Carpokes Porsche Forum

@CarBot i have a 1986 951 that cranks but does not star. What should I check/test to find the problem?

zooklm1 wrote: i have a 1986 951 that cranks but does not star. What should I check/test to find the problem?

On a 1986 951 that cranks but won’t start, I’d start with the quick checks that narrow it down to “no DME power,” “no RPM signal,” “no fuel,” or “no spark.”

First, listen/feel for the DME relay behavior. The DME relay is a very common failure and can cause a straight crank-no-start. If you have a known-good spare, swap it in. If you don’t, you can test for fuel pump operation during cranking (or briefly key-on depending on how it’s wired) and verify the DME is actually powering up.

On a 951-specific clue, watch the boost gauge at key-on. If the boost gauge doesn’t do its little jump at key-on, that often points to no DME power (relay, ignition switch feed, alarm/immobilizer interruption, etc.).

Next, check for tach bounce while cranking. If the tach needle doesn’t twitch at all during cranking, suspect the speed/reference sensor circuit at the bellhousing or the harness/connectors. On these cars, a bad sensor, bad gap, damaged harness, or corroded connectors can prevent the DME from seeing engine speed, and it won’t fire injectors or spark.

Then do a basic spark and fuel split test. Pull a plug wire and confirm you have strong spark while cranking. Separately, confirm you have fuel pressure and injector pulse (a noid light helps). If you have spark but no injector pulse, that again points back toward DME power or speed/reference input. If you have injector pulse but no spark, look harder at ignition components and DME output.

Also worth checking because they’re common on 944/951: cracked solder joints inside the DME (and KLR on the turbo cars), ignition switch electrical portion failing intermittently, injector harness shorts, and any aftermarket alarm system faults cutting power.

If you reply with whether the boost gauge jumps at key-on, whether the tach bounces while cranking, and whether you have spark and fuel pressure, I can point you to the most likely next test instead of guessing.

@CarBot Boost gauge moves on key on, tach jumps during cranking, swapped DME relay, fuel is getting to the injectors and still no start.

zooklm1 wrote: Wed Feb 25, 2026 5:02 am @CarBot Boost gauge moves on key on, tach jumps during cranking, swapped DME relay, fuel is getting to the injectors and still no start.

@CarBot did you respond to this? Either way, if the injectors are pulsing as they should and you check for spark and have none, you might start by checking the cap and rotor and then bypassing the KLR. If you do have both fuel and spark, think big vacuum leaks, cam timing etc. -- though I would expect some backfires and stumbling if you have both fuel and spark.

Tom wrote: did you respond to this? Either way, if the injectors are pulsing as they should and you check for spark and have none, you might start by checking the cap and rotor and then bypassing the KLR. If you do have both fuel and spark, think big vacuum leaks, cam timing etc. -- though I would expect some backfires and stumbling if you have both fuel and spark.

I didn’t respond earlier in that thread; I’m replying now because you mentioned me.

Your logic is solid. If you’ve confirmed injector pulse but there’s no spark, the next practical checks are distributor cap and rotor condition, coil output, and the ignition power stage/igniter wiring. On the 951, the KLR can also get involved in ways that confuse diagnosis, so temporarily bypassing it as a test can be a useful isolation step if you know the correct procedure for your year and harness.

If you truly have both injector pulse and strong spark, then yes, I’d shift to air and timing. A massive vacuum leak (big hose off, intercooler plumbing popped, brake booster line, J-boot issues) or cam timing being off will usually give you at least some sputter/backfire, so a totally dead “just cranks” with both present makes me re-check that the spark is happening at the right time and that the distributor rotor is actually turning (timing belt intact, distributor drive intact).

If you post what you saw when checking spark (where you tested it and how strong it was) and whether the rotor turns while cranking, that usually narrows it fast.

CarBot wrote: Wed Feb 25, 2026 8:04 am

Tom wrote: did you respond to this? Either way, if the injectors are pulsing as they should and you check for spark and have none, you might start by checking the cap and rotor and then bypassing the KLR. If you do have both fuel and spark, think big vacuum leaks, cam timing etc. -- though I would expect some backfires and stumbling if you have both fuel and spark.

I didn’t respond earlier in that thread; I’m replying now because you mentioned me.

Your logic is solid. If you’ve confirmed injector pulse but there’s no spark, the next practical checks are distributor cap and rotor condition, coil output, and the ignition power stage/igniter wiring. On the 951, the KLR can also get involved in ways that confuse diagnosis, so temporarily bypassing it as a test can be a useful isolation step if you know the correct procedure for your year and harness.

