car starts right up when cold, but has hard time starting when hot

[jacobzking]

i have a 91, well maintained, start right up like a new car when the engine is cold. but after the car has been running a while, if I have to stop the car for a few minutes, then start it back up I have to keep cranking the engine, or prime the fuel pump a few times.

the engine temp runs normal so it's not over heating. a friend said the problem is the gasoline getting to how just sitting in the fuel rails over the hot engine. any ideas? it's really a pain when I stop to get gas, or run in the bank... then my car wont start

 

[jacobzking]

k, so i tried the search button (before getting flamed, but after creating the thread...) is it usually the main relay next to the ECU?

 

[pgilliam1]

The main relay would cause your car to die or not to start, but usually not hard starting after warm up. That problem may suggest you need to clean the throttle body, replace fuel filter, replace air filter, replace spark plugs or have fuel injector problems (I had mine cleaned and balanced and it helped my start-ups alot). Good Luck and Happy Motoring!

 

[natcc99]

That problem may suggest you need to clean the throttle body, replace fuel filter, replace air filter, replace spark plugs or have fuel injector problems (I had mine cleaned and balanced and it helped my start-ups alot).

Hello. Bringing back an old thread to life because I have the same symptom = hard to start after engine is warm.
In reading another thread, post #14 and #15 mentioned about the main relay being the root cause since the car cranks, but does not fire up.
http://www.nsxprime.com/forum/showt...my-Main-Relay-pics-added?highlight=main+relay

Can someone help me confirm the right direction? Thanks.

 

[Honcho]

Heat related starting issues are usually the main relay. The heat expands the cracks in the solder joints and breaks the electrical connection. But, this relates to cabin heat, not engine heat, as the relay is separated from the engine compartment. A few questions:

What year is the car?
Are you running a non-factory ECU tune?
Do you have the factory fuel injectors?

 

[natcc99]

Heat related starting issues are usually the main relay. The heat expands the cracks in the solder joints and breaks the electrical connection. But, this relates to cabin heat, not engine heat, as the relay is separated from the engine compartment. A few questions:

What year is the car?
Are you running a non-factory ECU tune?
Do you have the factory fuel injectors?

Hello Honcho

Thanks for your response. My car is a 98 with a factory ECU and fuel injectors.
With the main relay mounted behind the seat on the "fire wall", I guess heat from the engine can soak onto the main relay?

Any other advice before I order a new one? Thank you.

 

[Old Guy]

First off it could be a number of problems.

It is slightly possible that it is the main FI relay. The main FI relay can suffer from solder joint fractures and it is possible that once the relay is warm, the fractures are opening up leading to intermittent non operation. However, this would occur any time the relay got warm / hot. Since you have not reported the car quitting while driving, I think it unlikely that you have an intermittent main FI relay problem. Although, if you are running AC all the time that could be keeping the relay cool enough to prevent the failure until the relay heat soaks a bit while parked. In such case I would not expect the engine to start at all until the relay has cooled off. If you can get the engine to eventually restart (i.e. its hard to restart versus a no startl) while everything is still hot I would be inclined to look elsewhere for the cause. However, if you want to eliminate it as a possible cause Rock auto sells replacement relays for less than $50 for the 1998.

The fuel system is supposed to retain fuel pressure when you shut the engine off. This speeds up restart and prevents vapour forming in the fuel line with a hot engine while the car is parked. If you have a drooling injector or the check valve on the fuel pump is leaking the fuel rail will lose pressure and you will not get a quick start. This occurs infrequently on my 2000 and when the engine does start on a second try it runs rough for a couple of seconds before operation smooths out. It would be unusual for this to be a 'hot only' problem; however, if it does occur when hot and you get significant fuel pressure loss, the loss of pressure can lead to fuel vaporizing leading to vapour pockets in the fuel lines if the engine is really hot. Once the vapour forms in the fuel line the re pressurization of the fuel line by the pump will not cause the vapour to re condense. The engine won't start on 'vapour'. The vapour will disappear as it is gradually purged through the injectors or purged through the by-pass on the regulator as you crank the engine. Depending on how bad the pressure loss is and how much vapour formed, this may take a bit of cranking to get things happening and once it starts the engine may run poorly until the fuel mixture gets sorted out. This type of problem can occur randomly or all the time if the fuel system loses pressure all the time. The definitive test is to connect up a fuel pressure tester or install a fuel pressure gauge and check the static fuel pressure before a hot restart.

