austin944

I have an intermittent no-start problem on an 84 944 with 175K miles. It cranks fine, but it does not fire. When it does decide to start, it runs perfectly, revs smoothly while sitting, does not feel fuel starved in any way.

When it fails to fire, I get no fuel pressure, and no fuel out of the rail after/during cranking. The fuel pump is getting power, but I cannot hear it run with a jumpered DME relay with the ignition on (relay pins 3,8,9). Battery is fully charged. Spark is good.

After not starting yesterday with no fuel pressure, the car started today and revs fine with a regular DME relay. The temperature has been about the same, maybe a little cooler.

It's done this twice now, where it would not start, then start and run fine.

It did have a long history of intermittent hard starting, which I attributed to the fuel check valve, which I am planning to replace. But it would always eventually start after enough cranking. The car had been mostly sitting awaiting repairs, but I drive it periodically and have added fuel stabilizer, and the engine had always run well prior to this problem.

Is it possible that the fuel pump is intermittently failing? AFAIK it is the original. Like maybe a motor winding is marginal? Can the check valve block the fuel line? Any other ideas?

931guru

Electric fuel pumps do fail. In my experience with Audis and VW cars, the pumps fail intermittently a couple of times before failing permanently. Carry a short broomstick, know where the fuel pump is on your car, and hit/bang/tap/etc. it if the car won't start... if it starts at that point, you know it's a failing fuel pump.

After nearly 40 years, I'm not surprised it's time to replace it.

The FPR and the fuel damper and the check valve are much more reliable in comparison.

Breakaway944

From what I understand you have narrowed down the starting problem to no gas.
If you were at the fuel pump, and verified it is getting power, and it makes no noise, it does not work. Problem solved!
If however you did not verify the fuel pump was getting power by putting a voltmeter on the leads of the fuel pump it could be several things.
Your symptoms of cranking a lot to start first made me think of one thing that causes it but had to dismiss it since you verified no fuel pressure.
How did you verify the fuel pump was getting power? How did you verify there was no fuel pressure? The more information you provide is helpful.

The two symptoms you mentioned, no fuel pressure and long crank to start, could be caused by two separate issues. Since you do not mention checking fuel pressure on previous "hard start" conditions I would not assume that was caused by fuel pump issues. It could be, but without testing that can not be verified. First fix the current problem that you have, verified as a bad fuel pump (gets power but does not run). Once that is fixed if you still have hard starting/long cranking (and have verified fuel pressure) look at the DME temp sensor. Actually I would check the wiring under the boot that plugs into the temp sensor. When these wires loose the insulation under the boot you will have an intermittent long crank/no start that will get more frequent as time goes on. Car will run fine when/if it starts.

austin944

It's not "cranking a lot to start", it's cranking a lot and NO start and NOT catching at all when it fails to start. And when it does decide to start, it starts right away with very little cranking, maybe 3 turns.

The prior history of the car is that it had very intermittent hard starting due to a probable check valve issue, but I've not seen any of that lately. Now it's always either starts right away, or never starts with a lot of cranking.

I verified pump power by attaching alligator clips to the pump electrical connections, with leads to a DVM. The DVM has the ability to record max voltage. So I crank the engine, it does not start and I do not hear the pump running with a jumpered DME inside a closed garage. I go to the back of the car where the DVM is connected, and I see a max of around 12.0V (not great, but OK) where it had previously recorded a max of 0V.

I had removed the fuel rail cap before the cranking session with a jumpered DME, and afterwards, and see no residual fuel come spitting out like it should, nor do I see fuel pressure on my gauge during cranking, nor does any fuel come out into a bottle when cranking.

austin944

I tried knocking on the fuel pump body with a plastic hammer a few times, and it's reliably starting today with a regular DME relay. I will try a few more times on different days to get it back to the fail to start mode, though it will be hard to tell if the knocking is doing anything, or whether other hidden demons are at work.

Breakaway944

If the fuel pump gets power but does not run it is bad. It is that simple.
I asked how you verified voltage and no fuel to see if it was done correctly. I am still confused by your answer. My understanding of what you are saying is - you jumpered DME relay, cranked no start, went back to the fuel pump to listen if it was running (I assume you went back because it is very hard to hear fuel pump from drivers seat), did not hear it running, checked DVM and DVM showed max voltage of 12ish volts.
Did the DVM show current volts? If it only showed max volts and showed current volts as 0 then it is an intermittent wiring problem. Knocking it with a hammer could temp fix a loose wire or a failing fuel pump, for a short time.
I interpreted this as Again, you have verified the fuel pump or wiring is bad, if I interpreted what you said correctly. After you replace pump or wires and you encounter This is textbook explanation of temp sensor wires being bare. The only difference is sometimes, not always, after a lot of cranking it will start, which you also mention. This will feel like a no fuel situation, ie NO start and NOT catching. This is a little confusing because most people would think a bad temp reading would create a hard to start issue, but the engine would "try" to start and catch. On the 944 it does not even catch or try, it feels like a no gas issue.

GPA951s

Unplug your reference signal sensors and plug them back in … i had one stall.. intermittent no start because there was just oxidation on the connections.. it was the “ when it does start its about three cranks and starts right up”

maint

+1 for [Unplug your reference signal sensors and plug them back in - it was the “ when it does start its about three cranks and starts right up”]
Do you get tach bounce? I chased a occasional no start for quite a while before digging into the reference sensors. Wire harness to both had broken & bare spots couple inches from the connectors. I changed both sensors and the harness (944online) and haven't had the problem since. I also couldn't hear my pump when cranking but I had to change it anyways because it started to leak at the joint. I originally suspected the DME relay and re flowed the solder to it (and a spare) but that wasn't the problem on mine

austin944

Bad sensors would not explain the absence of any fuel pressure at the rail with 12V getting to the fuel pump.

The car started and ran fine yesterday. Today it will not start and it's a little cooler from yesterday. Again I get no fuel coming out of the rail when I remove the cap at the end of the rail, no fuel pressure with the engine cranking, and am getting 12V at the fuel pump connectors. Hitting the pump with a hammer makes no difference so far. The engine sputtered for a second at the first crank, but I attribute that to residual fuel pressure in the rail from yesterday when the engine did run.

Looking more temperature related now, cooler does not start, warmer does start. I am thinking there is a marginal winding inside the pump motor, colder means the wire looses contact as it contracts.

I am going to replace the pump and perhaps put the old pump in the freezer and run some kind of bench test with the pump hooked up with gas. I hit the pump with a heat gun when it was not starting, but still did not start. It does not feel I am putting enough heat on the body of the pump but don't want to start a fire either.

GPA951s

Sounds like you found the issue.!

austin944

Just wanted to follow up on this thread I started. The engine is starting reliably now after replacing the fuel pump, fuel check valve, and fuel filter (while I was in there). It also starts faster than it did before, like instantaneously after turning the key, whereas before it was about 1 second to start. Thank you for all the help!