High Temps with Alekshop Oil Cooler

[Gossypiboma]

Originally Posted by s85e90

It's known that it's design is more for a quick visual aid than anything else. This also goes back to the e46 gauge. Take a look at it. It's a 90 degree sweep with the dot being a 45 deg interval. I'm not saying it's terrible, but if you're SC or even trying to measure the effectiveness of an aftermarket cooler, you need to see individual degrees or line markers and not use the OEM gauge as a reference at all.

It's a known fact the gauge design is like it is so that it doesn't move around too much. If it was designed with being a real measurement tool and not just an Oh! Shit car is getting too hot reference, the needle would have a much wider range of movement and BMW did this to eliminate that and prevent every silly M3 owner from going to the dealer to ask why the gauge is moving around.

Sorry man, you are wrong.

The dash gauge is quite accurate. I have had an oil temp gauge and they read the same. Oil temp doesnt fluctuate rapidly so inherently the gauge doesnt have large swings frequently.

[Leonardo629]

Originally Posted by bdaddylo

After the installation of the Alekshop oil cooler, I'm also seeing higher temps during highway driving than with the OEM cooler. I was expecting to see a drop due to the supposedly increased cooling capacity of a larger cooler. Not very happy with the results.

Just for reference, my oil temp is the same as yours, but I was doing 10mph stop and go on the freeway in 95F weather. All stock cooling hardware for my car, no supercharger.

Maybe Alekshop can give you an explanation?

[surlynkid]

Does anyone know where the AIM solo gets its data versus the dash display. They don't match up well.

[warp10]

Solo pulls off CAN/ECU figures. So I would assume the same as the place the gauge get it?

[regular guy]

Originally Posted by BigBoosting

Sorry man, you are wrong.

The dash gauge is quite accurate. I have had an oil temp gauge and they read the same. Oil temp doesnt fluctuate rapidly so inherently the gauge doesnt have large swings frequently.

I think what he's trying to describe is the gauge has smoothing built in. Whether it's done in the ECU or the gauge itself, he would be right about that.

[regular guy]

Originally Posted by warp10

Solo pulls off CAN/ECU figures. So I would assume the same as the place the gauge get it?

Here's a CAN bus view of the oil temperature. This is how AIM Solo sees it.

On the ECU CAN bus, the oil temperature is specified in degrees-C. That's probably not a surprise to anybody. It's an 8-bit value, which means its actual values span 0-255.

The actual formula conversion is:
Tc = Av - 48, where Tc = Temperature(C), and Av = Actual Value on CAN bus.

That means the oil gauge input values will span from -48 C to 207 C, -54 F to 405 F.

That doesn't mean the gauge in the cockpit will show those values.

[BPMSport]

There is no smoothing built into the oil temp gauge. It's very accurate.

BMW coolant gauges on the other hand, do have to have a large variation in temperature to move left or right of the center mark. This gauge is not linear unlike the oil temp gauge.

[regular guy]

Originally Posted by Mike Benvo

There is no smoothing built into the oil temp gauge. It's very accurate.

Have you ever data logged the output of the sensor itself and compared it to the data on the CAN bus?

Like I said, whether it's in the ECU or the gauge, there absolutely is smoothing. And smoothing has nothing to do with accuracy.

If you care enough, I can look at my data logs, and tell you exactly where the smoothing is done (ECU or gauge).

[BPMSport]

Maybe you can describe what you mean by "smoothing" since what you described in the previous post is merely a correction factor/formula applied to the raw value of the sensor output (converting it to F). This isn't "smoothing".

[regular guy]

Originally Posted by Mike Benvo

Maybe you can describe what you mean by "smoothing" since what you described in the previous post is merely a correction factor/formula applied to the raw value of the sensor output (converting it to F). This isn't "smoothing".

The smoothing definition I'm using here is the same concept and definition as the Dynojet software, vBox software, etc. Smoothing simply means averaging some number of samples so the needle appears "smooth" and doesn't jump around. That's exactly what s85e90 was describing whether he used the correct terminology or not.