Cars before 2000 (or around that date) use separate temperature sensors for the dash gauge and the ECU. This is significant because your dash gauge can read just fine; but, if the temperature sensor for the ECU has gone out of spec, the ECU will not be reading the correct engine temperature. If the sensor causes the ECU to read a low temperature the ECU may be applying cold engine fuel enrichment during the start attempt which is creating the equivalent of a flooded engine. In order to get the engine to start putting the ECU into 'flood clear mode' by putting the throttle pedal to the floor during cranking is the normal procedure. The definite test for this is to check the resistance of the ECU temperature sensor as set out in the service manual. Note that if you have this problem it will typically always occur on a hot engine. It won't be intermittent unless the problem is caused by a loose electrical connection on the sensor.

 

[natcc99]

First off it could be a number of problems....

Hello Old Guy, I always appreciate your detailed responses.

The car runs fine once it has started. Thankfully, it does not quit while driving. Therefore, your note about the main relay makes sense that it may or may not fix the issue. For a replacement without the "nsx-tax", it might be worth eliminate a variable. I will look into this. How long have you been using your aftermarket rock-auto version? I'm assuming your non-OEM version has been functioning without issues?

Thanks for the suggestion for the fuel pressure test and ECU temperature sensor check. I will check the service manual to see what the spec are for the fuel pressure, but feel free to share the spec if you have it at your finger tips.

Thanks

 

[Honcho]

Hello Honcho

Thanks for your response. My car is a 98 with a factory ECU and fuel injectors.
With the main relay mounted behind the seat on the "fire wall", I guess heat from the engine can soak onto the main relay?

Any other advice before I order a new one? Thank you.

As noted by Old Guy, I was suspecting a fuel pressure issue. You could always find another owner with a good main relay and swap it into your car to see if that eliminates the problem. It would only take a minute or two and you would know right away if the relay is your problem...

The fuel system is supposed to retain fuel pressure when you shut the engine off. This speeds up restart and prevents vapour forming in the fuel line with a hot engine while the car is parked. If you have a drooling injector or the check valve on the fuel pump is leaking the fuel rail will lose pressure and you will not get a quick start. This occurs infrequently on my 2000 and when the engine does start on a second try it runs rough for a couple of seconds before operation smooths out. It would be unusual for this to be a 'hot only' problem; however, if it does occur when hot and you get significant fuel pressure loss, the loss of pressure can lead to fuel vaporizing leading to vapour pockets in the fuel lines if the engine is really hot. Once the vapour forms in the fuel line the re pressurization of the fuel line by the pump will not cause the vapour to re condense. The engine won't start on 'vapour'. The vapour will disappear as it is gradually purged through the injectors or purged through the by-pass on the regulator as you crank the engine. Depending on how bad the pressure loss is and how much vapour formed, this may take a bit of cranking to get things happening and once it starts the engine may run poorly until the fuel mixture gets sorted out. This type of problem can occur randomly or all the time if the fuel system loses pressure all the time. The definitive test is to connect up a fuel pressure tester or install a fuel pressure gauge and check the static fuel pressure before a hot restart.

This is exactly what I was getting at with those questions. Most NSX injectors are "drooling" at this point. Just think of 20 years of gas varnish running through them. I would send them off to RC in Torrance, CA and have them cleaned and blueprinted and see if that fixes the issue. It's easier and cheaper than replacing the fuel pump, which would be the next stage. Also consider that your fuel pump resistor might be going bad. If the resistor crapped out after it gets hot, it would replicate the issue you are having almost exactly.

 

[Old Guy]

Hello Old Guy, I always appreciate your detailed responses.

The car runs fine once it has started. Thankfully, it does not quit while driving. Therefore, your note about the main relay makes sense that it may or may not fix the issue. For a replacement without the "nsx-tax", it might be worth eliminate a variable. I will look into this. How long have you been using your aftermarket rock-auto version? I'm assuming your non-OEM version has been functioning without issues?

