system is operating as designed, ie, no airflow what so ever below 100c coolant temp while stationary, where does the temp gauge normally sit while moving ?

Do these electronic fan clutches have some sort of speed limiter? how does the fan not explode if locked 1:1 with the engine at high revs

Exactly the same way any other engine fan doesn't explode. It is designed for the stresses which could be applied in anticipated usage.
Tony

Post by **Ivanerrol** » Fri 24 Jun, 2016 10:26 pm

If the cooling systems has a good circulatory water pump and there are no leaks in the system and the system is operating at correct pressure then the electric fan should only operate when the vehicle has been left sitting idling for an extended period.
Even then, if the electric fan clutch fails (or non electric variety) then the auxiliary fans will come into operation and should prevent engine damage.

Post by **Tony From West Oz** » Sat 25 Jun, 2016 12:44 am

Ivanerrol wrote:If the cooling systems has a good circulatory water pump and there are no leaks in the system and the system is operating at correct pressure then the electric fan should only operate when the vehicle has been left sitting idling for an extended period.
Even then, if the electric fan clutch fails (or non electric variety) then the auxiliary fans will come into operation and should prevent engine damage.

Providing the temperature switches are working.

Post by **Christo C** » Sat 25 Jun, 2016 9:02 am

Quite right Tony, exactly what I had in mind

; .... indeed I had something like that rigged whilst testing this all out, with LEDs gaffered under the rear lip of the bonnet centred over the windscreen wiper mechanism, just visible from driver's seat even in bright sunshine, but decided on the simpler approach as hacking extra holes in firewall & instrument panel seemed too much hard work (besides looking non-standard), and there never was any failure indicator for the Coolant Temp Switches except the Temp Gauge and driver observation.

As shown, in your diagram, under “normal” conditions the lower LED (at bottom of Relay) would be always illuminated. I suggest it would be preferable to change the LED connection slightly so as to only indicate when the Relay (& 1-pin temp switch) is ON, as per this diagram:

Modified Circuit. (which is included in the PDF in originating topic above)

TEMP_MOD_CIRCUIT2.png

MERCMOD01_CCT_MOD2.pdf

Actually, if you think about it, albeit that 3 LEDs gives a more complete situation, all that is really needed is LED3 (the LED on the Relay), perhaps in Parallel with a small Buzzer, as, under normal conditions this Relay should be never energized; only partial or complete failure of the 3-Pin Coolant Temp Switch, or perhaps extreme overheating of the engine, should ever result in the 1-Pin Coolant Switch operating the Relay as far as I can ascertain.

Post by **AMG** » Sat 25 Jun, 2016 9:50 am

Just an FYI and observation.

the m102.983 16v engine has an electric main fan, plus an auxillary HVAC fan.
The operational arrangement is via the same 3 pin temp switch in the front of the head / thermostat housing and the HVAC pressure switches control the aux fan.

the 2 stage switches will cut in at one of the 3 temps - 100, 105 or 110, depending on the 'colour' (part number) and the HVAC aux fan trigger is at 110, 115 or 120 depending again, on colour.

The HVAC pressure switches are there to protect the compressor, firstly to disable clutch against high pressure, the second is to trigger the fan when the coolant operating temperature is above 65ºC...

So, there is clearly a 'method' in the madness - if you're acutely aware of an electric main fans inability to operate, then turn on the AC, and the HVAC auxillary fan will very quickly bring the temperature down and maintain it (without the main fan clutch engaged) to the thermostat opening temperature.

I have proved this many times with thorough observation and numerous cooling issue troubleshooting in the 16v.

While your backup system is a foolproof solution, a keen eye on the temp gauge should easily mitigate this issue.
Further to that, if you have the HVAC on and crawling in traffic, then your w124's auxillary fans should quite literally blow the pants off your radiator, and keep it cool.

So I'd be looking into firstly, the operation of the fans (are they pushing the correct amount of air?) the cores of the condensor and radiator (blocked with crud to reduce airflow?) and the core of the radiator (water flow), as well as the water pump (impeller = known point of failure in m102's, 3's & 4's).

All of these things are quite literally on the 'short list' of things to check when the cooling system starts to go ape. Unfortunately for you, the temp sender failed. But believe it or not, it's as common as meadow cakes on a dairy farm. they fail ALL the time. So, diligence would say, carry a spare, change it every couple of years, when you drain the coolant, or at least test it for operation in isolation (like you would do with a thermostat).

Nonetheless, I like your schematic, and it adds a layer of protection which let's face it, the factory did not think about. I'm not a big fan of the cooling system in the 124 or 201.034, and your original post pretty much nailed it on the head as to why.

thanks for sharing.

Post by **Christo C** » Sat 25 Jun, 2016 10:11 am

Thanks AMG for your observations and kind comments!

Seems this was the first time the 3-Pin Coolant Temp Switch failed, so I was "lucky".... ha!

