Temp gauge markings and thermostat

[Dennis Nicholls]

The coolant temperature gauge has some markings with numbers but others without numbers.

Between the marks labeled 180 and 250 there is another unlabeled mark. Is the Porsche gauge linear - which would make this middle mark at 215* F?

The spec. for an OEM thermostat is 83* C which is 181.4* F. After spending the day cleaning my radiators, my car now fairly quickly heats up to the 180 mark then levels off, going higher only with demands on the engine (e.g. hill climb). This sounds like my cooling system is working properly.

Oddly enough, Wayne Dempsey in his book "101 Projects....Boxster" says the OEM thermostat starts to open at 187* F and is fully open at 210* F. This doesn't sound right to me. His own Pelican Parts says the OEM thermostat is nominally rated 83* C.

[Ahsai]

Best is to use an OBDII scanner or Durametric to read the engine coolant temp and "calibrate" your coolant temp gauge accordingly.

[JFP in PA]

The coolant temperature gauge has some markings with numbers but others without numbers.

Between the marks labeled 180 and 250 there is another unlabeled mark. Is the Porsche gauge linear - which would make this middle mark at 215* F?

The spec. for an OEM thermostat is 83* C which is 181.4* F. After spending the day cleaning my radiators, my car now fairly quickly heats up to the 180 mark then levels off, going higher only with demands on the engine (e.g. hill climb). This sounds like my cooling system is working properly.

Oddly enough, Wayne Dempsey in his book "101 Projects....Boxster" says the OEM thermostat starts to open at 187* F and is fully open at 210* F. This doesn't sound right to me. His own Pelican Parts says the OEM thermostat is nominally rated 83* C.

Wayne is correct. The biggest problem with the temp display in these cars is that they are both non linear and woefully inaccurate; which makes them little more than "an idiot light that moves". It is not at all unusual to measure actual coolant temps of 205-210F when the needle is sitting at 180.

[Dennis Nicholls]

Well that's no good.

According to Dempsey, a 50/50 mix of OEM coolant with an OEM cap should boil at 255* F. I know you should stay well away from the actual boiling point as bubbles of steam will appear well below that point and they are highly corrosive.

I'll have to hook up my Durametric next time I come in from a run.

By the way cleaning the radiators is a pain, but I sure removed a lot of junk in there.

[JFP in PA]

Well that's no good.

According to Dempsey, a 50/50 mix of OEM coolant with an OEM cap should boil at 255* F. I know you should stay well away from the actual boiling point as bubbles of steam will appear well below that point and they are highly corrosive.

I'll have to hook up my Durametric next time I come in from a run.

By the way cleaning the radiators is a pain, but I sure removed a lot of junk in there.

If you want to do something positive for the cooling system, install a 160F stat.

[Dennis Nicholls]

Wayne is correct. The biggest problem with the temp display in these cars is that they are both non linear and woefully inaccurate; which makes them little more than "an idiot light that moves". It is not at all unusual to measure actual coolant temps of 205-210F when the needle is sitting at 180.

This is a modern trend. After 1994 Mazda put in their Miatas an oil pressure "gauge" that only indicates two positions: zero and "some" oil pressure, driven by a switch that toggles at 7 psi.

At least racing heritage Porsche puts in a real oil pressure.....hey wait a minute.

Porsche puts the key to the left of the steering wheel to celebrate their racing heritage at Le Mans but deletes the oil pressure gauge.

[Dennis Nicholls]

Well I used my Durametric to check the engine temperature just now. I let the car sit until the gauge's needle was on the mark for 180. Durametric reading at that point was 82.5* C (= 180.5* F). That sounds right on the mark. The oil temperature reading tracked the engine temperature a few degrees lower.

But sometimes my Durametric acts funny. It also turned up a P0300 and P0306 pair of codes. Normally these are the highest-priority codes and not only light the MIL but flash the MIL. My MIL only comes on at startup so I know the bulb isn't burned out. And I passed ID smog a few weeks ago where they check the "monitor program" status and passed with all monitors completed. Perhaps my Durametric doesn't like running over the long 50' cable but it is THEIR cable that I paid a small fortune for. Maybe I should hunt up a long USB cable like John suggested.

[JFP in PA]

Well I used my Durametric to check the engine temperature just now. I let the car sit until the gauge's needle was on the mark for 180. Durametric reading at that point was 82.5* C (= 180.5* F). That sounds right on the mark. The oil temperature reading tracked the engine temperature a few degrees lower.

But sometimes my Durametric acts funny. It also turned up a P0300 and P0306 pair of codes. Normally these are the highest-priority codes and not only light the MIL but flash the MIL. My MIL only comes on at startup so I know the bulb isn't burned out. And I passed ID smog a few weeks ago where they check the "monitor program" status and passed with all monitors completed. Perhaps my Durametric doesn't like running over the long 50' cable but it is THEIR cable that I paid a small fortune for. Maybe I should hunt up a long USB cable like John suggested.

Something is not right, once the engine is warmed up to a steady state, the oil should always be hotter than the coolant, which is why the car has an oil to coolant cooler system.

[Dennis Nicholls]

I took the car out and drove it another 10 miles. Didn't need to drive it so much as it was already warmed up.

I unplugged and reseated the connection between the basic Durametric cable and the extension cable a couple of times. I then ran it again. I might have read the engine temp and oil temp upside down the first time. This time when the engine temp was 87.0* C, the oil temp was 90.8* C. That's more like it.

Also funny was that I had just erased any trouble codes. This time I went and checked the monitor program "ready" status. It showed a "pass" on all 5 monitor programs. It would require a "drive cycle" to make all the monitors come ready and I doubt driving just 10 miles with a hot car would do so. This again makes me think the two P codes were spurious.

I am running software version 6.3.2.5 which according to the website is the current version.

[JFP in PA]

I took the car out and drove it another 10 miles. Didn't need to drive it so much as it was already warmed up.

I unplugged and reseated the connection between the basic Durametric cable and the extension cable a couple of times. I then ran it again. I might have read the engine temp and oil temp upside down the first time. This time when the engine temp was 87.0* C, the oil temp was 90.8* C. That's more like it.

Also funny was that I had just erased any trouble codes. This time I went and checked the monitor program "ready" status. It showed a "pass" on all 5 monitor programs. It would require a "drive cycle" to make all the monitors come ready and I doubt driving just 10 miles with a hot car would do so. This again makes me think the two P codes were spurious.

I am running software version 6.3.2.5 which according to the website is the current version.

Temps are still a bit low.

Normally, if you clear codes, the SAI takes a while to reset to "ready" status, so what you are seeing is unusual.

[mikefocke]

SAI ???

[Ahsai]

The Secondary Air Injection readiness /test.

[Dennis Nicholls]

The weather warmed up so I was able to take a second reading of engine temperature. After a 60 mile jaunt I was able to hook up the Durametric and discovered that when the needle was at the left edge of the "0" in "180" the engine temperature was 96* C = 204* F.

Summary:

Needle reading "180" is nominally accurate at 180* F.

Needle reading left edge of "0" (midway between two marks) is about 204* F.

I think I got lucky and got a more-or-less accurate temp gauge.

I would guesstimate that cleaning the radiators lowered operating temperatures by about 8* - 10* F.

Edited May 17, 2014 by Dennis Nicholls