mhr650

Our last race was 24hrs, within the first hour we got a crack in our header and with all the mighty rotary heat going out into the engine bay it caused all kind of gremlins. One of the more interesting problems was when the throttle sensor failed, we run a GM sensor mounted directly on the throttle body, and it failed in a way that the MS3Pro sensed that the throttle was 100% open all the time. Since I run speed density the car would run fine with the bad sensor, I have a strong engine in a light car with a lot of gear so it uses a very small amount of AE anyway. The problem was that when we came in to do a pit stop the car wouldn't restart and if you stalled or had to shut the car down during a red flag you would be stuck on track and have to be towed in. Turns out that when the ECU sees 100% throttle during cranking it goes into flood clear mode and wont inject any fuel, I could disable that feature but it seems likely that sometime in the future I my genuinely want it to shut off fuel to clear a flood. I am half tempted to wire in a bypass so I can throw a switch and eliminate the TPS in case I get another failure. We were able to get the car started during pit stops by spraying brake cleaner into the intake until it would catch and run on its own, it was a serious pain to say the least.

Not megasquirt related but another thing that caused a problem was the fuel pump shutting off. I run a Holley fuel pump that is actually 2 separate pumps inside a single housing with a switch so I can change from one pump to the other. This would seem like a great idea since one of the most common failures in endurance racing is fuel pumps. Turns out the possible downside is when the switch itself that you use to select the pumps fails then it doesn't matter how many perfectly fine working fuel pumps you have in the car.

BTW the MS3Pro was mounted right on the other side of the firewall from the leaking header and it never missed a beat during the whole time we were running.

Aaron Cake

iTrader: (1)

While a switch with a resistor to bypass the TPS with a set value is quick from inside the car, the same thing can be accomplished in the pits by just disabling flood clear in software or disconnecting the TPS.

Interesting failure.

peejay

MS3Pro should have a feature where it recognizes a failed sensor and ignores it/uses an assumed figure. Sensors never are railed to ground or 5v, so a sensor that reads 0 or 5v is assumed bad.

I found this while doing an install on a car where someone replaced a coolant sender with a fan switch. The computer was showing 170F on an engine that had never been fired. That was fun to find, thought it was a wiring problem.

There are other neat failsafe features you can use, and are one reason why I want to switch to the Pro unit.

The fun part is that a TPS failsafe would have to be a HARD failure and not just a resistance jump, since even if it sensed a TPS failure by the fueling not jibing with the Alpha-N failsafe map, it still doesn't remember faults after cycling the key, and you may reset the computer when you try starting the car...

mhr650

I did unplug the TPS when I had a little more time to hook up the laptop and saw what the problem was after we pulled the car out of the tire wall at 2AM. After that it would inject fuel during cranking, after we got back on track the fuel pump selector switch finally failed completely and at the time I didn't understand that it was the problem so at 4AM it was time to call it a night…

header crack that caused all of our problems



She looked a little rough after pulling it out of the tire wall at 2AM