A while ago a young man in the street introduced himself, mentioned he is a an apprentice at a Merc stealership, and started chatting about my girlfriend .... eventually he asked if she had a new head gasket fitted lately .... now I know why!!!

Post by **ChristoC** » Sun 20 Nov, 2016 7:39 pm

Many months later, after a pleasant road trip to Victoria during winter, sadly I found I didn't entirely dodge the bullet after that one overheating incident - to be fair maybe 30 years of ageing had something to do with it - anyway I found a small amount of engine oil frothing in the coolant.
Diagnosis was bad head gasket, so reached deep into wallet (too busy to do it myself despite the desire) and had head / valves refurbished and head gasket replaced, as well as timing chain and associated tensioner/guides & crankshaft sprocket replaced.
What a shame I didn't dream up this mod before the overheating incident; anyway the mechanic said I should be good for another 30 years now.... and at least I should never have that particular problem occur again.

Sent from my iPhone using Tapatalk

Post by **compress ignite** » Mon 03 Jul, 2017 11:43 am

Thank Y'all !

'Converting 124.128 from Mechanically Driven (Serpentine Belt) [Viscous Fluid Filled Fan Clutch]
to Electric "Puller" fan mounted behind and "On" the Radiator.
(Using GSXR's tutorial,upgrade of Alternator to 143 Amp from 90 Amp)
http://www.peachparts.com/shopforum/car ... hotos.html
Electric Fan will:
Save the Wear and Tear on:
1.The Harmonic Balancer
(One Way Clutched Alternator Pulley)
2. Water Pump Bearings
3. The Serpentine Belt Tension-ing system

AND This post answers the "How To" of Backups and Alerts for a System
that was originally designed for non-Electric Fan Operation !

Post by **Christo C** » Thu 06 Sep, 2018 7:42 am

Two years later I suddenly realized I had overthought the solution to the failure of the main 3-Pin COOLANT TEMPERATURE SWITCH S25/4

After much thought I came up with a simple but elegant solution.

My car is not the SLS variant, therefore no SLS pump on front of cylinder head; just a couple of threaded bolt holes.... so this will only work for non-SLS vehicles

Simply add another main 3-Pin COOLANT TEMPERATURE SWITCH and wire in Parallel to the existing switch; chances are that two switches will not fail simultaneously!

So I made a simple mount from scrap Aluminium bar, and bolted it to the front of the Cylinder head, just next to the Thermostat, thus:

3-pin Temp Switch Add1.JPG

I also discovered there is a variant of the 3-Pin COOLANT TEMPERATURE SWITCH S25/4 Part # 006 545 42 24 available, which appears to switch approx 5ºC lower @ 95ºC/105ºC; = Part # 006 545 42 24 L

Post by **John Green** » Sat 08 Sep, 2018 8:54 am

Christo C wrote: ↑Thu 06 Sep, 2018 7:42 am Two years later I suddenly realized I had overthought the solution to the failure of the main 3-Pin COOLANT TEMPERATURE SWITCH S25/4

So I made a simple mount from scrap Aluminium bar, and bolted it to the front of the Cylinder head, just next to the Thermostat, thus:

I'm not so sure that will work. The temp sender needs to be mounted into a coolant passage. If you have access to a non contact thermometer check the temp of you original sender and the new one.

Post by **Christo C** » Sat 08 Sep, 2018 9:26 am

John Green wrote: ↑Sat 08 Sep, 2018 8:54 am
Christo C wrote: ↑Thu 06 Sep, 2018 7:42 am Two years later I suddenly realized I had overthought the solution to the failure of the main 3-Pin COOLANT TEMPERATURE SWITCH S25/4
So I made a simple mount from scrap Aluminium bar, and bolted it to the front of the Cylinder head, just next to the Thermostat, thus:
I'm not so sure that will work. The temp sender needs to be mounted into a coolant passage. If you have access to a non contact thermometer check the temp of you original sender and the new one.

Actually John, at first I thought "It can't be as easy as this!" but it works fine - and I did use a non-contact thermometer to check.
The added aluminium bar reaches and typically tracks the cylinder head temperature within a poofteenth of a degree.
I also used electronics heat-sink compound between the Al bar and the cylinder head to increase the thermal conductivity.

What I didn't mention is the 6 months of research that went into this solution - careful tests with the additional switch mounted but isolated, and both connected to some LEDs (mounted under the lip of the bonnet) so I could monitor the switching whilst driving, in all sorts of conditions - generally this coolant switch only operates when crawling in heavy traffic on a warm day - in all cases the switches worked in perfect harmony (given that when tested in hot oil both switches were slightly different anyway, but probably within half a degree of each other).

If you look carefully at the picture you will see that I made the bar from two short lengths of 20x12mm bar - only the upper piece has a threaded hole for the switch: when joined the bottom face of the switch is in contact with the lower bar (again greased with heat-sink compound). The two pieces of bar are simply screwed together from underneath.

I think this is a far more elegant solution as it is far less complicated and doesn't let the engine reach 110ºC before it kicks in, as was the case in the original method.