Blogs Car Info Our Show Deals Mechanics Files Vehicle Donation

A Test

What are the four primary inputs into the computer when the engine is started cold?


I’d say the crank sensor, coolant temp sensor, incoming air temp (charge temp), and the MAP or MAF sensor, depending on what the vehicle is equipped with. On some vehicles, the cam sensor too.

Input gas, input air, input spark, input oil?


I don’t know how it could be narrowed down to 4 and applied across the board to all PCMs on all cars so I’ll abstain…

I’m going to completely steal come text from some guy I see posting about cars online all of the time:
“There are only four primary inputs into the computer when the engine is cold. These are the coolant temp sensor for the computer, the crankshaft position sensor, the MAF/MAP sensor, and the throttle position sensor.”
~ Citation:

I know. I cheated.

Well, in my '61 Beetle it was… oh, that’s right, computers hadn’t been invented yet!
Of course, in my '89 Toyota pickup (carbed) it was…
Does a Breathalyzer count? My uncle’s car, by order of the court, has…

Yeh I know, gal dang systems, ok for my 68 itt needed air gas oil and spark, sure I am making fun of the brains in cars, but years were not specified so I want a gold star! Of course then we have to define computer, is an abacus a computer? sure. What is a computer? it is a machine that understands 1 and 0, extrapolated ad infinatum. So a spark plug is like a computer 0 no spark, 1 spark. basically a binary situation as a computer.

Interesting point, Barky. The difference is (I think) that “computers” can be programmed.
Which raises the question, are microprocessors that cannot be programmed computers?
Mike and a few others are computer guys. Perhaps they can provide us with a working definition.

Four primary inputs: gas, electricity, air, and oil.

ECT, TPS, CKP, MAF/MAP would be my answer to your question.

Although if the car is equipped with a CMP as most are nowadays, the absence of that signal may or may not cause a no-start. The ECM also need to see many other things to get the engine started though. It needs OK from the security system and to see a start signal to the starter motor.

Ase, I would argue that an engine just started cold needs to know the engine temperature, the crank speed, the crank position, the air intake or manifold absolute pressure, and little else to meter fuel and control spark. The throttle position I suppose is arguably necessary since even a throttle in its static idle position provides a signal that tells the ECU that it is, but I believe the ECU has a “default” output even if the throttle has no output. Therefore, the definition of “necessary” gets muddy.

Confession: I don’t have access to the data to verify my theory about the “default” output. So if you have that access, I’ll appreciate your correction.

TSM, the throttle position is for fuel control, an engine with a closed throttle needs far less fuel than one at 75% throttle. And at 100% throttle at starting no fuel will be delivered because the ECM will be in “clear flood mode” like the old carb days when you had to start a flooded engine. True, if the ECM sees an open circuit indicating sensor failure it will substitute an assumed reading to get the thing running. TPS input will also be used by the ECM to operate the Idle Air Control, to keep the engine running once it starts.

Don’t forget the CMP sensor. CKP alone often isn’t enough to control spark timing and injector pulse. Go out to almost any Mitsubishi and most Chrysler products, for example, and unplug the camshaft position sensor. It won’t start.

Let’s not forget something extremely important . . .

The PCM needs to recognize a valid key

Ase, you’ve confirmed for me that a TPS output isn’t necessary to get a cold engine to idle.

Re: crank speed/position vs. cam speed/position sensor; either can be used to tell the ECU both the speed and position of the crank, valves, pistons, etc. However, if the system is designed to use both or designed to use one or the other, unplugging either or it will not allow the engine to run.

Thanks for the feedback, by the way.

@the same mountainbike: No engine will run with the crank sensor disconnected. Some will with the cam sensor not in the circuit. I have confirmed this with my car. Supposedly with no cam sensor input the computer blindly fires all injectors like in pre-sequential systems. So economy suffers and drivability is a bit weird, but it will run and drive.

BTW, I would say that a CPU that is not programmable is still computing–its programming is just not alterable. You could argue that simple microcontrollers like that in small appliances, your cell phone battery charger, etc. are not programmable, but they are still processing data and reacting to it.

At that level it’s technically firmware. Even lowly PLDs have firmware. Gates configured to react to inputs and produce outputs. The line between discrete hardware, a collection of the same hardware components in a single package and a very large scale integration of the same hardware becomes blurred…

Thanks for the inputs guys. I suspect that even within the industry the term “computer” has long been debated.

Oblivion, that’s interesting info.

Well, in my '61 Beetle it was............ oh, that's right, computers hadn't been invented yet!

Back then…Computers had bugs…Bugs didn’t have computers.

First electronic computer (Eniac) was invented in 1946. But by 1961 they still weren’t fast enough or small enough for a VW Bug to carry it. Very few mac trucks could carry a computer fast enough to do what an ECU does today.

Speaking of computers in Bug’s…

Below is a pic of a VW Bug with a Dec PDP-8 in the back seat.

That’s a much later bug than a '61!

But you’re right, computers did technically exist in '61. Of course Eniac was big enough to fit a '61 VW inside the computer rather than the other way around. I wonder, can one even buy glass tubes anymore?

That's a much later bug than a '61!

True…Especially since the PDP-8 wasn’t invented until 1964.

I wonder, can one even buy glass tubes anymore?

Glass Tube computers were used in the Air Traffic Control as late as 1994. That’s extremely scary to even think about. The GPS system in a small single engine plane is more powerful then those computers…has more memory…and is smaller then just one of the 1,000,000+ tubes that made up the computer circuits. Not to mention the electricity (about the same as a small town consumes in a year)…or the heat (where you needed 5 industrial AC units to keep it at a tolerable temp of 75 degrees.