Thanks for the suggestion for the fuel pressure test and ECU temperature sensor check. I will check the service manual to see what the spec are for the fuel pressure, but feel free to share the spec if you have it at your finger tips.

Thanks

My 2000 is still running fine on the original main FI relay. I purchased the Rock Auto (Standard Motor Products) relay some time ago after reading all the comments on Prime about the inevitability of main relay failures. The spare has been riding around in my trunk ever since. So, no advice to provide on the durability of the aftermarket stuff.

 

[Old Guy]

This is exactly what I was getting at with those questions. Most NSX injectors are "drooling" at this point. Just think of 20 years of gas varnish running through them. I would send them off to RC in Torrance, CA and have them cleaned and blueprinted and see if that fixes the issue. It's easier and cheaper than replacing the fuel pump, which would be the next stage. Also consider that your fuel pump resistor might be going bad. If the resistor crapped out after it gets hot, it would replicate the issue you are having almost exactly.

Indeed. If the pressure loss is caused by injector drool, that is a much easier fix than the check valve which sits on the pump outlet. I don't know whether the check valve is a separate replaceable item. Even if it is the hassle of dropping the fuel tank says replace the pump at the same time because you don't want to be repeating this in 5-10 years.

In the early years of first gen EFI the injectors were attached to the rails with barbed connectors (think Datsun 260Z). You could pull the rail up with injectors attached, pressurize the system and then visually check for injector drool. That is not so doable anymore. Unfortunately the pressure test just confirms that you are losing pressure. It doesn't tell you whether its the injectors or the pump / check valve. The car is a '98, so cleaning the injectors would never be a bad thing. However, unless its a garage queen a 21 year old fuel pump also has to be getting up there. Decisions, decisions!

A flaky fuel pump resistor is in the realm of possibilities. However, the resistor gets really hot all on its own without any help and stays hot until you switch into VTEC operation. If a bad connection at the resistor was causing the problem I would expect that it would also materialize while driving rather than just on restart. Also, during the prime cycle (and I think cranking) the ECU by-passes the fuel pump resistor so the car should show a start during cranking and then die when switching to run.

 

[natcc99]

Fun times with owning an old car.

Thanks gents for the advices. Let me try a few of them and report back.

 

[Old Guy]

Fun times with owning an old car.

In this price range, its vintage or classic or exotic, not old - like good wine as opposed to plonk.

On the up side, you are probably not having nearly as much 'fun times' as owners of similar age Ferraris, Lambos, Maseratis, Porches, Lotus and a bunch of other exotics.

 

[natcc99]

In this price range, its vintage or classic or exotic, not old - like good wine as opposed to plonk.

On the up side, you are probably not having nearly as much 'fun times' as owners of similar age Ferraris, Lambos, Maseratis, Porches, Lotus and a bunch of other exotics.

haha, nice way to making lemonade out of lemons. I agree, we are fortunate that our maintenance procedure does not start with remove motor.

 

[Old Guy]

Indeed. I briefly toyed with the idea of purchasing a Ferrari 355 spider that came up for sale close to me at what seemed like a nice price. That ended when research showed that cam belt replacement was absolutely an engine out operation and that the typical cost for major maintenance was around $10,000 Cdn. The official Ferrari replacement schedule was every 3 years - give me a break! That and issues with excessive valve guide wear and a very hot engine leading to exhaust manifold failures and early failure of CV joint boots made it a pass. Its apparent why most Ferrari's become garage queens.

 

[natcc99]

The F355 is a gorgeous car. Although, I always thought it was odd looking how long the front overhang is.
The maintenance is just ridiculous. It is probably why you see photos of the engine out of the car on BaT listings to prove to the seller that it has been performed.

Perhaps it is best you did not get that F355. You could toy with the idea of a Fiero with a F355 body.

 

[Old Guy]

Perhaps it is best you did not get that F355. You could toy with the idea of a Fiero with a F355 body. :biggrin: :biggrin: :biggrin:

By coincidence, the place where I purchased my NSX had a Fiero with a Countach body. A genuine Easter Egg car (fancy shell over a cooked interior); but, probably more reliable than a genuine Countach. They also had a Ferrari 360. The owner of the place said the 360 was entertaining to drive around in for about 20-30 minutes. After that it was either ear plugs or acetaminophen. His opinion was that if you had to actually go some place and get there the NSX was the preferred piece; but, he was selling the NSX not the 360. Probably a different song for a 360 customer.

