Cranks, fires then dies after rebuild

Look who is out of the garage!! Ran a 15 to 20 min varying throttle from 2000 to 2500 for the rings. Now fluid check, cool down and I'll take it for a nice easy drive of long pulls (according to the break procedure post on here)

I wanted to upgrade to 51R and battery tray so bought a battery. Slepped it in. Same results.

On a whim....a whim, I swapped the new mass air flow and and tada. It's alive!

What a weird process and thank you to everyone. When in doubt MAF it out I guess.

 

Did you never unplug the maf and try to start?

 

I did and repeated it peridocally as i checked various things! That's what's crazy. It makes absolutely no sense whatsoever.

Looking at the data related to alternator and vpwr, starts at 12.3v and it stays strong at about 11v cranking, immediately returns when fires and alt output is the 14v.

 

Finally got back out to the car. I got a code of p2187. I reviewed the other posts on the forum. I did the suck blow test on the purge valve. It was not stuck open but didnt open either. I pulled it and cleaned using maf cleaner as suggested. Ordered a dorman replacement. Will be here today.

While I was waiting I got to thinking about the maf I put in. The maf I purchased from Amazon A Premium MAF01130D-C, R112428167, which turns out to be a replacement for L32113215 https://a.co/d/6Hehqwf
was showing for a regular 3 and not an ms3. So, I got a oe replacement denso from autozone #197-6040. Here is where it gets weird. The original crank and die problem came back when using this denso. Had a fault p113 for IAT. So ran a log, swapped the Amazon maf back in, and car fires up and runs. Cleared code and didn't comeback. I checked the IAT from both logs and they are the same. For the Amazon maf What I did notice was the O2S12.STFT.OBD stayed 99.2% and didn't move, SHRTFT1.OBDII(%) averaged 23.11 (extremely lean correct?). For the denso the code p0113 instantly comes on, the STFT was still 99.2%, the SHRTFT1 was 0 and didn't change with the denso replacement because the engine didnt run for long. Maybe it's bad too?

I took a couple of throttle blips and the AFR() and AFR_ACT changed.
My question for this is, with Forscan, should I be looking for change in the AFR data or the SHRTFT data?

But this maf sensor thing has me all messed up now. I wiggled the wires while running, no change, did a bubble test and listened for any air leaks, nothing noted . Disconnected the maf and attempted to start (again lol), no change.

Both the below numbers come up in the mazda system as being for a speed 3 but the one for my vin is the l3k9132159u and not the L32113215 (Amazon replacement)

Here are my thoughts and tell me if I'm an idiot before I go down a rabbit hole. Keep it simple right?
The denso sets a code immediately, this is a bad unit out of the box. Needs to be returned and replaced.
The Amazon, is the incorrect one for the car based on my vin. Need to get the oe replacement.
I can't make a determination on the 02 sensor until I get a healthy running car.
I do have an NTK o2 sensor, but don't want to just throw parts at it until I can figure out the maf thing.

Sorry for the novel and hope you can understand the ramblings of an old man.

 

When swapping MAFs like that, you *NEED* to clear trims by disconnecting the battery.

 

Disconnected the battery. Left it disconnected while I swapped in the denso unit plus an additional 60 secs. The car cranks, fires and dies again.

I've got be missing something here.

 

Get a voltage tester and check the wires to the maf for continuity etc

 

Will do. I won't be able to get back to it until Wednesday. I'll shoot wires and report back.

I have been working through the troubleshooting, obd and controls chapters in Sec 1. I will tell anyone who is not a metal member to become one. These manuals are a must have.

 

I couldn't wait until tomorrow. It's driving me mad. Maybe something, see below.

Open circuit check (continuity should be yes)
Maf terminal A to main relay: 0.0ohm. Good


Maf terminal B and PCM terminal 1P: 0.0 ohm. Good

Maf terminal C and PCM terminal 1AK: 0.0 ohm. Good

Maf terminal D and PCM terminal 1M: 0.0 ohm. Good

Maf sensor terminal E and PCM terminal 1AR: 0.0 ohm. Good

Short circuit check: (should not have continuity)
Maf terminal A and body ground:
33 ohm and beep on the meter for continuity.
There should be no continuity between the above. I read that lower resistance for automotive wires is more indicative of a short and 33 ohms would indicate no continuity.
Is this accurate, or have I found a problem?

