Testing cat temp sensors

John, they're simple Type "K" thermocouples, so get a high impedance DVM and measure their output (in millivolts) at ambient and compare to any commonly available Type K thermocouple charts. You'll need to disconnect them from the Cat ECU before doing this.

Some DVM's are equipped to measure thermocouple outputs, and give the output directly in either degrees C or F. This is a very handy function to have.


It's useful to have 2 points on the curve, but don't try measuring them while engine is running, as when you disconnect the sensor lead from the Cat ECU, it'll set a CEL and you'll lose that bank of the engine (AMHIK).

You can, however shut the engine down after it's warmed up, and pull the sensor leads from their respective Cat ECU's and measure their output then, and it still should be 400 degF or higher.

 

The problem with testing is they tend to fail intermittently. The factory supplied the dealers with test equipment and we often didn't use it because of the intermittent nature of the failures. It became pretty standard to swap components side to side if there were no symptoms of a real overheat to test them.

 

Isn't that their most common failure mode?

 

Yes

 

You already know all you need to know. It is showing an overheat when it can't possibly be overheating. You have swapped ECUs and the problem did not follow.

What more do you hope to learn?

 

The main ECU? I am not sure what you are calling the main ECU but the cat temp ecu turns on the light on the dash with no other influencing components. If you are getting a slow down light either the cat is too hot, the thermocouple is bad, the cat temp ecu is bad or there is a wiring issue between the ecu and the instrument panel. On the last one, I have never heard of that happening but it is technically possible.

 

Brian, with due respect to your good self, I don't believe the way you described is the way this circuit works. (Refer to 550 WSM, page C15.)

The "Cat ECU" as we & Ferrari like to call it is actually just a voltage amplifier that converts the analog, nonlinear, sensor level (millivolt) signal from the thermocouple, up to an analog, semi-linear logic level (1-5 VDC) signal.

This logic level analog output of the Cat ECU does not go directly to the "slowdown" lamp in the dash, but it arrives on pin 14 of the main Motronic ECU, and it's the Motronic which turns on the Slowdown lamp via pin 66.

While I don't have an internal diagram of the Motronics, I assume this pin 14 is an analog input, which gets looked at by some voltage comparator software for 4 different voltage levels:

Less than about 0.5 volts means the Cat ECU, wiring, or thermocouple are faulty, and it'll set a hard CEL

If it's between 0.5 to 3.75 volts the Motronic believes the other components are OK, and the exhaust temperature is in its normal range.

At somewhere around 3.75 volts the Motronic will set the first flashing "Slowdown" alarm light, and is equivalent to about 1650 degF (the WSM manual says 1720 degF plus or minus 86 degF). It will disappear on its own if the cat temperature returns to normal.

At around 3.95 volts the Motronic will set the "Slowdown" light to solid "on" and it will shutdown that bank of the engine and produce a hard CEL which must be reset by an OBD2 reader/analyser. In practice, I found this level corresponds to around 1720 degF, but the manual suggests it's 1778 86 degF, plus or minus 86 degF.

The above levels are not absolute, but are very approximate, and I've seen them vary as much as 10% during testing, which jives with the WSM spec. I'm not sure if the inconsistency is due to the cheaply made Cat ECU's or the Motronics itself, however if I were betting, I would probably bet on the Cat ECU's as being the weak link.

Hope this clarifies this circuit a bit more.

So far, I have seen the Slowdown light produced by the following:
1. Faulty thermocouple sensor
2. Faulty Cat ECU
3. Faulty (blown) input on the Motronic ECU (self inflicted, when +12 was accidentally put on pin 14 by a dubious Cat ECU replacement)

 

Excellent write up as usual, John. It seems in the majority of cases, #2 seems to be the issue (bad CAT ecu)

 

If I understood your post, it's not the Thermocouple that is the problem - it's the CAT ECU. Daniel at Ricambi offers an aftermarket solution from Gothspeed - I'd suggest you get that.

 

Got it - I thought it moved. Now swap the thermocouples and see if the problem moves.

 

Do you have one SDL or two?

If you have only one sdl, you cannot determine where the problem lies, right or left or which part is faulty by switching components, either TC or SDECUs. If you have known good parts, you can keep swapping until the single SDL goes away.

Sometimes you will get a pending OBD code which will be right or left specific for cat temps which will point you to the correct side. If you have an OBD reader, check for codes.

You can tap or back probe the data line output of the SDECU and real time see the voltages and accurately determine where the problem lies. I have done this with long leads into the cabin of the car. Criibj gave a synopsis of the pertinent voltage readings that the main ECU wants to see to be happy.

I concur with the intermittency comments by Brian. I have see this real time while driving and getting subsequent sdl. This makes static voltage readings not so helpful.

In the end without real time data to guide you, swapping with known good components is about the only way to go.

Check the OBD codes if you have only one light, that may point you L/R and then its either the SDECU or the TC.

 

John, I hate to publicize this, but "only" for a test, you can temporarily replace the suspect Cat ECU and thermocouple with a 1.5v battery and see if that eliminates the problem with those components.

For those reading this with a thought toward permanently bypassing this system with a constant 1.5 volt source, PLEASE for the love of GOD DON'T DO THIS. This is a very important safety system, and as long as you're running cats, and driving a Ferrari with the crap Motronic 5.2 system, DON'T DO IT. There have been documented car fires due to bypassing this system, and the owners are still on here, although they can't/won't talk about it.

 

Given the very low voltages, would a good place to start is treating the connectors to remove any potential corrosion? Low cost, easy first step. Just a thought.

 

I did exactly this a couple years ago with a cold start SDL. The SDL has not returned. Nor has the near heart attack...

 

Outstanding advice. DeOxit D100: https://www.amazon.com/Booster-Electric-Connection-Enhancer-Lubricant/dp/B0002BBVN2/ref=sr_1_2?ie=UTF8&qid=1497148963&sr=8-2&keywords=deoxit+d100