 

[natcc99]

By coincidence, the place where I purchased my NSX had a Fiero with a Countach body.

Another example of how small this world is.

 

[DLockwood]

Guys bringing back an old thread. Can anyone chime in on this….

Only when the car sits under the sun and it’s hot after a drive (today it sat for an hour after a drive), (ambient temp around 25 degrees Celsius but here in Australia under the sun gets to 35 + it takes long to crank. As it’s cranking the engine almost starts and the flywheel grinds for a split second on the starter while both are spinning almost like the flywheel sped up for a split second to start running but not fully then the car starts. Symptoms all point to fuel pressure loss but im yet to hook up a fuel pressure gauge. Not sure if the tw sensor could be affecting the hot start.

Car is/ has

91 jdm na1 nsx

New oem fpr
New oem fuel pump assembly
Injectors flow tested and cleaned w new seals
Fuel filter done a year ago
Spoon ecu
New oem fuel main relay

I thought if it was a drooling injector it would have been picked up when it got flow tested and cleaned. But maybe not.
Any help appreciated

 

[Old Guy]

First off, if your injector cleaning service was reputable they should have flagged an injector that continued to leak after cleaning as requiring replacement.

The starting problem with fuel pressure loss after the car has been parked is a momentary problem. It can result in the engine failing to start right away on the first try. On the second try the fuel system will usually be fully primed by the fuel pump and the engine will start; but, typically run rough for 2-5 seconds until the fuel mix is sorted out. The pressure loss when parked can be caused by
- drooling injectors
- a leaky check valve in the fuel pump
- a leaky fuel pressure regulator
- fuel piping leaks which should be obvious by the ' l'eau d'refinery ' smell around the car

The most reliable way to check for pressure loss is with a fuel pressure gauge using the correct fitting at the fuel filter discharge port. Power up the car (don't start the engine) and let the fuel pump prime to establish pressure (make sure that it is at the factory spec) and then turn the car off and watch the pressure gauge. The pressure should hold just below factory spec. If it drops quickly after pump shut off then something is leaking. If the injectors are good and the pump is new it might be the regulator.

Fuel pressure loss while parked is not just a 'hot' problem. It can occur anytime. As such I suspect your problem is not a fuel pressure loss problem. The fuel pressure loss problem would also resolve itself with persistent cranking of the engine which does not seem to be happening with your engine. Rather, I suspect that you have a hot start fuel mixture problem.

You did not indicate whether this hot start problem is a recent change or has been a persistent problem since you got the car. If it a recent change then checking the operating fuel pressures during a hot start would be a good starting place. Checking the engine coolant temperature sensor parameters would also be a good idea. If you had an OEM ECU using the service check connector to trigger an error code display would indicate if there was a serious error with the temperature sensors.

If you had an OEM ECU, I would suggest that as soon as you get a hot start failure, try to start it by putting the ECU into flood clear mode which consists of pushing the throttle pedal right to the floor and then hitting the starter. If flood clear is going to work you will usually get success within about 3 - 7 seconds. If you are cranking beyond 7 seconds you are just wearing your starter motor out. If flood clear works then you would need to investigate why you have an apparent starting mixture problem.

However, you do not have an OEM ECU and I don't know what a Spoon ECU is. If this problem has been around since you got the car then I think you may have a hot start parameter configuration problem with the ECU. Is the Spoon an OEM ECU that has been chipped? If it is then I expect that the start up parameters are the same as OEM and that the flood clear mode should be functional so give it a try. If it is a chipped ECU then I also expect that the start up configuration parameters are not the problem unless Spoon screwed those up. If the Spoon ECU is a completely user configurable ECU it may have a flood clear mode which may or may not be enabled by whoever set up the configuration files. If you have the configuration software you should be able to check hot start parameters to see if it is enabled and then give flood clear a try to see if it 'patches' the problem. If the Spoon ECU is completely user configurable and the hot start problem has existed from day one then I expect that you have a start up configuration problem.

So, check the resting fuel pressure to rule that out as a problem and then you need to provide some more details about the Spoon ECU.

