Hey guys i just did a little experiment to find out when our temp gauge really starts moving. It is 230f and as a side note above 245 you cant reflash the car. If you have to unplug the coolant temp sensor and it will reflash.

Neat. That's using the OEM ECT, correct?

Yes

Good morning log.
What do you mean by 'reflash'?
I thought that was something only the dealer can do.

Morning Cajun. Cobb ap and mazda edit are also reflash tools. I have a cobb.

OK, thanks.

How did you get your car to overheat? IIRC the FSM states that all the way up to 243F is within spec.

Rob @ Pineapple told me that he has built a lot of dune buggies powered by the renny and older 13B's and he said they run in extreme conditions and that typically OEM coolant seals start to deteriorate if you spend any significant time around 220F.

My last engine died of a failed (well it was caught in a oil sample it didn't actually fail) coolant seal after 30,000 miles and I did often get in the 225F range (both oil and coolant) and spend some time there during our runs thru the canyons during the spring and summer. The runs are typically high RPM, low gear, fairly low speeds.

I know people have pegged their gauges and not had issues but I think overall they were just lucky.

OH and IIRC, the factory temp sensor maxes out at 248F according to the FSM.

It's not installed on the engine.

This is why I got an ultra gauge.

Oh, so he just removed it and heated it up, gotcha ;)

I just tuned it to high temp car was actually about 100f at the time.

You can tune readings from sensors!?!?

Maybe he changed the tune so the fans don't come on?

The iat and maf sensors are tunable on 04 and 05 models with cobb

So what did you actually do? I'm confused now.

:lol: I think he drunk.

Dang man its too early to be drinking.


I told the car no matter what voltage the temp sensor reported it was x tempeture. Then flashed the car. Then took a pic and reflashed it to a higher temp.

200, 225, 250.

The. It would not let me reflash and i got scared and figured it out. So i did

230, 237, and 245

Then i narrowed it down and did some of the finer incriments.

Well that's one way to do it, I guess. Thanks for sharing!

I got to the last photo at the track a couple of years ago.

The oil temp was just under 230 and I was in the heat of the battle on a hot summers day. I usually keep an eye on the oil temp and back off when it gets to 230.

The only signal I was getting from the car was hesitation as I floored the throttle on the back straight of the track. When I went around the next lap it did the same and it is then I noticed that the water temp gauge had moved I backed off for a lap and it immediately cooled down however, it returned again after a couple of laps of hard driving.

Car has been running fine after that episode.

I'm assuming this is the map you played with.... very tempting since I monitor water temp through my adaptronic that has its own ADC table.

https://cimg8.ibsrv.net/gimg/www.rx8...48a956fe41.png

so a little bit of anality here.

Sensor theoretically goes up to 284F.

To have the gauge jump sooner, rescale .620v through .224v. If you want it to jump like mad, simply set all cells from .620v on to 140. With a little bit of interpolation you could have it rise in a visible linear manner.

It's literally as simple as making the V vs T graph into a straight line from top left to bottom right.

so interpolate between the endpoints?

Like so?

https://cimg2.ibsrv.net/gimg/www.rx8...f8c0b5ff00.png

That doesn't look right to me. Seems like it'd be even worse.

It was right, except why do you care if the gauge moves below a certain minimum temp and above a certain maximum temp?

You could rescale the whole thing accordingly.

that's what I meant. some would want 180 as the mid point... others 200.

To my taste I'm thinking leaving cold scale alone up to around 160F... then 180F as midpoint with the gauge maxing out at 220F.

Problem is -- if I use poo's data as a reference point we're talking half the swing of the gauge over the course of like 4 cells of the map.

Try it and find out. You won't hurt anything if it's wrong.

I'm working out the lopsided math for it now. Will post a screenshot of what I plan to flash here shortly.

Problem is I'm "studying" for my emissions test so I don't dare flash it yet. I've got one last test to pass before I go in for inspection.

That would work just fine. Or you can just tweak it enough to skip the dead spot in the gague.


the way I see it. The computer uses it to control a few things with load and possiably other things, but by the time I'm over 220F I don't really care if it is calculating them right. I would rather have it alert me of the overheating.

I have noticed that the AC turns off at a Overheating temp. I am not sure exactly what it is. That makes it really noticeable when I start getting into the upper temps.

so for me the only thing I may affect is AC cutoff as every other function that may be affected is controlled by my Adaptronic and it has its own ADC scale for ECT.

Very cool. I'm surprised we haven't covered this in more detail in one of the AP or mazdaedit threads... certainly is another benefit of having the software.

Yeah, With the Adaptronic It would be nice to make it a nice sweaping scale. The thing is you'd have to determine the temp it starts at and the temp it finishes at. you could spoof it with a 0-5VDC signal and tweak it to determine exactly what voltage gives what on the scale. Then adjust it to sweep correctly. I think interpolating between the sensors endpoints is too extreme. You need to interpolate between the gauges endpoints.

and you've already given us that data... at least from 50% to 100%.

