Turbo problems!
I took the Colt to Yuubari for a spin, and on my way back to Sapporo, things happened.
Before you read: I didn't do any actual hands-on repairs. I just figured out the problem and sent the car to a Mitsubishi dealer.
Problems
Will be detailed later, but let's just go through everything first
- Unstable idling revs (minor jumps)
- Engine warning light lit up in the dashboard
- Dead boost gauge, constantly pointing somewhere around the zero mark (not exactly at 0 but you know, it wasn't moving)
I later came to know that my boost gauge is a fully mechanical one. The only electronics inside of it is the lamp, which is dead anyways. At the time the engine started to behave curiously, I didn't know that, so I couldn't rule a lot of other possibilities out very soon.
I have a OBD2 Bluetooth adapter (ELM327 equivialent) plugged in all the time and a car scanner application on my phone. Of course I would scan for any OBD error codes. And I got only one, P0170.
It actually came with explanations: Fuel trim (FT), bank 1 - malfunction / Fuel Trim Malfunction (Bank 1); Abnormal fuel system. That is informative but not enough to pinpoint what exactly was at fault.
A quick search on the Internet showed that this error code could be affiliated with various causes, but unsurprisingly all related to the fuel system or air intake.
DIY diagnosis
The first thing to rule out was anything injection-related. The instability of idling revs was too 'smooth' for a possible injection nozzle problem. By smooth I mean the interval of the high and low rev jumps are about equal.
I thought it was less possible that more than one injection port had problems if any actually did. And one jammed nozzle would probably translate to a 1/4 low rev versus 3/4 high rev because the Colt has an inline 4. I'm not sure my logic is right though.
So the problem would lie in the air route. On the Internet most of the accessible sources are about Lancer Evos and they usually have a faulty O2, or air intake sensor. First I thought that if that sensor, be it oxygen or MAF or whatever, was bad, then the OBD would surely pop a specific error for it. But then I realised as long as the sensor is giving numbers (in the form of voltage), there's probably no way for the onboard computer to know if the numbers are true or not. Thus it won't know whether or not the sensor is bad. So I will simply keep the air sensor as one of the possibilities at this stage.
But apparently I have more significant suspects. As I wrote before, the boost gauge stopped turning with a constant display of nearly 0, which means it's detecting a constant air pressure, and that air pressure is close to 1 atmosphere.
Considering it was a gloomy rainy day, the atmo was probably lower than 1 bar and the boost gauge was also showing slightly less than 1 bar.
I didn't know if my boost gauge was mechanical or electronic, though. If it was mechanical then apparently somewhere in the whole vacuum cavity leaked, making everything along the way atmospheric. If it was electronic, however, it might be just some cable snapped.
But a leak in the vacuum was by far the most possible. And knowing that I was assured that the vehicle won't explode if I, well, drove it. It was just a naturally aspirated engine with a wrongly mapped ECU.
Fix
Basically I told the shop guys what I thought. They actually said what they could do was plug in a scanner and look at the codes and then decide what to do. I was not sure how much effort I saved them because apparently scanning for error codes was never the most difficult thing to do in the world.
And it took them 40 minutes to fix the car. Turned out the leak was near the boost gauge, somewhere near the connection between the hose and the gauge socket. They just cut the leaking section and secured the still usable portion with zipties. Neat, I would've done the same.