 

[Honcho]

However, you do not have an OEM ECU and I don't know what a Spoon ECU is. If this problem has been around since you got the car then I think you may have a hot start parameter configuration problem with the ECU. Is teh Spoon an OEM ECU that has been chipped?

Yes, the Spoon ECU is just a OEM ECU with a EPROM chip. Most of these were sold in the 90's and early 2000's so there is no way to know what they changed. I suppose you could call Spoon and ask if they changed the starting parameters. Another thing to consider is that this is a JDM ECU, which is tuned for the high-octane petrol available in Japan instead of the crappy 91 we have in the states. It may have different starting parameters.

 

[DLockwood]

First off, if your injector cleaning service was reputable they should have flagged an injector that continued to leak after cleaning as requiring replacement.

The starting problem with fuel pressure loss after the car has been parked is a momentary problem. It can result in the engine failing to start right away on the first try. On the second try the fuel system will usually be fully primed by the fuel pump and the engine will start; but, typically run rough for 2-5 seconds until the fuel mix is sorted out. The pressure loss when parked can be caused by
- drooling injectors
- a leaky check valve in the fuel pump
- a leaky fuel pressure regulator
- fuel piping leaks which should be obvious by the ' l'eau d'refinery ' smell around the car

The most reliable way to check for pressure loss is with a fuel pressure gauge using the correct fitting at the fuel filter discharge port. Power up the car (don't start the engine) and let the fuel pump prime to establish pressure (make sure that it is at the factory spec) and then turn the car off and watch the pressure gauge. The pressure should hold just below factory spec. If it drops quickly after pump shut off then something is leaking. If the injectors are good and the pump is new it might be the regulator.

Fuel pressure loss while parked is not just a 'hot' problem. It can occur anytime. As such I suspect your problem is not a fuel pressure loss problem. The fuel pressure loss problem would also resolve itself with persistent cranking of the engine which does not seem to be happening with your engine. Rather, I suspect that you have a hot start fuel mixture problem.

You did not indicate whether this hot start problem is a recent change or has been a persistent problem since you got the car. If it a recent change then checking the operating fuel pressures during a hot start would be a good starting place. Checking the engine coolant temperature sensor parameters would also be a good idea. If you had an OEM ECU using the service check connector to trigger an error code display would indicate if there was a serious error with the temperature sensors.

If you had an OEM ECU, I would suggest that as soon as you get a hot start failure, try to start it by putting the ECU into flood clear mode which consists of pushing the throttle pedal right to the floor and then hitting the starter. If flood clear is going to work you will usually get success within about 3 - 7 seconds. If you are cranking beyond 7 seconds you are just wearing your starter motor out. If flood clear works then you would need to investigate why you have an apparent starting mixture problem.

However, you do not have an OEM ECU and I don't know what a Spoon ECU is. If this problem has been around since you got the car then I think you may have a hot start parameter configuration problem with the ECU. Is the Spoon an OEM ECU that has been chipped? If it is then I expect that the start up parameters are the same as OEM and that the flood clear mode should be functional so give it a try. If it is a chipped ECU then I also expect that the start up configuration parameters are not the problem unless Spoon screwed those up. If the Spoon ECU is a completely user configurable ECU it may have a flood clear mode which may or may not be enabled by whoever set up the configuration files. If you have the configuration software you should be able to check hot start parameters to see if it is enabled and then give flood clear a try to see if it 'patches' the problem. If the Spoon ECU is completely user configurable and the hot start problem has existed from day one then I expect that you have a start up configuration problem.

So, check the resting fuel pressure to rule that out as a problem and then you need to provide some more details about the Spoon ECU.

Old guy thank you so much for the detailed reply, the car has being having the hot start issue since I’ve got it and that’s before I changed the fuel pressure reg and fuel pump. After I did the fuel system refresh the problem seemed to go away for a while but is now back.

The spoon ecu is just a standard ecu modified by spoon but yea who knows what they did back in the 90s. Im going to check the ecu tw sensor and install the fuel pressure gauge.

