I'm having trouble tuning my megasquirt and it looks like the problem is with the stock MAT sensor. I am getting 18 degree spikes in 1/10 of a second plus the readings seem very high once it is warmed up. I have easythermed the values and they look good when the engine is cold but they freak out when it gets warmed up.

; Generated using EasyTherm4.exe,
;
; Air Temp Thermistor - Bosch 0 280 130 006 - Tested
;
; File generated for use with stock 2490 ohm resistor at R4
;
; Steinhart-Hart coefficients: A= 1.453347E-03 B=3.570061E-04 C= -5.084427E-07
;
; Input Data: Temp, degF Resistance
; 32 729
; 70 303
; 212 33

For those using the stock sensor, how do your logs look? Here is a sample of one of mine, sorry about the alignment. The MAT column is the one that starts with 197.

Time SecL RPM MAP TP O2 MAT CLT Engine Gego Gair Gwarm Gbaro Gammae TPSacc Gve PW Gve2 PW2 DutyCycle1 DutyCycle2 EGT Spark Angle Knock

2892.060 22 23 76 22 3.980 197.0 215.0 1 103 80 100 103 82 100 55 6.0 50 6.5 11.5 24.9 558 9 0
2892.116 22 22 78 22 3.784 197.0 215.0 1 103 80 100 103 82 100 59 5.7 50 6.0 10.4 22.0 522 9 0
2892.172 22 22 80 22 3.588 197.0 215.0 1 103 83 100 103 85 100 56 6.4 50 6.5 11.7 23.8 515 9 0
2892.227 22 23 76 22 3.843 197.0 215.0 1 103 83 100 103 85 100 57 6.7 50 6.2 12.8 23.8 536 9 0
2892.283 23 23 79 23 3.765 179.0 215.0 1 104 83 100 103 85 100 57 5.9 50 6.4 11.3 24.5 529 9 0
2892.339 23 22 79 23 3.745 179.0 215.0 1 104 78 100 103 80 100 56 6.0 50 6.3 11.0 23.1 522 9 0
2892.394 23 23 76 22 3.647 179.0 215.0 1 104 83 100 103 82 100 60 6.3 50 5.6 12.1 21.5 529 9 0
2892.450 23 23 78 22 3.745 215.0 215.0 1 104 78 100 103 82 100 51 6.4 50 6.4 12.3 24.5 536 9 0
2892.505 23 23 78 22 3.980 197.0 215.0 1 104 80 100 103 82 100 57 5.9 50 5.8 11.3 22.2 579 9 0
2892.560 23 22 76 23 4.000 197.0 215.0 1 104 80 100 103 85 100 58 5.3 50 6.3 9.7 23.1 565 9 0

The spikes are probably just noise. For something like MAT, I'd not worry too much about that. 18dF isn't enough to alter the fuel mix by enough to matter.

Your temps are indeed very high. I'd expect to see numbers in the 45-50dC range (110-120dF) on a 914. Did you measure the sensor yourself, or are you using some table for that part number? Also, are you sure you're using the correct table in MT? I'm pretty sure MT receives "raw" numbers and converts them at its end.

If you haven't already, ask this question on the MSEFI forum. You're likely to get a better answer there. The author of MT is there regularly, and usually has an answer for odd logged data.

I don't have my files here at work but, there is no way your MAT should be that high. I suspect you either have a bad .inc file, the sensor is bad or the wiring is bad. I don't think I have seen MAT temps over about 130. That may be the reason that the EGO voltage looks so lean.

The flucuation may be megatune reporting sensor failure. I see that all the time with the coolant sensor but since it reports over 160 degrees it doesn't affect anything. Unfortunately, MAT figures directly into the fuel calculation and the higher the temp the less fuel you get.

Check this out on the MS board. This sounds kind of similar. Maybe it's a bug in the program. I have felt that the air density correction is a problem for MS from the beginning.

Dave

I thought it didn't look right. I'll check to make sure I have the modified inc files in the right directory tonight. I did check the MAT sensor values myself for easytherm. I didn't notice any problems during testing but it is a EBay part, maybe it's bad. I'll check the resistance when it starts flaking out to see if it is the sensor or if it is noise. If I don't find a problem there I will try and reburn the MAT/CLT tables.

I'll post over at MSEFI once I do a little checking. I'm new to MS and I just wanted to confirm that this was not normal before I go looking for the problem.

Thanks, you guys gave me a lot to go on.

P.S.

I've been looking at Brad Ander's site on how the stock d-jet system works. From reading his site it looks like MAT doesn't play a big role in fueling but CHT plays a huge role during warm up.

MAT doesn't play a huge role in fuelling, for MS or D-Jet. The 18dF difference you see in the spikes will result in almost no change to fuelling. The 80dF change you see from reading 200dF where it should be 120dF will make a bigger difference, but still less than the usual "coolant" temp range makes for warm-up enrichment. Cold starting requires a really rich mixture, so there's lots of "authority" to add fuel there based on temp changes, but only on the rich side. High MAT will lean out the mixture. From memory, I think MS will usually only change the fuelling by a maximum of 7% based on MAT.