The input sensor is usually voltage based (or sometimes current based) and then converted from this analog value to digital. When the analog value after conversion gets very near a temperature boundary, if this data were "unsmoothed" you would see the raw values appear to jump up and down. For example, let's say the value gets near 60, at 10 Hz, you might see those raw values oscillate between 59 and 60 many times before the temperature gets solidly above 60. So instead of allowing that to happen, some number of samples are averaged to "smooth" that out.

I've gone to my logs, and I'd say the MSS60 is the guy doing the smoothing. I did see one slight occurrence of the values oscillating between numbers, but it only happened once. So I would be inclined to think the MSS60 is smoothing these values before passing them on to the CAN bus.

[BPMSport]

Thanks for the information, but I just don't see the relevance to the hotter temperatures observed and don't think the smoothing is a function of this.

The gauge reads accurately. If you log the oil temperature in real-time and watch the gauge, it's pretty much spot on. Yes, the gauge will be ever so slightly slower, but at any given second it's pretty accurate. It's not sampling 15 times a second because it doesn't need to. You can call the slower sample rate smoothing if you'd like, but this is not necessarily the case.

Just FYI because you might be interested, the raw values for some of the map axes in the DME (such as hot oil temperature protection) are actually 16-bit values represented in Kelvin multiplied by 10.

So to convert the map axis into C, you take the raw value in hex, divide it by 10 and subtract 273.2 to convert it from K to C. After doing this you will find that the DME starts reducing the redline of the vehicle after the temperature exceeds 150C. It uses an interpolation function to reduce the redline from 8300 to 7800 between 150 and 151C and continues that as temperatures rise.

[Leonardo629]

Originally Posted by Mike Benvo

There is nothing inaccurate about the oil temp gauge.

I'm glad Mike debunked the oil temp gauge is just a "dummy gauge" myth.

Mike you still running 8900RPM with your new S65?

[BPMSport]

Originally Posted by Leonardo629

I'm glad Mike debunked the oil temp gauge is just a "dummy gauge" myth.

Mike you still running 8900RPM with your new S65?

Supercharged now so back to 8,300. I was debating a raise to 8,600 but it's really fast already.

[s85e90]

Reassuring to know that the aftermarket gauges are reading accurately but as stated the swing is large and with no markers present I don't see it ( the oem)as a reliable reference.

[evilspiritM3]

Originally Posted by surlynkid

This weekend, my wife and I were sharing the M3 at a DE. Weather was 87F with about 75% humidity. Our sessions were not back to back, but the car had to do 8 sessions in the day. On one of her sessions, the high temp light came on with the little thermometer picture and the light that looks like bacon cooking. I pulled the AIM data before going back out. Oil showed 262F max and coolant at 224F max. The needle on the dash was much closer to the 300F mark though. It quickly dropped on a cool down lap.

What oil/water temps are required to get the high temp alarms? Long term, I only need to car to support one driver, but if she wants to run it in BM with BMWCCA, I don't know if it will make it thru a 60 minute enduro.

I realize this is speculation, but how much worse, or shorter would your sessions have been with a stock oil cooler?

[bigjae1976]

Originally Posted by Mike Benvo

Thanks for the information, but I just don't see the relevance to the hotter temperatures observed and don't think the smoothing is a function of this.

The gauge reads accurately. If you log the oil temperature in real-time and watch the gauge, it's pretty much spot on. Yes, the gauge will be ever so slightly slower, but at any given second it's pretty accurate. It's not sampling 15 times a second because it doesn't need to. You can call the slower sample rate smoothing if you'd like, but this is not necessarily the case.

Just FYI because you might be interested, the raw values for some of the map axes in the DME (such as hot oil temperature protection) are actually 16-bit values represented in Kelvin multiplied by 10.

So to convert the map axis into C, you take the raw value in hex, divide it by 10 and subtract 273.2 to convert it from K to C. After doing this you will find that the DME starts reducing the redline of the vehicle after the temperature exceeds 150C. It uses an interpolation function to reduce the redline from 8300 to 7800 between 150 and 151C and continues that as temperatures rise.

