Pls help diagnose cranking no start

2001 Honda Accord, F23A1 4cyl engine, 5spd manual transmission

I've been doing some work on the car the last few weeks and went to start it today and it cranks but doesn't start. Tried again a few minutes later and it started and died within seconds; tried again and no start - video of this below.

Work I've done along with starts in between (car hasn't been driven at all during this work, but I did start it occasionally as described below) :

1. EGR valve removed and cleaned. Started afterward, no problem.
2. replaced timing belt, balance belt, water pump, alternator and power steering belts, and several seals - crank, balance shaft, cam and oil pump/balancer case gear seals. Started afterward, no problem. Ran for 45min.
3. valve adjustment. Started afterward, no problem. Ran for a few minutes.
4. valve cover gasket and spark plug tube seal replacement.
5. replace brake rotors and brake pads. This is where I am - no start.

Battery voltage is 12.4V and checked scan tool - no codes confirmed or pending.

Unless it's a random issue unrelated to anything I've done, my initial thinking was that it must be related to the valve cover gasket and spark plug tube seal replacement as this is the first time I've tried starting the car after that (and seems it can't have anything to do with brake rotor/pad change).

When I was performing the valve adjustment, I had to unclip the wiring harness that sits directly behind the valve cover (and in front of the fuel rail) so that I could push it down to get the valve adjustment tool to seat on the intake side valve stem nuts. I thought I put everything back, and the car did start just fine after this valve adjustment, but I only ran it for a few minutes.

So thinking I may have damaged the wiring harness, this was my first point of diagnosis - I removed the plugs to the fuel injectors and checked - each of the 4 plugs is getting battery voltage on the power wire when ignition is ON, and when cranking, each plug is getting a pulsed ground on the control wire. I can hear the fuel pump priming when turning ignition to ON position. So between the fuel pump priming and normal signal reaching the fuel injector plugs, I'm temporarily ruling out fuel as the problem.

Then I checked for spark. I'm getting spark at each plug wire.

So at this point I started worrying that timing belt had jumped a tooth (even though engine idled fine for 45min after timing belt job). So I removed the upper timing cover and visually confirmed the cam and crank are properly in time, and I did 2 rotations of the crank (equals 1 rotation of the cam) to confirm that belt is on the pulleys and top dead center per the crank and cam marks is indeed top dead center of cylinder 1 (long socket extension down spark plug tube).

Pics confirming cam/crank timing (all pics taken at same time without moving crank or cam). The crank pulley pic shows the 3 marks that precede the TDC mark, and I had to shift the camera a little to the left so that the TDC mark would be visible as it's directly under the lower timing cover pointer.


The tooth spacing on the cam seems large enough that if the timing was a tooth off, it would be obvious in where the cam marks line up.




Fuel pressure light, brake light, battery light, all door and trunk lights are all on when key is in ON position (position 2) and they don't go away. First time I'm noticing this - not sure if this is normal and they go away when engine starts (which I can't get to since engine isn't starting). Or if this is abnormal (guessing it's normal).

I don't have a high level scanner, but I do have the Bluedriver OBD2 scanner, which does show some live data, so if there's some data that would be helpful to capture while attempting to start the vehicle, I can do that. I also have a remote start switch, so cranking and observing things simultaneously is not a problem. Also have DMM and LED test light.

What to check next? Compression test each cylinder? Does the cranking in the video sound like an engine that isn't achieving compression? I'm guessing crank position sensor is sending the appropriate signals, which is enabling pulsed ground to the fuel injectors, so I suppose that can be ruled out.

Video below. The surge in idle after it starts happened without me touching the throttle. Noise between first and second starts was just me pumping the brake to reestablish brake fluid pressure after changing rotors and pads.

Your timing marks look right on the money. No worries there, IMO. I can also see from the video that you've tried opening the throttle during cranking, good move. Stuck/dirty IAC's aren't common on these engines, but not unheard of.

LOVE the testing you've done. I know that the pump is priming with the key on, and that the fuel injectors are being controlled electrically, but I think I'd try adding some alternate fuel to the intake anyway. Carb cleaner, brake cleaner, whatever. Take the intake boot off, give the throttle body a few good shots and try cranking again. If you can keep it running with the cleaner, you're on track for a fuel problem.

If alternate fuel keeps it running, then we can dig into the fuel pump or a possible PGM-FI main relay issue.

That wasn't me opening the throttle, that was the engine. I don't think the throttle can be manipulated by the computer, so the engine must have opened the IAC in order for the RPM to rise?

Last few times the engine was run, it was just idling, so with the throttle closed, the IAC would have to be open. So if it did get stuck open, it shouldn't prevent starting, should it? Thinking out loud here - this is all new to me.

I can get some starting fluid from the auto parts store and spray that into the intake while cranking.