You can get some cheap thermistor beads at Radio Shack with the temp/resistance scale printed on the package to back up any measurements you make with the MAT sensor you're using. You also see people sometimes post on MSEFI with MAT problems caused by heat-soaking of the sensor. MAT can actually go UP after shutting off the engine. I think this may not be as big an issue on an air-cooled engine as it is on a water-cooled engine, however.

I think the MAT is suffering from heat soak. The temp slowly rise until it peaks. I will work with it to see if I can get a better reading. My engine is not sealed to the body due to A/C so that may be coming into play. Head and oil temps look ok however.

I took a close look at the matfactor.inc and saw that at the upper end of the table, the temps are spread pretty far apart 169, 179, 196, 221... This accounts for the notchness.

I did find out what my major tuning problem was. It was a bit of a riddle. I had to remove the tank to change out the carb fuel pump. I put in the new pump,changed the fuel lines, replaced the tank and put a couple gallons of gas in it and started it on my first try. Everything was running pretty good for not being tuned. So it was time to start road testing/tuning. I went and got some more gas, filled the tank and Wham-oh, it begins to run like crapola. I check this and that and find that I have to change my ve's from 30's to 90's just to get it to idle. I put a fuel gauge on it and saw the pressure jumping around from 15 to 30. I already had problems with the fuel regulator so I grap one from an old CB performance turbo kit and it does the same thing (after I fixed the massive leak it had). Weird! Bad fuel pump all of a sudden?

I think about it for a while and then it occurs to be. The fuel pressure was fine when there was only a small amount of gas in the tank but when I filled it up I had problems. The only difference is that the tank is heavier. The tank must be resting on the fuel line! Sure enough I lift the tank up a little and the fuel pressure levels off at 50 psi (Turbo regulator setting). I check the lines under the tank, rearrange them and drop the tank back down, change back to the stock regulator and now it runs like it should.

There is alway something...

First, if you are running 50psi through the stock injectors get it down to 30psi then leave it there.

Second, the TPS has nothing to do with heat. It measures throttle plate opening and only is involved in acceleration enrichment in speed density. In tuning the last thing you add in is AE so, don't even worry about the TPS yet. My car will run VERY well without the TPS connected at all.

Third, if you fooled with the tank check the fuel filter. you may be surprised.

Fouth, the MAT is an issue that you need to fix. James is right when he says it plays a relatively small part in the fuel calcs. But it is significant when you get to finer tuning. If the air density is off you will be like a cat chasing it's tail. You will constantly be retuning cause the car will run rich or lean for no real apparent reason until you look at air temps. Been there.

Fifth, heat soak is a big issue with the stock MAT in the stock position. When you do a warm restart wait a minute before driving cause your car will be lean. It doesn't take long for it to cool but remember.

Sorry cause this kind of looks like a lecture. I don't mean for it to be critical. This is stuff I learned trying to make my MS work

Dave

My bad, I wrote TPS but meant MAT. I'll edit the other post.

I put the stock regulator back on set at 28 psi. New fuel pump and filter and hoses throughout. Even flushed the pastic lines since the return line was exposed to the elements fro some time. Quite a bit of gunk came out.

I have a GM MAT, I wonder if I would get better results if I retrofit it in the air cleaner.

I used your site quite a bit to get me started, all the info helps!

I have all my ET info on another computer also...I did just use the stock one and stuck it under the passenger seat. make sure that you have a good ground. mine is a 2 wire sensor. duh, I think yours is a one wire. Do you have the stock sensor? try it. Mine is close, not perfect, but it runs ok.

I will try to dig up the stuff from my other computer and post it...

I've got a 2 wire sensor. I'm kind of stuck with the location since I build a harness for the engine. I probably have around 4 inches of travel in the wires, maybe enough to put it in the air cleaner, I'm not sure.

What I would like to do is drill out the stock sensor and just hang a thermistor out of the bottom so it is not touching any metal. Unfortantely Fry's and Radio Shack don't carry them. I miss the old days were radio shack was all about electronic components instead of phones! I'm not sure I could save the connect on the stock sensor and re attach it but I could probably make a pigtail to it.

Ha! Guess what I just found!? A thermistor I think, I don't have a vom to check here at work but I can't imagine what else it could be. It use to be attached to one of those indoor/outdoor thermometers in our server room but it had been hacked off (I swear I didn't do it!). This should do nicely!

QUOTE (agrump @ Aug 19 2005, 06:18 AM)

I have a GM MAT, I wonder if I would get better results if I retrofit it in the air cleaner.

I have my Air Temp Sensor in between the air cleaner and the throttle body.

what you could also do it bypass the MAT sensor for testing and just run a resistor with the correct Ohms to emulate ambient temperature...maybe go with a rating just above ambient to better mimic the hot engine bay.


here is neat page showing some FI stuff and explains how to trick/modify an ECU, some of it can be applied to the MS for testing.......FI tricks

Just an update, I ended up cutting out the stock MAT bung and welding in one to fit a GM open cage sensor. After reprogramming the megasquirt to match the sensors my MAT readings are good. I am still seeing 179 degree temps when idling for a long time but that was on a 95 degree day. Once I start moving the temp readings drop which they did not do using the stock bosch sensor. I'm going to work on my engine tin seals to see if I can improve the AC opening to lower the MAT temps.

Tuning is going pretty good.

Dean