Finally got some work done on the FC like, new oil and a Mazda oil filter, but more importantly about the problem. I unplugged the IAT and the CEL went on, so that works, turned the car off and did the self-test, got a long blink and a short blink so that's 11 to my understanding, which would be IAT. Now all I need is to drive the car and break down so I can do the test then.

You might be right, I do think it is the wiring, I also have a wire issue with the warning cluster, it keeps resetting and sometimes just stays off.

The warning cluster problem (i.e., clock resets itself to 1:00 or stays off) is most likely just the cold solder joint issues common to most FC electronic accessories - no relation to your engine problem, but it be could be a factor in why the CEL isn't always displaying errors.

Simple fix involves removing the cluster and resoldering the larger solder connections between the circuit board and the one connector plug. I did this when I first bought my '7 (over 10 years ago) and haven't seen that problem since. There's was a good "how to" article for this procedure with pictures on the web, but the link I had to it on my PC now returns a "page not found".

When I check the CEL codes the cluster is on, also thank you for advice I shall fix it asap

Well, the car finally broke down, didn't throw any codes and CEL didn't come on. Idles rough around 600rpm. It seems to happen when it's warm outside. Can it be TPS related?

It could be an intermittent vacuum leak for the rough idle, but if it is coupled with the inability to rev, it almost has to be limp mode.

This is puzzling.

I tried unplugging the TPS on my '91 and revving the engine and it revs just fine. I also tried without the MAP sensor - same results. It almost has to be an Oil Metering Pump feedback issue (I don't remember what all codes can set the limp home mode).

Did shutting the key off and turning it back on fix the issue this time too?

I wonder if the CEL is not setting because the fault is immediately clearing, but the limp mode is latched on.

Key method seemed to fix it again. Is there any way to check if its the OMP while the problem is happening? Check wires for resistance and continuity sort of thing.

Best bet is try unplugging the OMP and see if you get the same behavior consistantly (don't let it run too long with the pump unplugged) You could try cleaning up all the connection points between the ECU and the OMP.

I think that if I remember correctly, there is a motor in there to adjust the position and a potentiometer to indicate position back to the ECU (closed loop control) If the resistance does not react as expected, you can get the code. You could have corrosion on the wipe contacts of the potentiometer too (a 'dead spot').

I don't know how to tell you to test that.

You could always try a different OMP, I have one, but it hasn't been on a running engine in years.

My friend has an OMP I can use, he has been premixing for a year now. Guess I'll try unplugging the OMP for a short time, It wont take long to notice the problem. If same things happens I'll check wiring and if that doesn't help ill change out the pump. Should get back with the results this week.

Talked to a good a friend of mine, he believed it is not the OMP because it would throw a code and it will only cause limp mode after a while, mine starts right away. We think it is the MAF, since he had similar problem and his MAF was the source, he also didnt have any codes when the MAF went bad. Time to check MAF wiring.

Not a bad idea, keep us posted!

Your friend is wrong regarding the OMP. When my old wiring harness was going flaky, I discovered that even intermittent open circuit associated with the OMP wiring, either the OMP stepper motor or the OMP position feedback sensor, would both trigger the CEL and force the limp-home mode every time.

Looked and my MAF plug today, wires where under pretty high stress going into the plug, fixed that, cleaned both sides of the connector, need to buy Volt Meter thing sometime soon so I can check the wires internally.

He might have been wrong, but I doubt it. And I don't get CEL, so I don't think its the OMP.