Just watch it after your next cold start and see what temp it starts moving at. Then scale from there to 250

For your leaking hose/clamp issue you might try Gator Brand clamps from Comp Cams. I buy them from Summit Racing.

First mark = 125
Second mark = 130
Fifth mark = 140
Seventh mark = 150
Ninth mark = 160

All horseshoes and hand grenades.

This should give me what I'm looking for. I rescaled it so that 220F will report as 250F and then I linearly scaled everything between it and 130F.


https://cimg9.ibsrv.net/gimg/www.rx8...065b2fc744.png

Looks pretty good, to me. Just to note for anyone not familiar you are using the Adaptronic ecu to control the motor so those changes do no affect it at all.

Once I get a little issue fixed with the car I'll rescale the gauge and post it for a system that's still using the stock ECU.

So I flashed this update tonight.... going to have to go out for a drive to get the temp above 160F... sitting right below the halfway mark on the gauge.

Silly thing kicked the fans on right at 160 (makes sense because I already had them lowered) and it's too cold here for it to get any warmer... as a matter of fact it dropped down to 158 after a minute or two of them running.

Will report back shortly... time to build up some heat.

==== edit ====

I'm measuring actual temp with my Adaptronic. My 8's ECU thinks ECT is 190F

I'm working on the ECT scale to make my temp gauge read accurately on the top end of the scale. First I Decided to make my scale max out at 254F. instead of the 280 something that it was. I figure over 220 I don't really care how high it actually goes.

I am going to scale it from my needle straight up at 210F and almost maxed out at 220F.

Here is a scale I made from 201-254F to interpolate the temps per specific Volts.


https://cimg1.ibsrv.net/gimg/www.rx8...f5c019e6d8.png

234F is the needle straight up and 250 is almost maxed out. So I will make 234 line up with the volts for 210 and 250 will line up with the volts for 220. Then slide the whole scale over and dropped the -40 column since I'll never see that low of temps.


https://cimg2.ibsrv.net/gimg/www.rx8...a692f9f015.png

won't this make the car's ECM think its hotter then it really is?

I would be worried its possible it would alter the tune somewhere else, fueling or timing

It does make it think it's more hot, but it doesn't use that information for anything other than open loop startup. The only temperature that's factored into other things is from the IAT sensor.

Well I didn't really change anything below 210. So the car has to be at 210 or above. At those temps I don't plan on driving it.

So I guess I can chime in with my own questions about this gauge and hopefully you guys can help or give some incite on my engine and the coolant gauge..

I typically sit in the car for 30 seconds to 1 minute warming up before I drive, even then the gauge never seems to move off of the C all the way on the left.

But while I am driving, my gauge never seems to get even to the middle between C and H, it always sits a tad to the left of the middle at say 40-45% or so. Is that normal that it never gets that hot? Hearing that heat is the death of these motors maybe I should consider myself lucky then?

This past weekend my wife and I went for a jaunt up throguh the Rocky Mountains here in Colorado on some of my favorite motorcycle twisties. We had the car driven for maybe 80 round trip miles from start to finish, without any cooldown periods or turning the car off in between, so roughly an hour and a half of the car being on. We were booking it around in lower gears, almost always over 5K RPM, twisting and turning through the mountains and then drove home. The entire time we drove around and as long as we drove, the needle still never got to the 50% mark between C and H.

Is this normal or should I just consider myself lucky? I have a 2004 GT with the original engine and just shy of 82K miles.

Reason I ask is because the pics from the OP show his car warming up and going all the way to the H side. I have never seen it hit 50% even on that long aggressive canyon ride.

If you look at the pics at the top of this post you'll see what temp your car is actually at compared to needle position. If your needle is straight up that is 230F which is well into overheating.

Ah totally looked at it wrong. It's weird that anything borderline towards the middle is between 199-234 per that first post. You'd think there would be a wider range from 199 to overheating as far as the gauge goes.

I just got an ultragauge this week and set it to alert me if I hit 217 for coolant temp so hopefully it doesnt get up there that hot if I'm at the 40-45% mark on the needle.

It is very weird that is flat lines and just sits there. Then starts moving. I'm guessing mazda thinks 230-240 is the start of engine damage.

The common forum belief is 220 is the start of engine damage. If you look at post 41. That is an explanation of how to make the display max out at 220F, but you need a flash tuner to do that with.

I received my Ultragauge last week and finally went for my first drive today since getting it as I've been busy. I was stuck in traffic so I couldnt go all that fast but 6the Engine temp was sitting at 212 degrees F in traffic. Not knowing how to actually guage what is good/bad what are the best temps to look for when reading the meter.

I have it set to engine F and intake F on the first page, Honestly not sure whats the right range to be in. A post above says forum belief is 220 is the start of engine damage and I was sitting at 212 for the longest time, then again not moving very fast maybe 10-15mph in traffic.

212 is a good high point / peak.