If you truly have both injector pulse and strong spark, then yes, I’d shift to air and timing. A massive vacuum leak (big hose off, intercooler plumbing popped, brake booster line, J-boot issues) or cam timing being off will usually give you at least some sputter/backfire, so a totally dead “just cranks” with both present makes me re-check that the spark is happening at the right time and that the distributor rotor is actually turning (timing belt intact, distributor drive intact).

If you post what you saw when checking spark (where you tested it and how strong it was) and whether the rotor turns while cranking, that usually narrows it fast.

Could a cycle valve failure cause a no start condition?

zooklm1 wrote: Could a cycle valve failure cause a no start condition?

A failed cycling valve by itself almost never causes a true no-start on a 951. The cycling valve mainly affects boost control. If it’s dead or unplugged you typically get low/limited boost or overboost behavior under load, not a crank-no-start.

The only way it tends to matter for starting is indirectly, if there’s a wiring issue or short in that circuit that’s dragging down power or confusing the DME/KLR, or if a related vacuum hose routing mistake created a big vacuum/boost leak. Even then, you’d usually still get some sputter or an attempt to fire.

For a clean “cranks, no start,” I’d stay focused on spark present and correctly timed, injector pulse, fuel pressure, and the speed/reference sensor signals, plus basics like the distributor rotor turning and the DME/KLR connectors and grounds. If you tell me whether you have spark at the coil wire and at a plug, and whether the rotor turns while cranking, I can point you to the most likely next check.

CarBot wrote: Wed Feb 25, 2026 4:40 pm

zooklm1 wrote: Could a cycle valve failure cause a no start condition?

A failed cycling valve by itself almost never causes a true no-start on a 951. The cycling valve mainly affects boost control. If it’s dead or unplugged you typically get low/limited boost or overboost behavior under load, not a crank-no-start.

The only way it tends to matter for starting is indirectly, if there’s a wiring issue or short in that circuit that’s dragging down power or confusing the DME/KLR, or if a related vacuum hose routing mistake created a big vacuum/boost leak. Even then, you’d usually still get some sputter or an attempt to fire.

For a clean “cranks, no start,” I’d stay focused on spark present and correctly timed, injector pulse, fuel pressure, and the speed/reference sensor signals, plus basics like the distributor rotor turning and the DME/KLR connectors and grounds. If you tell me whether you have spark at the coil wire and at a plug, and whether the rotor turns while cranking, I can point you to the most likely next check.

I have spark at the plugs.

zooklm1 wrote: I have spark at the plugs.

If you have spark at the plugs, the next question is whether you also have the right amount of fuel and whether the spark is happening at the right time.

First, confirm fuel pressure with a gauge at the rail. “Fuel at the injectors” or even injector pulse doesn’t guarantee you have correct pressure. A weak pump, clogged filter, bad regulator/damper, or a split in-tank hose can give you flow but not enough pressure to start.

Second, verify injector pulse with a noid light (or a scope if you have one). If you have spark but no injector pulse, that points back to DME inputs/outputs even if the tach bounces.

Third, make sure the ignition timing isn’t wildly off. Pull the distributor cap and confirm the rotor turns while cranking. If it doesn’t, stop and check the timing belt/distributor drive. If it does turn, bring cylinder 1 to TDC on the crank mark and confirm the cam mark and rotor position line up reasonably. A slipped belt or incorrect cam timing can give you spark but no start, sometimes with little to no backfire.

Fourth, do a quick “will it fire on added fuel” test. If it will briefly catch on a small amount of starting fluid (used carefully), that strongly suggests a fueling problem (pressure, injectors not actually delivering, or gross mixture issue). If it still won’t even try, think timing, compression, or a massive air leak.

Also worth checking on a 951: the blue DME temp sensor connection. If it’s unplugged or the sensor is way out of range, it can cause an over/under-fuel condition that makes starting very difficult.

If you can share whether you have confirmed fuel pressure at the rail and whether the rotor turns and lines up at TDC, that will narrow it down quickly.

These CarBot answers are coming mostly from our no-start knowledge base, so it is giving pretty sound advice. Given your recent findings on the bad vacuum hose, and it subsequently starting but running poorly, I'd focus on ensuring the proper AFR, most likely vacuum related. When it runs, how much vacuum are you seeing at the FPR vacuum line, and what's the RPM. CC'ing @CarBot to keep it up to speed and see what it things now.