I can't shake the feeling that this no start is somehow related to the work I've done on the car, so I'm thinking fuel is unlikely, especially given fuel pump priming and electrical signals to fuel injector plugs. But maybe it's unlucky coincidence. Or maybe it has something to do with car sitting for about a week since the previous start, which itself was only for a few minutes.

Since I've got the spark plugs out, I may as well perform a compression test. Anything else I should try before putting things back together (spark plugs, fuel injector plugs, valve cover, upper timing cover, power steering belt) so that I can attempt starting with starting fluid?

Could the crank position sensor be at fault? I had to remove it from around the crank when replacing the crank seal. But I took a pic of how the sensor's parts were oriented before removal and referred to that pic when reinstalling.

As I understand it, if the crank position sensor weren't sending any signal back to the ECM, then there wouldn't be any pulsing ground at the fuel injector plugs. And I think on this particular engine, spark is also dependent on receiving a signal from the crank position sensor.

But could the crank position sensor be sending signals out of time? Everything else seems to be in time - cam and crank are in time with each other. Spark is in time because the plug wires are connecting to the correct cylinders (different plug wire lengths make it hard to screw this up). So maybe the crank position sensor is somehow throwing off the timing?

Is there a way to check the timing of the signal coming from the crank position sensor?

You're sure you didn't open the throttle? 'Cause I can see the butterfly open at about 26 seconds. No electronic throttle here, maybe you hit the gas accidentally? If not, definitely try holding the throttle open on the next crank attempt. You're right on that the IAC had to have been doing something previously, but those valves will usually default to closed with no current flow.

Nothing else I can think of that you should do before buttoning stuff back up. If you've got the plugs out, have a close look at the tips and in the cylinders (as best you can, anyway). If they're wet with fuel, then you may be dealing with a ring wash problem. Add some oil to the cylinders before reassembly. If they're completely dry, then that'd be another strike against the fuel system.

FWIW, I don't think you caused this. The only thing I can imagine is if you took the plug wires out of the tubes AND off the distributor when doing the valve cover gasket, and now they're out of order on the distributor. Otherwise, there's very little you could have done wrong.

Your crank does sound funny. I do hear a slight inconsistency. I'm not sure if the starter is doing that or the compression.

I would have done relative compression test using the injector as the sync. This will definitely tell you if the timing is right.

But it appears you are already going to pull the plugs.

Make sure to pull all the plugs before doing the compression test.

What did you determine?

I like the idea of spraying carb cleaner or propane into manifold, be sure to NOT hit the MAF when doing this. This is a super quick way to determine if fuel is the issue.

jrud wrote: Could the crank position sensor be at fault? I had to remove it from around the crank when replacing the crank seal. But I took a pic of how the sensor's parts were oriented before removal and referred to that pic when reinstalling.

As I understand it, if the crank position sensor weren't sending any signal back to the ECM, then there wouldn't be any pulsing ground at the fuel injector plugs. And I think on this particular engine, spark is also dependent on receiving a signal from the crank position sensor.

But could the crank position sensor be sending signals out of time? Everything else seems to be in time - cam and crank are in time with each other. Spark is in time because the plug wires are connecting to the correct cylinders (different plug wire lengths make it hard to screw this up). So maybe the crank position sensor is somehow throwing off the timing?

It's possible? But like you said, I think the presence of spark and injection pulse would rule that out.

The different plug wire lengths do make it tough to get the wire positions wrong at the valve cover, but you CAN get them wrong at the distributor. If you never took them off the distributor, then disregard.

Is there a way to check the timing of the signal coming from the crank position sensor?

Absolutely, but you'd need a scope to do it. A DMM will show you the presence of the signal, but not the timing.

A timing light would work, if you have one or can get a hold of one. Clamped around the #1 wire, you should see the light flash when the timing notches come up to the mark on the lower timing cover. The exact timing isn't that important yet, just that the spark is happening close to TDC.

Tyler wrote: You're sure you didn't open the throttle? 'Cause I can see the butterfly open at about 26 seconds.

Oh, I thought you were referring to the blip up in RPMs on the first attempt - that wasn't me. In the video, there is some movement below the air intake hose at the 4-5 sec mark. Not sure what that is, I'll have to look down there.

Yeah, on the second attempt where you see the butterfly open, that was me giving it a little throttle hoping to help the engine catch.

Tyler wrote: Nothing else I can think of that you should do before buttoning stuff back up. If you've got the plugs out, have a close look at the tips and in the cylinders (as best you can, anyway). If they're wet with fuel, then you may be dealing with a ring wash problem. Add some oil to the cylinders before reassembly. If they're completely dry, then that'd be another strike against the fuel system.