On the E46 M3...I believe the coolant gauge was buffered. It would read half once the engine was warm. I've seen it in the gauge cluster unlock function, the temp can vary between 75c to 84c and the needle will not move. I think the coolant gauge will start to climb once the thermostat opens and the temps continue to rise. I had an aftermarket oil temp gauge...it fluctuated a lot more than the OE temp gauge but I wouldn't necessarily say that the OE temp gauge is accurate but its good enough. I guess it depends on what you call accurate. Not sure if the S65 DME changed much from the S54 DME.

Keep in mind these are street cars. Most don't want to really know that coolant and oil temps are.

As far as the oil temps...is there a filter bypass and does it bypass the oil cooler? A bigger oil cooler won't help if the oil is bypassing the filter and cooler...the pump is just recircing uncooled oil which would significantly lessen the effectiveness.

As I said before...blow out your radiator.

If I were turning my E90 into a track car...I'd go dry sump. I don't see how you can effectively control oil and water temps with the space provided. Especially with a DCT and if you plan to use brake ducts (a must have IMO)...then all of your airflow choices are taken.

[BPMSport]

Originally Posted by bigjae1976

On the E46 M3...I believe the coolant gauge was buffered. It would read half once the engine was warm. I've seen it in the gauge cluster unlock function, the temp can vary between 75c to 84c and the needle will not move. I think the coolant gauge will start to climb once the thermostat opens and the temps continue to rise. I had an aftermarket oil temp gauge...it fluctuated a lot more than the OE temp gauge but I wouldn't necessarily say that the OE temp gauge is accurate but its good enough. I guess it depends on what you call accurate. Not sure if the S65 DME changed much from the S54 DME.

Keep in mind these are street cars. Most don't want to really know that coolant and oil temps are.

As far as the oil temps...is there a filter bypass and does it bypass the oil cooler? A bigger oil cooler won't help if the oil is bypassing the filter and cooler...the pump is just recircing uncooled oil which would significantly lessen the effectiveness.

As I said before...blow out your radiator.

If I were turning my E90 into a track car...I'd go dry sump. I don't see how you can effectively control oil and water temps with the space provided. Especially with a DCT and if you plan to use brake ducts (a must have IMO)...then all of your airflow choices are taken.

The coolant gauge is buffered on pretty much all BMW's. Oil temp is a different story. The MSS54 and MSS60 DME's are completely different architecture.

[bigjae1976]

Originally Posted by Mike Benvo

The coolant gauge is buffered on pretty much all BMW's. Oil temp is a different story. The MSS54 and MSS60 DME's are completely different architecture.

So smoothing is the wrong word and buffered is correct. Which means regular guy is WRONG!!!

In your face!

Seriously, maybe the increased oil capacity due to the oil cooler may contribute to increased temps? While you increase the size of the cooler, you also increase the volume fluid that needs to be cooled...I would think that this would end up being not very effective.

Maybe try running the crap e-dipstick on the lower half of the gauge?

[V1.47fan]

Originally Posted by s85e90

People here need to stop measuring exact oil temps from the dash and get a gauge before they analyze the effectiveness.

I think the op is going with the oil temp gauge's before and after delta. That'll give you a good idea of what's goin on, like on colder days it shows lower oil temps, with a better oil cooler, you should see slightly lower oil temps.

[Gossypiboma]

Originally Posted by V1.47fan

I think the op is going with the oil temp gauge's before and after delta. That'll give you a good idea of what's goin on, like on colder days it shows lower oil temps, with a better oil cooler, you should see slightly lower oil temps.

Exactly, its a relative comparison.

The term dummy gauge is usually reserved for one where the needle doesn't move in the least while the parameter being measured varys greatly.

[regular guy]

Originally Posted by bigjae1976

So smoothing is the wrong word and buffered is correct. Which means regular guy is WRONG!!!

In your face!

HAHAHA.

[Hujan]

Originally Posted by bigjae1976

So smoothing is the wrong word and buffered is correct. Which means regular guy is WRONG!!!

In your face!

Seriously, maybe the increased oil capacity due to the oil cooler may contribute to increased temps? While you increase the size of the cooler, you also increase the volume fluid that needs to be cooled...I would think that this would end up being not very effective.

Maybe try running the crap e-dipstick on the lower half of the gauge?

But more fluid also means more heat capacity, just like a bigger/heavier rotor.