Computer controlled engines bad

[biggoolies]

Okay you hear that for an expedition truck a fully manual truck with no computer control is the way to go as electronics have a good chance to fail out in the harsh environment we travel to. But we see vehicles like the Toyota Landcruiser or the Nissan Patrol that go out into these harsh environments every day and come back home. I know if the electronics fails out in the boonies your gonna have a harder time getting it back home. But how often does this happen? Wouldn't the computer/ electronic systems of current 4wd trucks be designed to withstand a certain level of punishment?

 

[SkiFreak]

The reality is that vehicle computers are built to be quite robust, but like any piece of electronic equipment, it is possible that they can fail.
The other problem you can have with newer engines is that they can be designed specifically for low sulphur diesel, which I would think is more of a concern if looking for a vehicle that can be driven anywhere in the world.

One problem with computer controlled engines is fault finding issues if something does go wrong. If you have the ability to read the error codes that are output from the ECU there is a fair chance that you can resolve the problem yourself, as you would on a non computer controller engine.
I have seen on occasions that the ECU can put the vehicle into "limp home" mode. Diagnosing why this has happened can only really be achieved from the error codes. Sometimes it can be as simple as a faulty sensor, or even the electrical plug for the sensor. With the error codes you can identify the location of the fault, but without them it is pretty much total guesswork.

If I were travelling afar, I would carry a tool that can read and reset the error codes in the ECU. Some of these are reasonably cheap, others (like for my FG84D) are not. Even if you were not able to fix the problem yourself, with this information you may be able to guide a mechanic to a possible solution, if they did not have their own diagnostic equipment (a possibility in 3rd world countries).

 

[alan]

We are seeing in our workshop common rail diesels with problems slowly increasing, and it's expensive! and you cannot rely on the computer to tell you fault codes, and if fault codes do appear they can be misleading, it's really a nightmare and it's all brands Toyota, Nissan,Mitsubishi,Landrover,Jeeps. the most common fault seems to be injectors and fuel pressure issues, the diesel in Australia is mostly poor quality we always tell customers to stick will BP and Shell Diesel and fit a secondary filter, poor quality fuel in a common rail is a disaster!

 

[1leglance]

I still love the story a member here posted about a new LR4 tapping an electric fence and the entire rig shutting down. Including trapping the occupants inside due to elec locks & windows. Now I would have thought there was a way to open an electric locked door from the inside but maybe that is an anti-theft thing.

Anyway I think I will stick with my phone or laptop being the only computer in my rig.

 

[gait]

several years ago now but stranded in shiny new Landrover Discovery.

Took three days and much communication with LR in UK to find the fuel cutout valve. It not only cutout but broke internally on a corrugated road. No warning lights, no diagnostics available, not in any available manuals (operating or maintenance). Zilch.

Not as bad but a Ford Falcon auto. Accidentally moved gear to reverse while moving slowly forwards. No indication of why the vehicle wouldn't then start. Disconnect battery trick failed. Industrial systems typically have a "reset" switch.

Really as bad as staring at an ATM that says "transaction not allowed". Little Britain's "computer says no". There are examples of planes falling out of the sky when fly-by-wire gets knotted up.

An early design guide for computer systems was to make things visible. That seems to have been lost somewhere along the way.

Its really a universal problem as computers are used more extensively but designers really haven't got round to designing for what happens when things go wrong. The assumption that the sensor they interpret for one failure mode only indicates that failure mode is fundamentally flawed. And rarely do they allow for sensor failure. Add to that the increasing regulatory environment in developed countries that insist on repairs being carried out by qualified persons, plus manufacturers protecting their information and after sales business, despite education, training and availability not having caught up.

When regulations were introduced to provide diagnostic connections (next to steering column) someone forgot to add that the diagnostics should be available to anyone, and useful.

Even the non-computer models are somewhat flawed, just that the design problem is amplified with the complexity designers can add to computer systems.

A trivial but good example is engine temperature measurement on my Canter. When water level is low the gauge reads low, on the stops. I first thought it was a gauge failure due to a loose connection after rough tracks. Coolant reservoir had visible liquid within limits. When checks of sensor and wiring drew a blank I looked further. At least the fuel gauge fails by reading low!

The fault finding guides in the back of operating and maintenance manuals were never exhaustive or complete. That's been transferred into the computer systems.

Basically, until design catches up with reality (and I'm not holding my breath) simple is good, knowing and understanding the vehicle is good, having a vehicle that can be understood is good.

end of rant!