Temperature gauge accuracy

My car always sits at the same tick on the gauge once warmed up.....one below the middle/horizontal

I am running an AEMv1 with Gaugeart output. Usually fully warmed up in in the 180 degree range,

I had to park the car with engine running in midday sun this week for a phone call. 30 minutes later I flicked on the Gaugeart display before taking off and was surprised to see the temp readout at 220. But the cluster gauge was still exactly where it usually sits.

1) Has anyone else noticed this?
2) Is 220 worrisome?

Thx!

Is your radiator fan working?


Sent from my iPhone using Tapatalk

Yes it is , or at least AFAIK. I think in that heat it wouldn't have topped out at 220+ otherwise, right?

I assume that the AEM is using a temp sensor in the same location as the OEM ECU? If so, there will probably always be a temperature differential with the gauge sensor which may get exacerbated under low coolant flow conditions. Second, the RTDs in the sensors are very non linear and there is a small variability between sensors. If the AEM is using the OEM sensor, then it may not have the exact calibration for the OEM sensor. Even if the RTD is an AEM sensor, the slight variability between sensors can lead to errors. Errors in the engine coolant temperature for the ECU is not something to get hung up about. The ECU uses the coolant temperature to determine the need for cranking fuel pulses and warm up enrichment. Once the engine is up to operating temperature coolant temperature generally does not enter into the fueling equation. Manifold temperature is the important one. The short answer is that there may be a significant temperature difference between the two locations and there is also no guarantee that the ECU temperature is necessarily super accurate.

The RTD for the gauge is also non linear coupled together with markings on the gauge face which are non linear. Net effect is that I believe that at higher temperatures the gauge may be less sensitive to temperature fluctuations.

If you think you might have a problem, one testing option would be to borrow an IR thermometer and replicate the conditions taking a temperature reading on the thermostat housing at the start of the 30 minute idling period and then take a second reading at the end. If the temperature readings remain relatively constant, then perhaps that provides a level of comfort around what the instrument cluster is reading (you could do a double check with the entrance temperature to the radiator).

you should be 175-190F. 220 is high

Thx [MENTION=26435]Old Guy[/MENTION]. Great write up. AFAIK it is using the factory sensor - I need to check with Mase and Stage 6 on that, but doubt they remember. I will pack my IR thermometer for the next hot day trip and see what transpires!

Search is your friend... according to this post here's the breakdown.
http://www.nsxprime.com/forum/showt...rature-gauge?p=1712505&viewfull=1#post1712505

The ticks on the gauge are:

---- H = 270 F
---- = 253 F
---- = 236 F
---- = 219 F
---- = 202 F
---- = 185 F
---- = 167 F
---- = 150 F
---- C = 133 F

I wonder if your AEM gauge has been calibrated or something else is causing your discrepancy.

RYU said:

Search is your friend... according to this post here's the breakdown.
http://www.nsxprime.com/forum/showt...rature-gauge?p=1712505&viewfull=1#post1712505

The ticks on the gauge are:

---- H = 270 F
---- = 253 F
---- = 236 F
---- = 219 F
---- = 202 F
---- = 185 F
---- = 167 F
---- = 150 F
---- C = 133 F

I wonder if your AEM gauge has been calibrated or something else is causing your discrepancy.

Click to expand...

I love seeing the old school members posting this stuff.

Honcho said:

I love seeing the old school members posting this stuff.

Click to expand...

You've had some gems over the years Paul!

On a side note, I can verify that one tick down from middle is definitely ~185 (+/- 5*) as per the OEM temp sensor reading from my HKS CAMP and verified with the Ztronics as well. My normal operating temps hover between 176-185 on most days even hard track days on the MASiV radiator. It's about 10-25+ deg higher on the OEM radiator.

Interesting. Low half ticks on the gauge match up kind of OK with the AEM output to Gaugeart, i.e. gauge doesn't even move until about 130 or so (need to check to be precise), and one tick down from center is usually around 180 or so.

same here temp needle always a click below halfway...the one time it went up was because of air after a coolant flush

boca rat said:

Interesting. Low half ticks on the gauge match up kind of OK with the AEM output to Gaugeart, i.e. gauge doesn't even move until about 130 or so (need to check to be precise), and one tick down from center is usually around 180 or so.

Click to expand...

