Error code 25

[Jacques]

HI all,

Since a few days ago, I have had an intermitent error code 25 on my Esprit Turbo SE.

 

According to Lotus Service Manual section EMH, it's an error related to the MAT sensor, which measures the temperature cia it's resistor dunction on the chargecooler above the engine, related to a 5 Volt signal from the ecu, ehich is them diminished or not via the resistance in the MAT sensor.

It is also dependant on a few factors, as I have read it:

- triggered because of a measured temperature byt the MAT sensor above 143 deg. del.

- triggered after a minimum period from  engine start greater than 240 seconds.

- something else I don't remember right now.

 

It shows as the "check engine"-light in the instrumentation.

Engine continue to perform as usual.

Anyway, to get rid of that, I read it out from Freescan on my win10 laptop.

It said error code 25 and in engine data, the MAT showed a temp of 148,3 deg cel.

Other than that, I think all other sensors and incoming data to the ecu, reads normal.

It starts up as say 29 deg cel, then raises slowly to say 31,3 deg.

After a few minutes of driving, it raises up to 148 and CEL goes on for between 10 -58 seconds.

No special periodical rereading or retriggering, but more like outside a premade scheme.

 

I have taken off the plug for the MAT sensor, and cleaned the plug. No change.

I have taken off the MAT sensor and cleaned it, No change.

I have installed a brand new original Lotus MAT sensor. No change.

I have taken off all four multiplugs on the ecu, and cleaned them all. No change.

I have wiggeled a bit on the wires to the MAT sensor. No change.

 

After having vleaned connections up today, I have taken another testdrive, adn after a few minutes, the same error code 25 came back on and off. But in Freescan it also showed a MAT temp of between 29-32 deg. cel. So no strange 148. But this is even stranger to me....?

 

I have not yet located the correct wires in one of the four plugs to the ecu.

I have not yet neasured the Ohm resistance in those wires.

I have not yet changed the ecu (have none).

 

Could it be, that the liquid is not circulating in that sepeate system?

Could it be a bad ground somewhere? For example on the chargecooler body onto somewhere?

Could it be a dead ecu?

Something else?

 

All help and suggestions most appreciated.

 

Kind regards,

Jacques.

[Barrykearley]

Jacques - you must check the wiring to the ecu. If you've swapped sensor and are getting the same result it sort of highlights either wiring or dry joint in the ecu....

[Qavion]

Could it be a bad ground somewhere? For example on the chargecooler body onto somewhere?

 

 

The MAT has two wires, one of them is the ground wire. The ground is inside the ECU, not on the engine. The MAP and Baro use the same ground. If these other two devices are reading normally, then I'd say that the earth circuit is ok (It depends where the splice is).

 

The MAT "A" terminal (not sure if the terminals are marked, but the wire is brown/pink) goes to ECU plug 3 pin "C4" (this is powered by the 5 volts from the ECU). The MAT "B" terminal goes to ECU plug 3 pin "C10". This is the earth (black/orange wire).

 

After a few minutes of driving, it raises up to 148 and CEL goes on for between 10 -58 seconds.

No special periodical rereading or retriggering, but more like outside a premade scheme.

 

 

Interesting. According to the manual, higher temperatures are produced by greater resistances, not shorts. If the plugs are clean and the wiring is firmly secured to the pins, then I can only guess that the problem is inside the ECU.

 

After having vleaned connections up today, I have taken another testdrive, adn after a few minutes, the same error code 25 came back on and off. But in Freescan it also showed a MAT temp of between 29-32 deg. cel. So no strange 148. But this is even stranger to me....?

 

 

Unfortunately, we don't know how the ECU responds to very rapid changes. The CEL circuit may respond to fast changes. The MAT temp display may be smoothed and not show rapid changes. However, I know on aircraft, most warning lights/messages are usually generated after certain time delays. You would think cars electronics would be the same.

 

Anyway, rather than wiggling the wiring quickly, I would move the wiring into a certain position and wait a few seconds for the ECU (and Freescan) to respond. If that fails to produce a temperature change, then it looks like you need a new ECU (I just hope the faulty circuit is in the base unit and not the Memcal).

Edited August 19, 2015 by Qavion

[Jacques]

oaky, I tried to measure resistance on both wires from the MAT sensor, to the ECU (unplugged).

On Black/orange, it's 1,5 Ohm.

And on the other wire, the Brown/orange, There is about the same.

I then tried to move the wires about and as suggested, hold it in different positions, while measureing the resistance and also the voltage, and it's the same and steady.

