Oil Temperature gauge Calibration????

patkeefe · 09-22-2007, 05:32 AM

Hi all:

Is there a way to recalibrate the oil temperature gauge???

Homework on the question is as follows:
My temp gauge reads a bit hot. It is an SC, non numerical type. I know the numbers are on the left margin of the gauge, and the middle of the gauge is thus 105C. I am seeing estimated likely 230+F on the gauge. I have replaced the sender...no change.

I have a three temperature test thermostat in the car, and have the probes strapped onto the the two lines to the front cooler, and one on the line to the oil tank return from the aux thermostat. I am seeing a bit less than 220F going to the front cooler; I will assume this is the actual oil temp being sucked out of the sump. I get a 24F reduction in temp returning from the front oil cooler, and I see maybe 210F going back into the oil tank. Please note these are temperatures on the track, and my CIS is richened sufficiently, and this is also a turbo car with oil return into the lrft side chain cover. I thus believe the gauge is out of calibration.

TIA!
Pat

patkeefe · 09-22-2007, 03:31 PM

I suppose I could just take it apart...

OldTee · 09-22-2007, 04:41 PM

Harbor Freight sells a laser thermometer. Go buy one and test away.

patkeefe · 09-22-2007, 04:50 PM

Thanks, OT, but I have all that stuff. I need to know if there is an actual calibration adjustment for the gauge itself. I can't seem to find any references to this procedure anywhere.
Pat

KevinB · 09-22-2007, 05:21 PM

Suspend the sender and place it in oil in a pan or pot that you can heat up on a hotplate. Check the temperature of the oil with a known calibrated temperature indicating device. Wire the sender with 2 wires back to the engine. One for ground and one to the wire going back to the wire for the gauge gauge at the engine. This should help you find out if the numbers on the gauge are accurate.

I have done this with my numbered temp gauge to verify that it was accurate.

I am not sure if the gauge has a way that is accesible to caibrate it. If you had to calibrate it, I would do so at 3 different temperatures to make sure it is accurate and repeatable.

304065 · 09-22-2007, 05:26 PM

Pat, search under early_s_man, Warren did this years ago.

patkeefe · 09-22-2007, 05:32 PM

Thanks Kevin:
I have previously shot the sensor with my infrared gun, and it reads lower than the gauge. If I had an open or short, the gauge should be pinned in one direction or another. I think I would have to introduce some resistance into the circuit (or take some out) to get the gauge calibrated. Three points of reference is duly noted.

John:
Good idea. If anyone knows of a procedure, Warren would.

Kind of like this setup?...(from thermostat testing)

Pat

fintstone · fintstone's Garage

You could probably add a variable resistor on the ground side of the meter and adjust it for the desired deflection of the meter...however, from your description, it would seem your readings are pretty accurate. If you measure 220 going to the front cooler, then the reading at the sender would likely be 5 to 10 degrees hotter.

sig_a · 09-23-2007, 01:32 AM

Porsche engineers designed oil temperature gauge to report heat content of oil within the engine, which is the only place it counts. It's likely your gauge is reflecting normal temperature. Also, those infrared temperature guns are not as accurate as a temperature probe connected to Fluke meter.

patkeefe · 09-23-2007, 03:29 AM

I would think the oil at the sending unit would be cooler than the oil in the sump. Oil in the sump would be the temperature of oil going out to the front cooler. I believe the oil gets hotter on the way to the sump due to stuff like, internal friction.

patkeefe · 09-23-2007, 04:52 PM

So, I checked the calibration of the gauge and sensor (actually two sensors) witha small can of oil, in which I suspended the sensor and my infrared temp probe (same as in above photo). Placed it on a jackstand behind the car, and tied the wiring in with long jumpers. Ground and open pin the gauge on either respective end of the spectrum (i forgot which is which). The gauge seems to read reasonably correct: about 220F at the gauge midpoint (which is 105C), pretty close to 300F at the first line of the red area (150C), and about 144F at the first mark coming out of the warmup band (60C).

The point of this exercise is to see just what the temps are on the track, which now appear to be somewhere in the 245F range, which I suppose is reasonable. Next, I'll calibrate my test sensors on the car with my Fluke at work (following sig_a's suggestion). I may need to look at the scavenge pickup tube, which was ok when I put it together six months ago, and flowrates to the front cooler. Maybe I need a third cooler.

I am still a bit baffled by having a 244F temp at the sender, which is after the internal cooler and before getting any heat transferred into it by the engine internals, when I have lower temperature oil going to the front cooler, which is directly out of the sump.

Pat

Ducman · 09-23-2007, 06:32 PM

Pat,

I'm chasing temp problems as well - frustrating. I'm getting about the same differential as you are between outgoing and returning oil lines (front cooler).

As well my guage appears to be accurate - by means of IR temp at sender

Do you know how to meassure flow rate? Concerned my thermostat is not fully opening.

