You're thinking of older carbureted engines. A vacuum leak will cause an increase in idle speed without an increase in power, and the engine will run smoother than normal. With a carbureted engine, the lean condition can cause preignition, especially under load, and that can result in a backfire through the intake and carburetor. There is no vacuum yet when you're cranking the engine, so a leaking small vacuum hose would have no affect on starting. If the leak was really big, there might not be enough air flow through the venturi in the carburetor, so no fuel would be sucked in. In that case, pressing the accelerator pedal would work the mechanical accelerator pump that all carburetors use, and that would squirt in some fuel and aid starting.
With fuel injected engines, on all brands other than Chryslers, fuel metering calculations are based on the weight of the air going through the mass air flow sensor, then modified slightly based on readings from the coolant temperature sensor, intake air temperature sensor, throttle position sensor, and map sensor which GM uses to measure barometric pressure. The map sensor can also be used as a backup strategy when there's a failure with the mass air flow sensor. Chrysler is the only manufacturer that has been able to make an engine run right with just the map sensor. That measures intake manifold vacuum. Those sensors are so accurate that if they wanted to, they could measure engine rpm by counting the pulses of tiny vacuum changes as each piston takes a gulp of air. They don't actually do that, but that shows that those sensors are really sensitive to changes in load on the engine.
You have to remember that during cranking, initially there is no vacuum, so vacuum can't be used as an input for initial fuel metering. It's all the other sensor readings that go into those calculations. When you have to press the accelerator to get the engine to start, there's a number of possible causes. If you have to hold it to the floor, that is "clear flood" mode and a good suspect is an injector leaking down after the engine is stopped. If you only have to hold the accelerator down 1/4", a better suspect is idle speed that is too low. That could be caused by a stuck automatic idle speed motor, a plugged air passage, or the computer not raising idle speed to the desired level. You would need to connect a scanner to view live data to see why the computer is not commanding a higher speed. GM never really had much trouble with the air passages getting plugged with carbon, and even the engines on which that WAS common, it isn't anymore with the better additives in today's fuel.
Your Engine Computer is going to command pulses from the injectors as soon as you turn on the ignition switch to provide the priming fuel for start-up. The amount of that fuel is based on coolant temperature and intake air temperature sensor. It has nothing to do with vacuum because there isn't any yet.
You also have to consider that having to hold the accelerator down can be confused with a simple extended crank time. If fuel pressure bled down while the engine was off, it has to build up to where the engine can start. Chrysler engines can run as low as 20 psi and still run okay, but 45 to 50 psi is normal. On many GM engines, they won't start or will run very poorly if the fuel pressure is just a few pounds low. Typically that is caused by a fuel pump that isn't running up to its normal speed, but it can also be caused by a plugged pickup screen in the tank, a partially-plugged fuel filter, or excessive varnish buildup in the injectors. If the only problem is extended crank time, and it's due to low fuel pressure, the fuel pump will not run until the engine is cranking; then the pump is naturally going to run very slow due to battery voltage being drawn down by the starter. That means pressure will build up very slowly, and it could take perhaps five seconds before fuel starts spraying from the injectors. Holding the accelerator pedal might be something you resort to just as the engine would start on its own anyway, or the Engine Computer could be interpreting the throttle position sensor's signal voltage as a need for additional fuel. In response, it will hold the injectors open longer and give them more time to dribble in sufficient fuel for starting.
One thing you can try to identify this cause is to turn on the ignition switch to the "run" position, then wait for the fuel pump to finish its one-second pulse. That is programmed in to insure fuel pressure is up for starting, but it's not long enough if the pressure had bled all the way down to nothing. Turn the ignition switch off, wait a couple of seconds, then turn it back on and crank the engine. That will give the pump twice as much time to get the pressure up. If that results in better starting, suspect a leaking injector first, but also a leaking fuel pressure regulator, and a leaking check valve in the fuel pump assembly.
Sunday, January 18th, 2015 AT 12:16 PM