 

One thing I realised though, is that my MAT sensor was not wrapped in 1½ rounds of teflontape to make it tight(?) as described in the Service Notes. So I did that.

I'll take a test trip later on.

 

Question: is it absolutely needed for the sensor to be isolated in the threads, and thereby isolated from the chargecooler?

(since it have it's own +5 Volt and Ground wires).

 

Now, I am out of ideas.

 

Cheers,

Jacques.

[esp88]

Hi Jacques  

 

I think your  problem  is the ECU.,  As I  saw last time  I think  grand connection is OK,  otherwaise  it will show  another problem  becuase  MAT and baro sensors  shared the ground. 

 

The MAT sensor  resistances  work when  low  temp.  resistance is high and   temp. raise  resistance  wil  down  to e.g. 3,4K Ohms for 20 deg.  and in your case it  showed  148 deg. it look like MAT sensor  has no resistance.  or  terminated to grand.  

 

So,  check again the  ECU connector J3C4 (Green connector   terminal 4 and 10 goes to MAT)

 

Alternative connect the dumy resistance e.g  3k Ohm insted for  MAT. and test run. 

 

Finally  you may disconnect the battiry  to  power off the ECU  wait  5 min  and  reconnect.

 

 

 

Good luck 

 

 

Yasuo 

[Jacques]

Hi Yasuo,

 

Very good idea about a resistor with the equivalent of say 20 deg cel (scale in Service Notes). I'll do that asap.

 

Cheers,

Jacques.

[Qavion]

oaky, I tried to measure resistance on both wires from the MAT sensor, to the ECU (unplugged).

On Black/orange, it's 1,5 Ohm.

And on the other wire, the Brown/orange, There is about the same.

 

 

Jacques, were you using a digital multimeter? What was the reading when you shorted the meter test leads together. For such a short length of wiring, the resistance should be in milliOhms, not Ohms. However, even if the wires are 1.5 ohms, this shouldn't affect the readings. 

 

is it absolutely needed for the sensor to be isolated in the threads, and thereby isolated from the chargecooler?

 

 

I'm not sure what the teflon tape is for. It may be to ensure no air leaks (in or out). It may be to help prevent thermal conduction from the metal parts of the chargecooler to the sensor. Are the threads on the sensor metal? Anyway, with the MAT removed, check for continuity between the pins/sockets of the MAT and the threads on the MAT. If isolated, then you know the thread is not part of the circuit. By the way, do you have an earth lead on your chargecooler? Even if the chargecooler is not part of the electrical circuit, it's good to have all metal parts of the car bonded together.

 

Note that the resistance of the MAT circuit is inversely proportional to the temperature. If you had a high resistance in the MAT wiring, say, due to corrosion or an open circuit, the temperature reading will go DOWN, not up.

[Jacques]

Regarding the reverse proportional resistance measureing: Yes, and that is why I think it's a ECM error.

Regarding the earth on the chargecooler:: Yes, the wire is there and intact, and measured to be functionin.

Yes, the thread on the MAT sensor is metal.

Inserting a brand new unopened original Lotus MAT sensor, did not make a difference.

Yes, I have a Fluke 179. I'll measure it Again as soon as my left knee is working better

Car runs fine otherwise. I also tried to drive and suddenly stop and run out and put my hand on the chargecooler, and it certainly cooler than anywhere else in there. I wasn't Cold, but more a bit like lukewarm. It was  35 deg cel yesterday, when I tried.

 

I have now borrowed a S4s ecm, which I will test with, and report back.

 

Also, I will get a 3KOhm resistor and short the MAT wires with that and see, if the temp readout gets normal and correspond to the scale indicated in Service Notes. I'll do that with both my own and the borrowed ECM, to try and get some reference behaviour.

 

Kind regards,

Jacques.

[fjmuurling]

Have you checked the resistance across the two sensor wires? ie measuring across terminal 4 and 10 on the connector when the green connector is unplugged from the ECU. If it is shorted than the problem is in the wiring.

Also measure the resistance between pin 4 of the connector and ground (engine). If low reading than also wiring problem.

 

Measure with sensor disconnected and connected.

 

Make sure that the ignition is off when (dis)connecting connectors on the ECM. Page 9 of EMH.4

 

The teflon tape is just for sealing.

The groung wire on the chargecooler is for preventing a build up of static electricity.

[Jacques]

Thanks.

Yes, I did measure resistance across both wires, being from unpluggen connector at ECM, to unpluggen connector at MAT.

