I remembered wrong. It was only 95%. This is MS1.

And forget everything else I said because my XD-16 is on the fritz and showing vastly different afr numbers than what is showing up in TS coming from the same LC-1. TS shows everything in the normal range when hot in boost. Sorry for all of the trouble.

Now I need to know why the xd-16 is only reading single digit values when the lc-1 is putting out good ones to the MS1. I thought the number one was burned out on the display because it was reading "3.7" or "4.5" but it isn't a burned out number. But that is a different search. Thanks guys.

That's really odd, especially for an XD-16. Unlike the DB gauges that uses a serial interconnect instead of analog.

First things first, make sure you have the terminator plug in place. It should look like a black headphone jack with nothing connected to it. If it's not connected chances are the gauge it's going to read right, but the megasquirt will get a good read since it's hooked up over analog.

Second thing, make sure the XD-16 isn't trying to display in lambda or some other funky output mode. You'll have to hook up to the gauge via serial with Innovate's configuration program.

The other thing to try is to hook into the LC-1 via serial and read the value with LogWorks. That's the same interface the XD-16 uses, so if it's not working the problem is the LC-1 and not the gauge.

After researching, my XD-16 needs to be reprogrammed. It must have reset itself somehow when the voltage was low during cranking. It is apparently displaying lambda instead of afr. :vash: The signals going to the MS1 in boost are 12.2 afr like they are supposed to be and my gauge reads ~3.something. I just thought the first digit was burned out on the gauge. My car never went lean in boost. Just ignore me and I'll sit quietly over here.



Now I require a stereo 2.5 to db9 cable that didn't come with my used lc-1/xd-16 :vash: :vash: :vash:

Sorry to bother you guys unnecessarily.

Mine did that too randomly at some point. Once I got it hooked up it had some weird custom calibration set...was strange.

MS1 Gair madness
 

I don't mean to thread-jack here, but I'm having trouble how to apply Brain's suggestions on my MS1 MSPNP. I just now noticed that I'm getting some *crazy* Gair going on, ranging from 88% to 152% on a pretty typical drive on a 65 degree (F) day. It's swamping all my other adjustments, like so:

https://www.miataturbo.net/attachmen...1&d=1342069603

Here's my "coolant corrected air density" page - I haven't touched it from the base map. (It's set on "IAT Corrected.")

https://www.miataturbo.net/attachmen...1&d=1342069603

What do I need to do to disable this IAT correction entirely?

don't. on your map youre adding 39% more fuel than your ve fuel table is asking for, yet youre still at 17.25:1 afr.

This may not be exactly relevant to the problem you guys have, but what some OE's do is to model actual charge temperature *at the intake valves* (which is what matters), by assuming it's between the IAT value, and the CLT value. The idea is that the air *heatsoaks in the intake manifold*. Some OE's use a 3rd temp sensor, on the metal of the intake manifold. The lower the motor's airflow, the more the air heatsoaks on the way in, and the closer the modeled air temp is to the CLT instead of to the IAT. At heavy load the air temp may be very close to the IAT instead.

What they do is to have a table with airflow as the X-axis (which is proportional to power, which is proportional to MAP x RPM x VE), and then the Y-axis is a percentage. This percentage is used to multiply to the difference between the measured IAT and CLT, then added to the IAT value. A value of 0% means the modeled air temp will be equal to the IAT. A value of 100% means the modeled air temp will be equal to the CLT value. A value of 50% means the modeled air temp will be halfway between them. The table will start out at 100% at 0 airflow, quickly tapering off to almost 0%

The Autronic BTW has this feature. I find that the Autronic software is elegantly designed to model engine behavior properly. It is instructive to study its tuning software setup. The tuning software in MS in contrast seems to be more designed by code jockeys. (In the same manner a lot of chassis dyno software is designed by code jockeys instead of combustion engineers or ECU tuners/calibrators)

Reverant, is this what your special IAT code does, or does it otherwise resemble this?

This method will prevent excessive richness when the airflow is low and the IAT reports low temps. It works well when the IAT reports an accurate temperature of the air. When the IAT is heatsoaked from plumbing after a hot restart, thus reporting a higher temperature than the actual air temp, then you have the opposite problem.


BTW the above is one example of why I think there has to be an "airflow" variable in the software. It should be proportional to MAPxRPMxVE. And there should be two 2D VE multiplier lookup tables. One with RPM as the X axis, and one with MAP as the X axis. The actual VE is now the product of the 2 VE multipliers. This method more models actual engine behavior pretty well. When you have these 2 tables, the 3D VE table that you still have, will be pretty close to flat all over.

Brain, wouldnt I be better off removing all this "correction" and re-tuning my base map? All this Gair adjustment is happening at moderate air temperatures (65 deg f) with a warm engine.

No, since the problem with Miatas is that the sensor heatsoaks. The GM sensor is unfortunately very prone to heatsoaking due to the metal body.

see post 109

Where does GM mount these IAT sensors?

I used to think that too. I swapped mine for a plastic Mazda NB IAT sensor and suffered the same heatsoak issues. If there is too much heat, the sensor will just soak it up regardless of the material it is made from. I think the brass metal sensor conducts and dissipates heat fairly well.

G

the GM sensor is actually a plastic body that's just inserted/molded/bonded to that brass sitting.


I think OEMs just do a better job of dealing with the temp deltas.

For one thing they use MAF sensors, not MAP (speed density) systems.

In a speed density system the IAT is a critical part of the mass flow calculation.

In a system that uses a MAF sensor, the airflow is measured directly, and the MAF sensor is used for retarding timing to prevent ping.