slomiata

Hi Guys,

This is not your TYPICAL "help me" question. I have a really weird thing going on with my miata race car.

I am a professional mechanic in Morro Bay, CA and have built a 1990 Miata turbocharged race car. It has a GReddy home-brew turbo/intercooler kit with 326cc injectors, Vortec FMU and Bipes ACU. Running at 15psi.

I have been really happy with the Bipes, but decided to hook up a 2bar GM map sensor to smooth out some rough running while freeway cruising.

Here's the problem:
*** After the MAP was installed, I started the car and it ran rough for about 15seconds, then rougher for an additional 10seconds. Then it died. When I went to start it again, it would START & DIE. With each attempt, START & DIE. START & DIE. START & DIE.
*** Spark and fuel pressure checked out ok.
*** I checked injector pulse with a "noid tester" and noticed that the injectors would PULSE on start-up, run for about 2-3 seconds, then the injectors' pulse stopped (causing the engine to die).
*** I disconnected the map sensor, replaced the ecm with a known good one, and the EXACT same chain of events occurred.
*** I noticed that I had forgotten to flip switch 1 on the Bipes from 1.6l to 1.8l, so I did...No change.

What I've done so far:
*** Tested every connection and sensor (all good)
*** Tested and tightened the grounds (one was loose, but was tightened BEFORE the 2nd ecm was installed
*** Returned the wiring to "pre-map sensor" configuration (no change)
*** Tested both ecms in a known running Miata and they acted the same as my racecar (described above)
*** Opened and inspected 1 of the ecms for burns/melt/fries/failures etc... NOTHING

I now know that the injector pulse is failing (possibly dropping the drivers), and that 2 ECMs have fried...

The only other component not tested is the bipes.... Could a loose ground OR having switch 1 flipped to 1.6 instead of 1.8 damage the bipes, causing it to eat ecms??

Has anyone ever experienced this type of thing? My next step is to disconnect the bipes and map sensor and try ANOTHER ecm!! AAAHHH!!!!!!!

Thanks!

Richard Wimmer
Service Technician
Jim's Automotive

Joe Perez

iTrader: (8)

How about the ignition system- are you getting spark after the injectors quit?

Are you getting 12 volts on the white/red wire that goes to the injectors, after they shut off?


I haven't seen a schematic for the ACU, however I can make an educated guess as to what the 1.6 / 1.8 switch does. Electrically, the VAF / MAF sensors on the '90-'97 cars are all pretty much identical, to wit: they output a 0-5v signal to the ECU. The major difference is that on the 1.6 cars, the signal is reversed. When there is no airflow, output is 5v. As airflow increases, voltage goes down. On the MAF-equipped cars, voltage rises as airflow rises.

So, this switch setting is most likely just telling the microcontroller inside the ACU which scaling to expect, and since MAP sensors work the same way as MAF (voltage increases as pressure increases), this is why you needed to flip the switch. It does not likely make one bit of difference how the box performs from an electrical standpoint, merely how the software runs. Ergo: you didn't cause any physical harm to the car by having it backwards.

So, at this point you've put everything back to how it was and yet you're still having trouble, right? Perhaps a wire that wasn't even part of this task got pinched or cut somewhere in the process?



EDIT: I just re-read your post again, and on a 1.6 car in particular, this START-DIE behavior sounds like the hallmark of a bad fuel cutoff switch which, incidentally, is inside the AFM that you've been ******* around with in a peripheral way. In the diagnostic box under the hood (drivers' side shock tower) put a wire jumper between FP and GND. You should hear the fuel pump running all the time when the key is on. Try that.

