Coolant Temperature Reading

I have a 2001 996 C2 that is running a little hot. In traffic, 85 degrees ambient temperature, I will run 103 C or higher at times. I have been using the HVAC head unit to monitor the temp. I recently bought an OBC reader (ELM327 WiFi) and installed ODB Fusion on my iphone 5. The OBC reader is reading about 3 to 4 C less that the HVAC display. Why would they be different and which one is the accurate reading? Am I reading two different sensors? Thanks for your help.

Quotes source post: 10610736

Can't explain why the two readings disagree. Since all cars support the coolant temperature via reading it via an OBD2 scan tool while only a couple of years of Porsches support the coolant temperature via some proprietary HVAC diagnostics mode, I'd tend to believe the OBD2 reading is the more accurate one. You can do a test if you want. With the A/C off -- so the radiator fans do not run until engine coolant temperature gets high enough --- warm up the engine and then monitor the coolant temp by the OBD2 port tool and the HVAC. When the radiator fans come on -- you may have to raise the idle speed to 1500 for some seconds -- note the two temperature readings. The one that reads 212F is the most accurate one. Maybe. My info is the fans come on at 212F and shut off at 205F. BTW, I do not believe your car is running a little hot. Coolant temperature can easily reach 212F in traffic on a 86F ambient temperature day. Heck with a bit of hard driving the coolant temperature can reach 226F which is still not too hot as long as the system is *pressure tight*!

I've seen the coolant reading on the HVAC display hit 114 C (OK, it happened once, in a Dairy Queen drive-thru ironically enough.) This did NOT trip the engine over temp. So clearly there are different sensors or the temperature data is being displayed (interpreted?) differently. For me, (and I live on the surface on the sun, also known as central FL) if I heat soak the engine on the highway, with A/C on, I run at about 91 C. When I come off the highway the temp will hit perhaps as high as 107 C until the waterpump catches up. Then, even in traffic, the cooling system will drive it back down to about 101 C worst case. Thanks for the heads up on ODB Fusion, have you compared this to DashCommand, Rev, or Engine Link?

I could not get Engine Link to work with the ELM327 WiFi. It would connect for two or more samples then lose connection with the ODB unit. I did try Rev Lite but it really is a demo and not very useful. ODB Fusion is the only one I could get to work. It disconnects every once in a while. I just wonder if the Bluetooth version works better.

I just hooked up EngineLink last night and ... WOW! There is one wrinkle to getting it going, you need to give it a router number in the WiFi connection page under general settings, then connect the software IP (in EngineLink, Rev, whatever) to THAT IP. I suspect this is why many people have reported problems with various WiFi ODB gizmos. But here is why I'm excited ... this thing (EngineLink) takes customizable PIDs. In other words, if we can sus out the Porsche specific "Durametric" DME codes then we can read all that stuff as well (misfires, over-revs, etc.) Soooooo .. anyone out there have the Durametric PIDs? If I can get the codes I will happily create the required data file for EngineLink and make that available.

Quotes source post: 10611641

Actually a fast food drive through is a good place to see coolant temperature climb if the day is mild enough that you have the A/C off. When I used to eat fast food I'd often be in the drive in line and hear the radiator fans come on, run a while, then shut off. Sometimes more than once. Even at 226F -- which is 10F over the temperature at which the fans switch to high speed -- the engine over temperature warning never appeared. The first time I encountered this I asked my SM about this and he asked the factory. The word back was if there was no warning light on either steady or flashing there was no overheating condition and no problem. The only time I have seen the warning lights come on have been once after hitting some road debris and damaging a radiator. The cloud of coolant vapor out the back of the car told me the worst and the warning lights confirmed it though they did not come on right away only after the coolant level dropped from the leak. Another time was when the coolant tank got low from what proved to be a leaking coolant tank cap. The water pump isn't behind and it doesn't need to catch up. It just takes time to circulate the coolant through the radiators and shed the heat. The water pump runs at whatever speed the engine is running at so coolant is always flowing and at high engine speed flowing quite a lot of coolant. I forget what the flow rate is. (The oil flows at up to 60 liters per minute at high engine speed. I remember that.) If at all possible after hard running it is best to continue to drive the car but at a slower pace, with less load. That the car is moving has the radiators receiving better air flow and the engine rpms are up some which means the water pump is flowing more coolant which means this will work to cool the engine down quicker than just bringing the car to a stop and letting the engine idle.

Hi Macster! Of course I didn't mean to indicate that the waterpump was actually running faster or slower, but rather the cooling system itself was "catching up." Which it is. When the thermostat opens up, the coolant flow RATE certainly increases (we hope.) Thus the waterpump is doing more work. This in turn drives the the coolant temperature back down again, which I (overly simplistically) referred to as "catching up." In response to the OPs original question, I also see the coolant temp reported 3-4 C lower via OBC than the HVAC reported temperature. Here is the coolant temp entry from the PID file ... 01 05 1 Engine coolant temperature -40 215 �C A-40 So the sensor is running from -40C to 215C. This provides 255 values, or 1 BYTE. So the formula is easy (A - 40) with no scaling required. If Porsche uses sensors that run from -37C to 218C then they will need to modify the formula to (A-37) This in turn will cause our generic OBD readers to read 3C low. So the HVAC reading is correct, notice I did NOT say accurate. I also notice there is ANOTHER PID entry for coolant temperature with a 3 byte return value (PID 67) It is also possible Porsche is using this PID instead. Unfortunately the generic PID file I have does not have this PID hooked up (yet) I will experiment further.

Thanks to you all for your insights. I was hoping that the scanner was the correct one. Thanks LordVicious for the explanation of the difference in readings. If your get a PID file for the 3 byte return I would appreciate getting it. Does anyone know how to get an accurate reading? Since this car has had a cracked head due to heat and water flow issues I may be overly concerned. With it getting up to 106C (223 F) and my antifreeze boiling is around 225 F, I just don't want an over temp issue. (I know under pressure the boiling point is actually higher.) Again, thanks.