Does anyone know how the fans on an auto diesel work?

[jimmo62]

Due to an occasional overheating problem, I'm trying to work out how the fans operate from the excellent wiring diagrams supplied by g8dhe and the workshop manual.

I can see how the scavenger fan works - from the temp sensor via the ECU to control the relay.

The other 2 fans are a bit more complex. From the wiring diagrams they seem to be wired in parallel so both should be doing the same as each other all the time. There are 2 control signals from the ECU and I think that if one goes low then both fans will be on a slow setting, and if both go low the fans will both be on fast. I'm guessing that the line marked 15 on the wiring diagram is the slow setting control signal and 16 is for fast but I'm not sure.
15 links to 3Z on the ECU which is "Cleaning fan control signal" to "Cleaning fan relay No.3" and should go low when the coolant temp is over 108C
16 links to 3S on the ECU which is "Cleaning fan control" but it does not say what temperature this triggers at.

Has anyone ever worked this out? Are all 4 possible combinations of the controls ever used (LowLow, LowHigh, HighLow, HighHigh) or does it only use 3 of these?

I may try to work it out by connecting a variable resistor in place of the temperature sensor and see what happens as I reduce the resistance.....

[g8dhe]

I've wondered that myself, but having a V6 its not a problem I can test
I do have some ECU's here for reverse engineering, but its going to be a while before I get down to them and I suspect even then it will only be to handle the I/O structure rather than the whole device as they are not going to be repairable devices in general.

[jimmo62]

Agreed - reverse engineering an ECU is going to be hard so I think it's the black box approach - I'll get a variable resistor and put is across where the temp sensor should go and wind it down to simulate rising temp and see what happens to the control signals.... The temp sensor should be 2.6k at 20C and about 300 ohms at 80C so I'll get a 4.7k pot and see what happens!

[g8dhe]

Yup that sounds like my approach as well! Lets here the results!

[mikexgough]

One quick thing to remember...... the Air Con also uses the Fans.... so factor that into any investigations...

[jimmo62]

OK, been to Maplins and got the potentiometer....hopefully will work it all out tomorrow.....

[hijimhere]

96 2.5 td only ever heard the fans run when the air con is "on". If you feel she is running to hot turn the air con on !!??

[widdowson2008]

15 links to 3Z on the ECU which is "Cleaning fan control signal" to "Cleaning fan relay No.3" and should go low when the coolant temp is over 108C
16 links to 3S on the ECU which is "Cleaning fan control" but it does not say what temperature this triggers at.
..

From tests I did, the fans cone on LOW SPEED when the coolant temperature at the front of the head (hottest point) reaches 96/97 C - i.e. when the thermostat is no longer in control.
and HIGH SPEED comes on at 108 C.

As has already been stated, this does not apply when the air-con is active.

[jimmo62]

That seems right and agrees with my tests.

The potentiometer burned out pretty quickly (it was only a low power one....) so I had to use fixed resistors.

With a 330 ohm resistor in place of the temperature sensor and a/c off, none of the relays/fans are on. With the a/c on, relay B6-02 is activated. So that corresponds to low running temp - fan is only on for the a/c. 330 ohms corresponds to water temp around 80C (spec says between 290 ohms and 340 ohms at 80C).

With 150 ohms and a/c off, relay B6-02 is activated. That would correspond to high temperature running.

With a short circuit and a/c off, relay B6-02 is activated and so are relays B6-01 and B6-03 for the high fan speed. That would correspond to well over temp.

All my fans seem to be running OK so that rules out another cause of overheating - could be the temp sensor I suppose so may have to try swapping that next.

[mikeonb4c]

Just so we are clear, have the other (mechanical parts) possible causes of a hot running engine been investigated?

[jimmo62]

New radiator, new thermostat, new water pump, new pressure caps on rad and header tank, and bled properly - all by a Bongo-savvy garage. Garage looked for the usual symptoms of a blown head/gasket and did not find anything which indicated a failure. Has lost coolant since this work was done on 2 occasions, never enough to trigger the LCA but enough for the level drop to be noticeable.

Only seems to happen on really hot days - e.g. stuck in traffic. With a/c turned on so at least the fans were going at slow speed. But only twice this year so far. Scavenger fan has been seen to operate during bleeding.

It dumps coolant out of the overflow on the header tank (I am catching it in a bottle) - so pretty sure it is not leaking - but I have not been able to catch it in the act so far.

Cylinder head temp gauge fitted and has been briefly up to 105C, usually runs between 85 and 95C which I believe is about right - but again I have not had it overheat since fitting this so not sure if it goes any higher.

LCA fitted just in case.

I suppose if could be a cracked head which only leaks when very hot? or could the temp sensor be faulty and not turn the fans on at the right temperature?

Any ideas welcome at this stage!

[mikexgough]

Worth eliminating the Temp Sensor on the block.... for the cost.... I know many say use the Mazda one ( a denso made job) but the Intermotor one is just as good I've found...

[Northern Bongolow]

i experimented with original mazda fan switch sensors and copy parts, i noticed that there were slight differences in the trigger points when fitted in the head.
they all also suffer from the plastic ends breaking away from the brass probe portion of the sensor, this too altered the resistance/trigger points massively, from not working at all to coming on early.

this sensor also controls the trigger temps of the egr valve and the glow plugs so is an important bit of kit. it may also be involved in the top gear lock up although the gearbox has its own internal temp sender.


good to see someone putting some numbers to the sensor.

[jimmo62]

Thanks guys, I will change the sensor next. Appreciate the help!