Maf terminal B and power supply: no continuity. Good

Maf terminal C and power supply:
No continuity. Good

Maf terminal C and body ground:
No continuity. Good

Maf terminal D and power supply:
No continuity. Good

Maf terminal D and body ground:
No continuity. Good

Maf terminal E and power supply:
No continuity. Good

Intake air temp sensor check: measured old maf
Resistance between terminal D and E at ?: 68 sb 2.21kohm. Is 2.38kohm.

 

The fun continues. Thoughts on the below?

I was pouring over the various data logs I have (incorrect MAF at start and operating temp with throttle blips, correct MAF start attempt and after warming up with incorrect MAF, and various other ones).

What I noticed:
APP1 (v & %) never changed from 0.4v and 7.84%.
APP2 (v & %) never changed from 0.4v and 7.84%.
APP(%) never changed from 0.

I am going to run a KOEO throttle sweep when I get home later. The APP1 and APP2 do not align with what the mazda manual states.
Based on that, I will check the wiring from the PCM and report back.

 

As long as it's not WOT, I would think it should still start and idle.

 

Throttle sweeps for APP looked good. Let us not talk about me paying attention to APP and not looking at TP.

Anyway, following the manual with key on engine off and depressing the APP, see below. I am going to check the wiring to the PCM (it appears sheilded but no disconnects between it and the PCM) and pull the throttle body off and take a look. It was caked with oil deposits, and I cleaned it and the intake before reassembling.
*ignore the colors

 

So. A little update. No issues with the wiring from throttle body to pcm. Manuals States to replace tbody. Replaced and voltages and percentages improved but didnt meet the specs. Car still fired and died. Then tbody voltages and percent went back to now change.

While I was in the general area, I decided to replace the map sensor because it was also nasty when to tore the engine down. I cleaned it, but for peace of mind, I replaced it with the throttle body.

Now...it starts with maf disconnected. I shut it off and Rechecked all the connections I disturbed for replacement and that the gasket was in place. All good. Fired it up, runs like poo. Using throttle body cleaner worked my way around the engine bay looking for change in idle or stalling when sprayed. Nothing changed as I worked my way around. There are instructions for checking vacuum disconnecting the check valve and installing a gage. I will try that approach a other day. I'm just beat for now.

I think I'm going to take a vacation.

 

If it runs with the maf unplugged the maf is bad

 

Or something is wrong with the intake/filter setup that's causing it to read improperly.
Or the MAFCAL is bad/not right for the intake.

It's possible you fixed a leak that was tuned around or something, too.

 

Sadly, I tried the denso replacement and it acted the same.

It's a factory intake, filter, etc. When I feel more motivated, I'll try running it without being connected to the filter box.

I've got the full Forscan now. Is it possible to reflash? I haven't tried to search what can be done yet. But I will.

 

Found this thread https://mazdaspeeds.org/index.php?threads/how-to-ecu-replacement.10717/page-3

Downloaded my as built from the website to avoid any corrupted data.
Will attempt to reflash when I get home later.
Disassembled the battery box to get my PCM number L3BG 18 881A just in case. Did a visual inspection of pins and general condition lol.

 

Logged into ForScan and tried to write as built and got an error: number of blocks in file and module doesn't match continue anyway? I click no because i don't want to brick the ecm.

ForScan didn't give me the option to upload my as built it asked if I was connected to the internet and got the as built itself.

I uploaded my downloaded as built and images of what's in the car and error message. It's all above my head.

I rebooted the module and IC, same results. Now it starts, hit 2000 rpm and starts to come down to idle but seems to die as it starts to come down pretty consistently.

If the blocks don't match, does that mean my ecm is bad?

 

Reading on the ForScan forums, it said to bypass this and attempt to write ...seemed sketchy. Tried it anyway and got a "Configuration Structure Missmatch".

Going to talk to some of the controls engineers at work to get some feedback.