Did not yet measure from one leg to the other at the 3 ecm plug. Will do!

 

Kind regards,

Jacques.

[Jacques]

update:

I've installed a new S4s computer, and the error comes still, but kind of intermittedly. Sometimes the 148 deg cel goes to 32-39 deg cel.

That's strange.

I don't understand this.

Before installing the new ecm, I took the MAT plug off the MAT sensor, and inserted a 3,6 KOhm resistor. Measured in the other end at the ecm, and it showed 3,6 KOhm.

 

No more ideas.

 

Jacques.

[Qavion]

You have my sympathies, Jacques Intermittent problems are a pain.

Do the wires and plugs seem in good condition? Now that you've replaced the ECU, wiring can be the only problem. If the air temperature really was 148C, you would burn your hand if you touched the chargecooler.

All I can think of is a short between the two wires (perhaps in the wiring plugs) or interference from other wires in the same harness. Where the wires run through the side panel, is there any sign of chafing?

[Jacques]

Yes, that is also my latest thought. So, imagine that there is endless resistance, then the readout in Freescan will show -40 eg. cel.

Imagine that with very little or no resistance, the Freescan pgm will read plus lots of temp.

So, if a wire is broken, it is the same as high resistance, so it should be a very very low temp.

But in my case, there is a very high temp (fake of course, as the chargecooler is cool or just lukewarm, depending.

This lead me to think that there is something fishy going on with the +5 Volt supply wire.

Why? Because I cleaned up the connectors, which were already clean, and I also installed another ecm (S4s).

I took another drive late Friday evening, and had a friend read the sensor readout on Freescan.

The cel light goes on, when the readout suddenly jumps from say 34,3 degrees cel, to 148 deg. cel. an back.

And that happens in both going fast an going slow, as long as there is a moving up or down of the car, like hitting a small bump, or crossing rough pavement etc.

This leads me to think that the + 5 Volt is having it's isolation rubbed off and shorting the 5 volt to another circuit.

I'll investigate this later today.

 

ZZzzzzz....

Jacques.

 

ps: An S4s ecm is a VERY nice thing The turbo comes on earlier, an faster, and a lot less hesitation, and no stumbleing at 3000.

A must have.

[Qavion]

ps: An S4s ecm is a VERY nice thing The turbo comes on earlier, an faster, and a lot less hesitation, and no stumbleing at 3000.

A must have.

 

 

Don't forget the S4s engine/turbo is different from the SE and its ECU/Memcal may not be completely suitable for a standard SE. I have an S4s style chip in my ECU, but have a slightly larger turbo internals, WC engineering injectors, air inlet trunking mods, etc, so may be a little closer to the S4s.

[esp88]

So, my ECU work ok then you have to re-check The there Two wires between ECU and MAT.

Good Luck

[fjmuurling]

Make some new wiring between the sensor and the green plug.

[Escape]

As said above, high temperature means low resistance. So you are looking for damage to the wiring to the sensor allowing them to make contact and thus bypass the (higher) resistance of the sensor.

Have you measured the resistance at the ECM plug with the sensor installed and connected, as suggested? You have eliminated all other likely causes, so it must be wiring.

 

Good luck,

 

Filip

[Jacques]

Right, so I wnet to my Esprit, battery off, ecm plug 3 off, cleaned the contact cleaner off, and sprayed a Little of the same from another brand I trust. Then I took some small cotton Sticks and added a bit of isoprophylalcohol and cleaned up the pins on the ecu.

Then I measured the resistance, using a heatgun (at a safe distance) and measures Ohm. Up and Down like it's supposed to.

Then I connected battery, ignition on, Freescan on and then a Voltmeter, Ohmmeter and an amperemeter. No strange behaviour.

Then I disconnected the MAT sensor, cleaned the plug in a similar way, and measured in that spot the same as before.

Got a 3,3 KOhm resistor and inserted that in series.

Got something strange, and took a very small screwdriver and slightly bent the small female pins in the plug inwards. Back together showed no errors. Took a testdrive, and Error 23 came up and showed -38 deg cel.

All right: no error code 25, took the small screwdriver and bent the female pins a bit more, and then driving -  no problems so far. But I only drove 5 kilometers home. So couldn't say for sure.

Now off to test it in the countryside.

 

Ziiiiip,

Jacques.

[Jacques]

it Works (on the borrowed ecu from Yasuo).

 

Now to change it back to my own ecu.

 

Cheers,

Jacques.