Kinda looks like the IAT (red) log might be richer ............ but yeah , not enough info to call it .

 

I was just looking at my logs and comparing timing vs ECT and coming to post a Q about it when I found this thread... What I have found is that timing gets pulled at higher ECT. According to my logs, at 203F and 214F ECT, no timing is pulled. At 223, 228 and 230F it is pulling 2 degrees.

I found this, which might help you. However, in my case I am under 40C (104F) intake temps the whole time. So, I wonder if there are some columns missing in MazdaEdit for higher coolant temps? Note that I have not yet tried editing this table to see if it has any effect...

 

There is another thread here that mentions this. There is a maximum ECT over which the ECU will not accept a flash! You found the solution - unplug the ect sensor.

 

I am not seeing a noticeable change in AFR, but I have not looked at much data for this either

 

I see that one in the series 2 maps but can't find it in the series 1 maps
I'll email epifan and see if he is willing to help.

 

woot woot

 

Let us know if you get anywhere, and if you could also ask about additional columns in the s2 map I'd appreciate it... in the meantime, I'll try and find time to do a couple of experiments using the map I have.

 

I wonder if this can be used outside of a tune; it may be an idea for a nitrous car to have a resistor on a switch that raises the IAT sensor reading right before the nitrous kicks in another stage.

 

I was thinking something similar to the guys who can manipulate the sensor cal they could just raise all the intake temps to something more real world. but when it boost the intake temps decline so probably wouldn't work. I like your idea, maybe you could tie it to a cobb switch so when you hit atmosphere the twitch triggers the resister to get the IAT up.

my GM IAT mod is nearly complete. I just need to do the wiring then recal the tune.

 

He got back to me saying there was no similar map for S1 then asked for more info to make it easier for him to find . Have not heard back since

 

Glad you got it figured out. Cobb had the same issue. I think it's a ecu thing.

 

Holy thread revival. I'm looking in ME for that holy IAT timing retard in a S1 tune and I found/guessed this:


Not tested yet; looking for more tables to test on a mule car then will return with findings.

Anyone want to try find out theirs, here's my XML:
<map name="IAT based Timing Retard" type="2" help="??? looks like some hot temp protection|||" class="High temp protections">
<rows count="11" offset="#7B6AC" storagetype="float" format="%.0f" metric="deg C" caption="IAT"/>
<data offset="#7B6D8" func_2val="[x]/2-50" func_val2="[x]*2+100" format="%.1f" metric="BTDC" caption="Ignition Retard "/>
</map>

Offsets are NOT cross-calibration compatible sadly, but you can search for this sequence of numbers: 100 80 60 as 1 byte long integers/decimals. All the tunes I have on my PC also have this sequence.

 

Doesn't seem to tie in with my actual results ...see page 1 of this thread as it will give you clues on what to look for on the map.

 

Bit of discussion about this lately so ...even though I've kinda done this test on page one previously ...I felt the need to do it again
Test car: 2003 6MT N/A mostly stock with minor porting . Intake stock . Ambient temp 29C . Engine fully heat soaked.

Method : 1/ Get intake normalised by driving for about 1 mile ,
2/do full WOT run in 2nd gear from 2500 to 7000 and log. (didn't take to redline due to road conditions)
3/Change IAT sensor calibration with ECU reflash (all flashes had closed loop disabled so no fuel trims could be made)
4/ repeat
Red is IAT at stock setting , blue is IAT values increased by 20% , green is IAT values reduced by 20%





Notes
Actual intake air temp after heat soaking is around 6 degrees above ambient as seen in the red 'stock' setting .
load readings mirror what you see above with AFR & maf readings.

Results :
I think this proves pretty conclusively that the IAT value reported to the ECU in open loop has NO effect on maf readings / load or more importantly ......... AFR.

Note: I did the same with idle after each wot run, to test if the same is true for closed loop but unfortunately left A/C on which sent rpms up and down too much to see a definitive result. Need to redo that test but will have to wait till temps go down a bit.
Edit : Tried again with AC off but this time the cooling fan was cycling so that needs to be taken out of the equation as well to get something definitive.

 

Similar test to above showing idle (closed loop disabled)
Red stock calibration , Green less 10 deg , Blue plus 10 deg



Unfortunately I don't have screens in the intake which might be why maf sensor is quite erratic.

Averages : Red 5.0 , Green 5.0 , Blue 5.0
So I'm leaning toward no correction from IAT at idle.

 

Another test done attempting to maintain 55km/hr (closed loop disabled)

Red stock, Blue plus 10 deg , green minus 10 deg.
Averages Red 14.3 / Blue 14.2 / Green 14.4





Cruise range under 20g/s : Does not appear to be any influence from IAT here either.

 

very nice results. however part of it still stays open:
does the ecu take MAF reading without IAT adjustment in open loop but it does in closed loop - this could explain since your VE vs uncorrected MAF is bang on (i would assume) at your normal operating temps (not much change between 20 and 30degC) and you were also in a constant ambient so the ECU it didn't need to correct it, and it corrects it only dooring closed loop as it is expected that IAT doesn't vary that fast
does the ecu have an internal "not mapped" conversion table that doesn't use the IAT sensor calibration table (I have often found this that one sensor has two or even more calibration tables in my WinOLS days..) this is also since the IAT vs cooling of sensor is not really simple math like other parameters so could be "deep" in the code.

you could test both by adding a resistor or voltage splitter to the IAT sensor and not changing the calibration in the ecu.

 

This is what Ciprian plans to do I believe so I'll leave that to him. I wanted to do the test I did for myself because I was having doubts after hearing what Ciprian went through. Plus..... I would be very surprised if there are maps in the ECU using the sensor voltage to make adjustments without going through the sensor calibration map as well.
I believe I found the culprit for my issue at the track recently (dirty sensor) on my setup with the iat sensor separated. Pretty sure I've had that happen before as well. Other than that I've never seen issues like Ciprian has been reporting in the maybe 10 years since I seperated the IAT sensor from the maf.

 

Another test ....

First 5 seconds : steady air sensor temp, air flowing through maf tube.
5-19 seconds : air temp sensor heating up to it's max. (in a separate airstream) , air flowing through maf tube
19 seconds on : air temp sensor cooling down, no air flowing through maf tube

 

Please explain if the maf unit itself had a working and connected IAT element to both ECU(usual connection, OEM stuff) and MAF unit in the 0-20000 x axis points. Then repeat the same experiment, but with the iat disconnected from the ECU(and true, hotter and hotter air being pumped through at constant flow). I'm too lazy to build such a setup with a hair drier or something.

There's my far-fetched-but-dunno-how-far-fetched theory of mine that the maf unit uses its built in IAT sensor to output a constant maf voltage at constat flow, regardless of air temp. Disconnect the built-in IAT sensor from the ECU, and that compensation no longer works.

 

These two parts are counter indicatory. It either corrected before the voltage is output from MAF unit, and ECU receives an "invariable" curve of MAF vs Voltage,
Or the voltage output is dependent on IAT and the ECU needs IAT in order to compensate the curve

 

That's why for best results you'd delete the ecu entirely from the test setup. You'd need to bias the IAT NTC element(I think it's 2.2kOhms in the ecu but can't tell 4 sure, have to look it up), at 5V, then feed the maf 12V and take notes of the maf voltage output. Too much time churned on that for me to do it, at least when I know that leaving the stock OEM setup works fine 4 me, and I don't have high IATs post-intercooler.