The only part I think is suspect in metric is that I believe the OEM gauge scales exponentially. I believe all OEM gauges do that to provide a more effective warning signal to the driver. In [MENTION=18194]Honcho[/MENTION]'s post he showing it scales up 17 degrees between each tick. I'm not 100% sure that's correct but I really have no way to test this since i'm running the S2000 gauge cluster now. I could very well be wrong though.

For example... if this is correct on the S2000 the increments are all over the place

boca rat said:

My car always sits at the same tick on the gauge once warmed up.....one below the middle/horizontal

I am running an AEMv1 with Gaugeart output. Usually fully warmed up in in the 180 degree range,

I had to park the car with engine running in midday sun this week for a phone call. 30 minutes later I flicked on the Gaugeart display before taking off and was surprised to see the temp readout at 220. But the cluster gauge was still exactly where it usually sits.



1) Has anyone else noticed this?



2) Is 220 worrisome?

Thx!

Click to expand...

I responded to this some years back but will do it again , our temp gauges are what is called "buffered" ,that why most people see the needle one click below middle ,regardless idling or driving . If you want to see real temp swing drive with scanner hooked up ,my car will run between 180 -200 with no movement of needle . Heres why, years back most manufactures had real time accuracy temp gauges ,this resulted in customers complaints of overheating [gauge would climb to 220 before thermostat would open and then drop to normal range 190] . I know for fact GM was experiencing this resulting service complaints . You can google "buffered water temp gauges "and read about this .

Agree about the buffered / high latency effect in dash gauges; however, the OP had been parked idling for 30 minutes. If RYU's temp vs marking data is correct, I would expect that the OP should have had enough time for the buffering / latency in the gauge to have washed out and the needle should have been displaying one tick above the center mark - presuming that the ECU temperature reading of 220 was correct.

RYU said:

For example... if this is correct on the S2000 the increments are all over the place

Click to expand...

Hi RYU, these ticks should be accurate to within a few degrees. The increments have been programmed by myself. I could have chosen a completely linear readout but chose to work with the scheme that you now have, as it has been recommended elsewhere (The purpose of the "wide" central reading bucket of 179-208 F is that this is the engine's normal operating temp range and as a driver you shouldn't expect the ECT to start ticking up over that tick mark).

Best regards,

Guys, I am totally confused now. My motor blew the hell up due to coolant leaks/head gaskets which I had no warning on. So, the dash gauge is just a dummy gauge? Because I had no signs of foul play on the gauge before things went KABOOM! More details please....

Jinks said:

Guys, I am totally confused now. My motor blew the hell up due to coolant leaks/head gaskets which I had no warning on. So, the dash gauge is just a dummy gauge? Because I had no signs of foul play on the gauge before things went KABOOM! More details please....

Click to expand...

I'm sorry for derailing your thread a little bit! Here's my guess to what might have happened. The coolant system is an airtight, pressurized system working close to 1 bar. If there's a leak to atmosphere somewhere, coolant fluid will leak out and air will get it. The air forms air bubbles that travels around your cooling system. When exposed to high engine temperatures, these bubbles will expand rapidly, and deprive the heated components in your engine from being constantly flushed by coolant fluid, thereby quickly and locally lead to very high engine temperatures that might lead to warping and catastrophic engine damage. Or destroying the head gaskets due to the rapid expansion of the air cavities.

Why wouldn't such an event be seen by your ECT gauge? The ECT gauge is connected to a temperature sensor that measures the temperature of the coolant fluid, not the engine temperature per se. The overall coolant temp might be only slowly increasing, but the air cavities expand rapidly causing high local temperature gradients that causes the engine failure. You won't have time catching such an event in this case.

Just a guess, I'm no engine expert. Sorry to hear about your situation!

Best regards,

Thanks for confirming Johan.

Folks, so it looks like the S2000 gauge readout I posted was a custom one. Ignore that post. Back to the drawing board...

Johan@SjoebergDP said:

Hi RYU, these ticks should be accurate to within a few degrees. The increments have been programmed by myself. I could have chosen a completely linear readout but chose to work with the scheme that you now have, as it has been recommended elsewhere (The purpose of the "wide" central reading bucket of 179-208 F is that this is the engine's normal operating temp range and as a driver you shouldn't expect the ECT to start ticking up over that tick mark).

Best regards,

Click to expand...