Let me take a pic of the spark plugs. They look normal to me, and when I took them out, I thought I smelled a whiff of fuel when I put the plugs close to my nose (not soaked, but not bone dry), but I'm sure that will be gone (evaporated) by now.

Tyler wrote: FWIW, I don't think you caused this. The only thing I can imagine is if you took the plug wires out of the tubes AND off the distributor when doing the valve cover gasket, and now they're out of order on the distributor. Otherwise, there's very little you could have done wrong.

Thanks Tyler. No, I didn't pull the plug wires off the distributor. I was excited that I had finished up all this work on the car (all first-time maintenance items for me), so it was a huge letdown that the car wouldn't start, and I immediately thought I had screwed something up.

GeekDIYMechanic wrote: Your crank does sound funny. I do hear a slight inconsistency. I'm not sure if the starter is doing that or the compression.

I would have done relative compression test using the injector as the sync. This will definitely tell you if the timing is right.

But it appears you are already going to pull the plugs.

Make sure to pull all the plugs before doing the compression test.

What did you determine?

I like the idea of spraying carb cleaner or propane into manifold, be sure to NOT hit the MAF when doing this. This is a super quick way to determine if fuel is the issue.

Sounds like I'd need an amp probe and labscope to perform a relative compression test, neither of which I have.

I'll pull all the plugs before doing the compression test and post the results.

Thanks for the tip regarding the MAF.

Yeah, on the second attempt where you see the butterfly open, that was me giving it a little throttle hoping to help the engine catch.

Right on, good idea. Scratch the IAC. Some kind of breathing problem is still on the table (intake/exhaust restriction), but that's further down on the no start checklist.

Let us know what you find!

Did you have the tps and map disconnected? I vaguely remember those connectors will reach and fit either one - unless I'm thinking of something else that is possible on. Maybe see if they got switched around.

Compression test results:
cyl #1: 220
cyl #2: 215
cyl #3: 180
cyl #4: 220

This was done using a remote starter switch (I was standing in front of the engine), so throttle was closed (i.e., air pedal up), and engine is cold (obviously since it doesn't start).

Not sure why so much deviation in cylinder #3, but I'm guessing the overall compression results rule out insufficient compression as the culprit driving the no start condition.

Off to get some starting fluid.

Those are excellent compression numbers. Perhaps even a little high? :huh: Doesn't matter right now, that's more than enough to start. Fuel washed cylinders don't make those kinds of numbers, either, so that's out.

GeekDIYMechanic wrote: I like the idea of spraying carb cleaner or propane into manifold, be sure to NOT hit the MAF when doing this. This is a super quick way to determine if fuel is the issue.

Sorry sir, no MAF on this one. :blush: But good advice on vehicles equipped with MAFs!

Let us know what happens with the spray.

I agree with the suggestion to add a fuel source, if you don't have a fuel pressure tester. If the thing starts while spraying, the first thing I would do is go to the main relay down near the brake pedal, and the black/yellow wire coming out of the back should have full voltage while cranking, as it is the power to the fuel pump. Those main relays have a history of causing problems. Just because you hear it buzz initially doesn't tell you much, the pump can have low voltage or lose power completely if the relay is internally acting up.

As far as "not spraying the MAF"... someone correct me if I'm wrong, but don't all those old Honda's work off of Speed Density? (doesn't have a MAF). You should still be careful when spraying in the intake, as if you drown it, you could have a small fire on your hands if it backfires - just keep your face away, haha.

You do have a compression issue on that one cylinder, or at least it's washed. That wont keep it from starting though, that's not your issue.

Does the car have any aftermarket stuff wired in, like a viper alarm or anything? And does any security light stay on while cranking?

Let us know what you get with the fuel situation

Sorry, just saw Tyler's message, looks like I repeated a few things he said.

John Z wrote: Did you have the tps and map disconnected? I vaguely remember those connectors will reach and fit either one - unless I'm thinking of something else that is possible on. Maybe see if they got switched around.

No. I don't even know where they are exactly, as I wasn't doing anything to the intake.

Tyler wrote: Those are excellent compression numbers. Perhaps even a little high? :huh: Doesn't matter right now, that's more than enough to start. Fuel washed cylinders don't make those kinds of numbers, either, so that's out.

Yeah, I'll have to check Honda spec for this vehicle, but it does seem high. Maybe carbon buildup? Anyway, as you note, it's not particularly relevant to the no start and we now know that compression isn't the issue.

Hadn't thought about it, but it makes sense that fuel washed cylinders could have abnormally low compression if the oil is washed away, preventing a good seal.

Anyway, it seems that we can pretty much definitively rule out spark, cam-crank timing, and compression. And as you noted earlier, exhaust isn't likely - seems like that's the kind of issue that gives you lots of advance warning, not this kind of sudden failure.