Braineack

iTrader: (62)

My question to you is: How are you preventing your AFR from achieving well over 10:1 after 4.5-5k?

slomiata

Thanks for the responses,

1) Spark never quits until engine dies and check engine light comes on.
2) I e-mailed Dick Bipes and he confirmed the same information as Joe about the switch and MAF.
3) Does the AFM cutoff control the fuel pump or the injector pulse??? My fuel pump is wired to a relay that is switched manually, so the computer cannot turn off the fuel pump. I will check the input values at the ECM to see if the AFM is telling the ECM to turn off the fuel...
4) AFM bench tests well, however at idle, the AFM should read ~3.3volts.. It is reading 2.23volts... (HOWEVER, these ECMs are acting the EXACT same way my racecar and in the OTHER "test" miata) :-/

My next step is to do a pin check at the ecu and check all wires for shorts/breaks. Hopefully I can find SOMETHING wrong.

Braineak... I've played around a bit with the FMU disks to achieve a reasonable AFR. I need to get a secondary wideband O2 to wire into my AFR gauge and run the stock O2 to the ECM. That will hopefully give me a good idea what my AFR truly is.

I'll keep you posted on progress with this problem.

Thanks again,
Richard

Matt Cramer

The AFM cutoff controls the fuel pump directly. It is possible to bypass this in a couple different ways. Of course, a zero air flow reading could also be translated into zero injector pulse, although usually the extent of our work with stock AFMs is documenting how much power you can gain by removing them.

Joe Perez

iTrader: (8)

^^ This.

On the 1.6 cars, the fuel pump is not controlled by the ECU. There is a physical switch contact inside the AFM which is closed whenever the flapper is not fully at rest, and is pushed open when the flapper does come to rest in the zero position.

This contact drives a relay (called the Circuit Opening Relay, or COR) via the light green wire. Specifically, when the contact inside the AFM is closed, the light green wire at the AFM is closed to ground via the black wire. Voltage for this coil is supplied into the COR via the white/red wire.

There is a second path for operating the COR as well. During cranking, the AFM is assumed not to give a reliable contact due to insufficient airflow. Thus, a violet colored wire comes from the starter motor into the COR, and this wire has +12 on it whenever the starter is operating. This causes a second coil within the COR to energize whenever the starter is engaged, to ensure that the fuel pump operates during this time. The return path for this coil is a black wire at the COR. (The violet wire also goes to the ECU, as a signal to tell the ECU that the starter is engaged, and this to switch to the cranking fuel and ignition tables.)

So this is why I am suspicious of your AFM and related wiring. It sounds to me like the fuel pump is only operating during cranking.

I cannot directly explain how this may be related to the failure migrating into other cars, but I would be remiss in not mentioning it.

Here are the relevant schematics:

slomiata

Joe,

Thanks for the diagrams. The fuel pump is wired seperate from the AFM. The AFM fuel switch circut is not a part of the fuel pump circut. It is simply wired through a switched relay.

I am conducting a PIN CHECK at the ECM plug right now, and have found that PIN 2K (LG/R wire, 5v MAF reference) has 280 ohms of resistance from the ECM Connector to the MAF sensor. When I unplug the MAF, the resistance is 0 ohms. So it seems that the MAF sensor is shorted to ground inside the MAF ?????

Does anyone have resistance valves for a PIN CHECK? I am searching for the correct values as we speak. I have found voltage values, but not resistance.

CHEERS!

slomiata

FYI,

I have access to ALLDATA to assist in this PAIN IN THE *** PROBLEM!!!

CHEERS!

slomiata

Well,

I hate to do this, but I think I am going to have rewire my harness making checks all the way.... I spent 4 hours today checking all the wires, and haven't been able to find anything WRONG.

I have a few things wired "differently", but the CAR RAN!! Doesn't mean it was right though...just means it worked (for a while)

SIGH.......

Next steps:
1)Open up the harness AND Inspect all wires
2) Replace AFM (MAF)
3) Replace ECM

Keep you posted......

slomiata

I'm sending my ECM to a company in TEXAS (Automotive Computer Supply - ECM To-Go)
www.ecmtogo.com

The lady said that they will test the ECM and tell me what happened to it. They can repair it too...