The next thing would be to get a proper programable ecu to tune the car properly to Australian fuel and also some of the mods the car has. I believe if the car has a fuel pressure issue I would get other issues like running rich or also cold start issues like you mentioned it not being a hot start only issue, the weird thing is it only does it when the car is hot as in parked under the sun and not show signs of the problem when the engines hot, so hot ambient air temp is the only common thing I’ve pin pointed to make the issue re occur… it’s quite odd.

 

[Old Guy]

If it is just a chipped OEM ECU, then chances are it may still have the flood clear function enabled. In that case, I would be tempted to give the flood clear a try. If the flood clear is successful in starting the engine, then the problem is likely over-fuelling during start ups. That may be a tuning problem or some other problem like excessively high fuel pressure during a hot start.

Since the ECU is essentially OEM, also try the service check connector to see if there are any stored error codes that might indicate what the problem is.

You mentioned 'some of the mods the car has'. What other 'mods' does the car have? You haven't switched to a different fuel injector by any chance? When cold, engines are moderately tolerant of a little more than necessary fuel delivery on start up; but, a little extra fuel can make hot starts a pain. Those other mods might be contributing to a hot start issue so those details would be useful.

My understanding is that the chipped OEM ECUs from Japan primarily modified the ignition timing map to presumably take advantage of the 100 RON premium (JIS standard) fuel in Japan. That would be about 94-95 by the R+M/2 method used in North America. According to Google, Oz also uses the RON rating method and the common premium is minimum 95 with up to 98 available. That is significantly below the JIS premium 100 RON; but, 97 RON is in the range of the typical 91-92 RM/2 in North America. So, if you are burning 97-98 RON there is no advantage to using the revised ignition maps since you are knock limited by the fuel. There is a down side to running knock limited with a revised ignition map. If the map is set to use 100 RON it is possible for you to initiate detonation with 98 RON fuel using the 100 RON map. Presumably, the knock sensors will detect the detonation and retard the timing. However, detonation will have to occur for the knock sensor to work. If you run 98 RON with an ignition map meant for 98 RON you are timing limited and you will never incur detonation if everything else is fine. The knock sensor is just there as a back-up safety system for a bad batch of fuel or the time you had to fill up with regular because they ran out of premium. You have moved the knock sensor from a back-up safety function to your first line of defence. Not a good idea, Particularly if your knock sensors are getting deaf with old age.

Personally, unless the compression ratio has been increased, I am undersold on the benefits of bumping the ignition timing to take advantage of fuels with higher octane ratings. Higher octane fuel actually burns slower than lower octane fuel so it needs more timing advance to complete the burn at the optimal point after TDC. You can tweak the timing to fix the burn time problem; but, if you have not bumped the compression ratio the big up-side is not there.

Rather than change the ECU, you might want to check out the work that @MotorMouth93 and Honcho were doing with the characterization of the pre OBDII ECU. With a small amount of soldering work on the ECU and the correct software interface you will have access to the fuel and ignition maps. One of them might be able to supply you with the ignition map for a North American na1 which are set up to burn 91 R+M/2 which should work just fine with 98-98 RON in Oz. The big up side is that the start up and warm up configuration in the OEM ECU will be miles ahead of anything that you will get for a stand alone ECU because you do not have enough time / money to pay a tuner to spend the time to get that stuff sorted as well as Honda has it sorted (been there, done that, got the Tee shirt). Plus, the OEM ECU has error code reporting (although not as extensive as on the OBDII ECUs.

 

[DLockwood]

Thank you @Old Guy , does my nsx have a fuel pressure sensor? I will also check the vacuum lines for the fpr are ok.

The car has gt one headers, high flow cats and a grf exhaust. Intake is stock and the dash pot has been deleted. Apart from that everything is is factory.

As for the ecu, wouldn’t a standalone ecu be better in terms of modern ecu technology and tuning flexibility? We have a tuner here that’s very knowledgeable with Honda (previous Toda racing tech and aircraft engineer) so I was going to get him to completely tune the car to Aussie 98 RON as opposed to the Japanese MON fuel, plus dial in the engine to run better with the bolt on mods.

First things first I will install a fuel pressure gauge and check / replace the tw sensor and go from there

 

[NSXGB]

Mine has done this for many years. Every time I start the car, even from cold now, I wait for the fuel pump to prime and light on dash to go out before cranking. Starts straight away every time, hot or cold.