Racedriver

Has anyone seen this happen before with the temperature gauge? 2006 C4. It constantly jumps all over the place below 225 but never exceeds it. It moves randomly whether I'm coasting or accelerating. I showed this to Gaudin Porsche in Las Vegas when I first got it back in 2007 but they said it was normal which I felt was a blowoff but I never pressed the issue since the car drove fine. Due to my procrastination, its now out of warranty but I'd still like to know what could be the problem and if its worth fixing with an independent shop. Could this affect performance? Certainly not bringing it back to Gaudin.

Here's a short video of it.

http://vimeo.com/25471791

Vjgtrybno1

Looks more like a tachometer the way it was moving in the video - there is a dead spot in the gauge but that is above 200 degrees - your gauge appears to be faulty. Do you know what the true temp was in that video? My guess would be that it is 190 ish but I wouldn't trust that reading. Hopefully something simple like a shorted wire...

Racedriver

I was driving for about an hour on the freeway so I think the car was at its regular temp whatever that may be. You're prob right about 190. Just not sure if the gauge affects other electronic regulations of the car.

Macster

Video link doesn't work for me but a gage needle that jumps about is generally an electrical connection problem. (You think a coolant gage needle jumping about a bit is worrisome, wait until you see your car's oil pressure gage needle drop a couple of bars' worth of pressure with no warning. My 03 Turbo's gage does this and the behavior is not that unknown.)

Anyhow, obviously engine coolant temp can not change much in a short time frame. The biggest short time frame change I've seen in some cars is when while the engine is warming up the coolant temp rises to the point the t-stat opens and lets in a bit slug of cold coolant and the coolant temp will drop a noticable amount as this happens. But the drop takes a few seconds and is at best only a couple of gage needle widths. I do not know what the actual temp drop is. I do not see this happening with my Porsches.

Now, if you are a bit concerned about what the coolant temp is actually doing buy an OBD2 code reader/data viewer and view the coolant temp in 'real time'.

This will provide you peace of mind the coolant is not actually jumping about.

But... what I have seen doing this over a long period of time is once in a while the coolant temp reading will be at its maximum. 215C. Every great once in a while the coolant temp is reported by the DME through the OBD2 port at its max. value. The temp gage (or instrument cluster controller) probably has a relatively filter which filters these 'spikes' out so you should not see these at the needle, unless these spikes are happening too frequently and are being treated as they were real readings. Then the needle might jump about a bit.

The only way I know would be to monitor the coolant temp in real time as I mentioned above.

Sincerely,

Macster.

Hella-Buggin'

seems like a short between the sender and the gauge.

cvazquez

1-Check the fluid level and bleed the coolant system
2-Add coolant if required
3-Check Gauge, if faulty replace

simsgw

The movement is not random. At least not in that video segment. The high reading is stable at what I take to be the current oil temperature. When it falls, it tries to fall all the way to the key-off position, but doesn't always make it before the needle shoves back up to the reading that is steady in our cars.

Porsches probably use the CAN bus, which is a wired network of data distribution developed specifically for the relatively low speeds, but harsh conditions and criticality of data transmission in cars and similar modern equipment. Also, the 'gauges' in modern cars usually are not, and I doubt they are in a Porsche, though I don't have a source to check off hand. What I mean is that a gauge measures some aspect of the car: speed, engine oil pressure, temperature, what have you. In my early days, the dashboard had real gauges so we had a small tube full of oil running up to the instrument. The tachometer had a spring-based cable driven by a gear at one of the wheels, and that cable ran to the tach where increased revolutions of the cable would increase the position of the needle. Sometimes we used a transducer, a device that changes a phenomenon in one energy mode, like pressure, to another like electrical voltage. Then the gauge in the dashboard would be basically a voltmeter. And so forth and so on. (We also walked seven miles to school in the snow. Uphill. Both ways.) Enough.

A modern car does not have gauges on the dashboard, it has displays. The various information is gathered elsewhere and then sent to displays on the dash.

Some observations:

• It ain't normal. That was a blow-off.
• I doubt it's the CAN bus. I can think of failure modes that would do that, but they are extremely unlikely. The CAN bus if present is also driving all the other instruments and you'd see a lot more symptoms than a jumpy oil temp needle.
• I doubt it's the instrument itself. As I say, it's just a display. Might as well be a circle drawn on an LED screen and since it is holding a reading that looks plausible for much of the time, I'd say the display is just fine.
• Ten to one, it is losing its source of data. That leaves three possibilities: the connection to the wire delivering data, whether that is the CAN bus or some other wired connection to the sensor in the engine; the connection at the other end to the sensor; or finally the sensor inside the engine is intermittent. I can't say offhand, not having worked with them, but I have serious doubts about an oil temp sensor going intermittent like that and staying that way this long. That leaves an intermittent connection between the sensor and the display as the villain of this piece.
• The connection may be a simple wired one. In which case, it could be shorting somehow. If it were the CAN bus, that would affect all the instruments. Since the CAN bus is very likely in a car design so committed to light weight, I can't see a shorted wire being the problem.
• Now we get to the likely problems. First, let's assume for argument's sake that your car is using the CAN bus. Like the network board/chip in your computer, the CAN bus must have a protocol chip at each connected device. The chip may be defective at one end or the other since you say it's been happening since you bought the car. Either the temp display itself or the 'sender', the sensor in the engine that measures oil temperature and creates a small packet of information that it puts onto the CAN bus. That's possible, but it would mean the defective protocol chip is losing its connection to the bus for a terribly long time in computer terms. Notice how long the needle goes without data. Still possible and fitting the symptoms, but I'd sure want to see the diagnostic data if someone settles on this answer.
• The classic, and for good reason. Is it plugged in? This very much looks to me like an old-fashioned intermittent wire, and classically the one we check first is the ground. It is more likely. (Just works out that way. Don't worry about why.) Either the ground at one end or the other is loose, or the signal wire is.

Don't let anybody sell you new parts unless you're really convinced the loose connection is inside the sensor or the display, that is, internal and requiring replacement of the device. It can be, but if the explanation given doesn't start with "it is losing connection" then it's someone trying using replacement in lieu of diagnosis. If it were one of the protocol chips that almost certainly is a problem requiring replacement of the display or the sender. But I seriously think you just have a problem with the ground connection on one end or the other. Failing that, the signal connection is bad.

Gary

P.S. I did see the coolant needle jump that wildly on another car. The head gasket was blown and bubbles of steam were passing over the sensor intermittently. That cause exhibits different needle behavior. It lies normally on the true coolant temperature and then jumps wildly up toward the top peg when a bubble of steam passes. Your gauge is fluctuating in the other direction. As if a bubble of zero temperature were passing. Since that doesn't happen, what we have is an intermittent data connection that causes the needle to home to its default position.

slicky rick

i think gary hit it smack on the head. for me i would like to ask the op if the temp just does it when driving when the car is stationary does is do the wiggly thing? when engine is not running, and hot of course, does is pointer do this? my assumptions would be that it just happens when driving hitting bumps or any noticeable vibrations to the engine. when car is hot and you just turn on the ignition no running engine the temp stays stable. it might be safe to assume connection to engine would be loose. find it and check that.crimp a bit and fit back on.

Racedriver

Thanks! Extremely helpful to know going into the shop!

Alan C.

I've had this on gauges before and it almost always was a poor ground.