I rarely get above 200 on the street - but as soon as I'm on the track it shoots up. I get nervous at 240,

Don

patkeefe · 09-23-2007, 06:36 PM

Don:
Hangin there. I will get answers soon enough. I have to see if my ultrasonic flowmeter will work on this. I can say, I heated some Mobil1 up to 350 degrees today, and saw no adverse reaction in the oil.
Pat

patkeefe · 10-06-2007, 04:58 PM

This oil temperature gauge thing is really annoying me. I was at the track last week, with my three temperature instrument secured to the crotch strap on the passenger seat...it is my constant companion. I was reading temps into and out of the front cooler, 25F differential, and the oil line FROM the sump was maybe 210F. The gauge is reading something like 250F. Even around town, I'm reading 220F on the gauge, and maybe 190 FROM the sump. Consequently, I think my gauge is fibbing.

I tested this before, but got inconclusive results. So, today, I figured I'd try to perform the tests in a more orderly fashion:

Here's the setup. The spare oil temperature sender (they both give the same readings) is fitted into the bar across the top of the "pot of boiling oil"...actually a 2-1/2" copper cap. The probe is adjacent. I just hit it with the propane torch, let the temperature stabilize, and see what the gauge says.

The sender is grounded. The green/black to the gauge is jumpered. It seems that at lower temperatures, the gauge is fairly close to agreeing with the probe. After heating the oil to over 325F, it never actually pinned in the red. At the bottom of the red, which is around 304F, the oil was really only about 270F. At the gauge midpoint of 105C/221F, oil temps were consistently below by about 20-25F, so 200F actual reads 220+F..

What disturbs me is this:

I read lower, or at best the same, temperatures directly from the sump, even after correcting for the alleged calibration error in the gauge. So, I'm thinking perhaps I am not getting the full cooling effect from my internal oil cooler. I am confident the intenal thermostat works OK, as I tested it when I put the engine together earlier this year.

Is there a possibility I am bypassing oil internally? Maybe getting some relatively cool oil back from the tank, and proceeding to dump it right into the sump? Meanwhile, the engine runs hellishly hot?

I'll get a new gauge, and try that. I hate being a parts replacer.

Pat

patkeefe · 10-06-2007, 05:49 PM

BTW, here's a picture of my new companion. Just for the heck of it, I stuck one of the probes into the fins of the oil engine cooler and took a drive to warm it up. Dash gauge said 220F, and the probe said 200F.

The dreaded gauge...

dvkk · 10-06-2007, 06:33 PM

Quote:

Originally Posted by patkeefe

Congratulations, you are the first person on any forum I have seen that calibrated an oil temp gauge with something that was traceable. All other comparisons to untraceable, uncalibrated instruments are meaningless.

For anyone that does not have access to such an instrument, use pure water instead of oil. Then take a reading when it boils. As long as you are at sea level, you will get accuracy at 212ºF, which coincedentally happens to be right in the range we are interested.

patkeefe · 10-06-2007, 06:48 PM

Thank you, dvkk
The sign in my office says "if you can measure it, you can manage it". I hate taking WAGs on a $10k engine.
Pat

Eagledriver · 10-06-2007, 07:08 PM

The bottom of your engine is exposed to a 100 mile per hour breeze on the track. The oil is also going through a line allowing it to cool on the way to wherever you are measuring it. The scavenge pump has about 50 percent more capacity than needed to scavenge all the oil therefor it sucks air as well as oil into the line. There may be some cooling due to the pressure drop in the air from the sucking of the pump or the cool air being drawn in through the breather. This is a good experiment you are doing but sometimes occams (sp) razor applies.....you need a better oil cooler setup.

-Andy

patkeefe · 10-07-2007, 08:01 AM

My problems seem to be these:

1 Gauge possibly reading incorrectly
2 If gauge is reading correctly, and the primary circuit (pressure pump/engine oil cooler) is working incorrectly.
3 The secondary circuit (scavenge pump/external cooler) works just fine

So, putting on a bigger cooler is not really treating whatever the root cause of the problem is, IMHO, it is just masking the problem.

I can pretty much say with some certainty that the airflow over the sump and the cooling of the line are negligible, heat transfer-wise. I have shot the sump plate with my infrared, and it tends to agree with the reading on the scavenge tube going to the aux thermostat. The sump has air washing over it, but add in thermal mixing of the air being exhausted which cooles the cylinders, and the stock heat exchangers. The lines will basically offer no heat transfer, because they have no fins, which is why effectiveness of oil cooers goes from (low to high efficiency) trombone, 28 tube brass Carrera, Mocal,Setrab (which I have), Carrera fender mount, and front mount cooler. Of course, the fineed coolers also require airflow, which is what makes the last three listed better. Any fender cooler will cool as effectively as a front mount, provided the same amount of airflow is provided.

Also, is it possible that I have the only oil temperature gauge on the planet that is possibly out of calibration? I found only a couple of threads during searches regarding similiar problems to mine. There are tons of people hunting down cooling problems, and they are treating the temperature being read on their gauges as gospel truth.

Pat

mca · 10-07-2007, 08:18 AM

Subscribed.

Dealing with same issue. This summer my temp gauge has been reading 240 - 250 with normal, non-spririted driving. Last summer it was dailed in at 9 o'clock ... not sure what has changed.

Keeps us posted.