So it's back to fuel, which I had only temporarily ruled out pending other now-completed and passed tests, and which everyone here seems to be eyeing as the likely missing factor in combustion.

TheTechWhisperer wrote: As far as "not spraying the MAF"... someone correct me if I'm wrong, but don't all those old Honda's work off of Speed Density? (doesn't have a MAF). You should still be careful when spraying in the intake, as if you drown it, you could have a small fire on your hands if it backfires - just keep your face away, haha.

Yeah, I looked it up and this gen has a MAP, not a MAF. Subsequent generation Accords have a MAF.

And the MAP is farther down the intake, so if I disconnect the hose at the air box, I think I should be safe spraying starting fluid in at that point.

TheTechWhisperer wrote: Does the car have any aftermarket stuff wired in, like a viper alarm or anything? And does any security light stay on while cranking?

Bought the car used, I think the alarm is factory (if that's even possible), or perhaps dealer installed. No, the security light is off.

TheTechWhisperer wrote: Let us know what you get with the fuel situation

Since I just installed the valve cover gasket and used gasket goop at the sharp bends in the gasket, I'll have to wait at least a couple of hours before trying the starting fluid, so that will probably have to wait until tomorrow. I must say that gaskets and Permatex and cleaning surfaces for gaskets is the part of DIY automotive work that I hate the most! :pinch:

TheTechWhisperer wrote: I agree with the suggestion to add a fuel source, if you don't have a fuel pressure tester. If the thing starts while spraying, the first thing I would do is go to the main relay down near the brake pedal, and the black/yellow wire coming out of the back should have full voltage while cranking, as it is the power to the fuel pump. Those main relays have a history of causing problems. Just because you hear it buzz initially doesn't tell you much, the pump can have low voltage or lose power completely if the relay is internally acting up.

Yeah, I was doing a little googling this evening, and I saw a number of mentions regarding the main relay and the fuel pump - including crank but no start symptoms. Seems promising, given what has been ruled out from prior tests.

And thanks for pointing out that the initial buzz from the fuel pump isn't definitive proof of anything. It's always tempting in these diagnostic situations to make assumptions, which is an easy way to miss a problem.

In addition to backprobing the black/yellow wire while cranking, does it make sense to pull the main relay out and run tests on it? The Honda service manual has a number of tests. Have seen some videos where people, including Eric The Car Guy, have pointed out a cracked solder joint - and the fix is as easy as resoldering the joint.

Well done using sealant on the corners. :B What Permatex sealant did you use, exactly? I've started those engines up immediately after a valve cover install with no leaks, so I think you're good to go if you want. If you'd rather not chance it, no worries!

SD has an excellent series on the PMG-FI main relay as well, if you're interested. The classroom lecture video is especially good.

I wouldn't recommend doing the Honda bench tests. They're unloaded tests, and the relay isn't easy to dig out from under the dash. :silly: Check voltages in the car first, or try tapping the relay, see if you can get it to work temporarily.

Tyler wrote: Well done using sealant on the corners. :B What Permatex sealant did you use, exactly? I've started those engines up immediately after a valve cover install with no leaks, so I think you're good to go if you want. If you'd rather not chance it, no worries!

Permatex Ultra Grey. No way am I chancing that - I'd never forgive myself if there was a leak and I had to deal with that valve cover gasket again! :pinch:

Tyler wrote: SD has an excellent series on the PMG-FI main relay as well, if you're interested. The classroom lecture video is especially good.

I wouldn't recommend doing the Honda bench tests. They're unloaded tests, and the relay isn't easy to dig out from under the dash. :silly: Check voltages in the car first, or try tapping the relay, see if you can get it to work temporarily.

Yeah, I watched the Part 1 video where he shows the bad solder connection. In fact, I've watched a few such videos, and out of curiosity, I thought I'd take a look at the solder connections on the relay despite your suggestion to leave it alone. Getting the relay out wasn't too difficult (pulled the relay out from the housing, didn't try to remove the housing). As I was pulling the relay out with the connector still plugged into it, there was a little spark - DOH! In my haste, I had forgotten to disconnect the battery. I think the exposed copper on the top side of the relay came into contact or very close to metal housing that was in the path as I was pulling the relay down.

I may have hosed the relay. Worse, I may have hosed an ECM since the connector was still plugged into the relay. :sick:

Not sure how to proceed other than continue on the original diagnosing path and see if other issues arise. I'll watch SD's part 2 video on PGM-FI and see what I can learn from that. And I guess after diagnosing original issue, it might make sense to run through the tests in the service manual to see if all the relay and connector pins are testing normal.

Anyway, the solder joints didn't look as bad as the one SD showed in his video, but from the very awkward position under the dash, I can't say for sure. Would have to pull it out and examine it closely.