Odd engine quit

[Wavey1]

We headed out on a short road trip in the NSX yesterday. At about 25 miles the engine just shut off and the check engine light came on. I turned onto a side road, turned the key off, and it restarted immediately. We turned around a drove home to switch cars, and it ran perfectly all the way back. There were no other warning lights or anything weird with the other gauges. It just quit like the key was turned off.

I checked for codes this morning and there's nothing, zero codes stored. It's a '92 with about 140K miles, extremely well maintained. The main relay was replaced at about 110K as preventative maintenance. I figure the likely suspects are the igniter, the fuel pump relay or the fuel pump itself, which are all original. This doesn't seem like the typical ignition switch problem. The battery is at 5 years and I planned on replacing it, but again this doesn't seem like a typical battery issue.

The cats were replaced a few hundred miles ago with Pride high-flow cats, and the O2 sensors were also replaced with NTK parts. No other recent changes.

Without any codes and with it being a one-time occurance so far, I doubt that running through the troubleshooting sequences for the suspect parts is going to reveal anything. Any ideas on what to look for?

EDIT: I think I remember that the fuel pump relay is part of the main relay, so not that.

 

[drew]

Investigate the ignition switch.

 

[SMYLDOC]

I had 2 separate occurrences with my 92 NSX with 220,000 + miles
1st was about 1 mile into my office commute at 25 mph and the engine just stopped, all dash lights came on, BUT, the car would crank, but not start..... the ignitor failed.
I replaced it with the AS Motorsport ignition kit that does not use an ignitor
2nd incident 6 months later, same circumstances as yours, but I was able to restart the car, this happened 4 to 5 times over a one week period.....the fuel output on my AEM Series one ECU failed, had the ECU rebuilt, then replaced it with the upgraded Infinity model.
Good luck finding your problem, let us know what it was
And Merry Christmas

 

[Wavey1]

Investigate the ignition switch.

I thought of that, but would that cause the check engine light to come on? Wouldn't it just die with no check engine light? I have a new spare here, may throw that on for good measure.

 

[Old Guy]

I thought of that, but would that cause the check engine light to come on? Wouldn't it just die with no check engine light? I have a new spare here, may throw that on for good measure.

Late at night and recovering from eating too much and probably drinking too much wine Xmas day - so cut me some slack if this doesn't necessarily hang together.

Did the check engine light come on after the engine died and stay on or did it come on and then go out? If something momentarily cuts power to the ECU (ignition switch, main FI relay, bad battery connection ..all sorts of other stuff) and then power is restored the ECU will go through the power up cycle. During the normal power up cycle the MIL will come on and then go out as part of a normal power up cycle. If the ECU powered up fast enough with the car still in gear and the clutch engaged I would expect that the motor would have restarted. If the MIL came on and stayed on until you turned the key off and then the car restarted normally with the MIL staying off, that is odd. Not sure that I can comment on that; however, if there was a general loss of power to everything, that could erase any stored codes in the ECU; however, I am having trouble working my head around the particular scenario that could cause that.

You said that the warning lights were normal; but, by any chance did you notice whether the warning lights happened to go through the self test sequence just after the engine quit (i.e. the self test sequence that occurs when you first turn the ignition key to run)? If they did, that might be an indication of a momentary primary loss of electrical supply to everything. If that occurred, then one of the first places I would look is the condition of the battery post clamps. Those clamps stretch and can result in intermittent connections. However, there are a whole bunch of other loose or bad connections that could cause the same problem. The ignition switch could also cause this problem.

There is a fuel pump relay which is actually the fuel pump resistor by-pass relay. The fuel pump (resistor by-pass relay) relay is not the source of your problem. Below the VTEC operating region the relay is open so a failed relay will not cause the engine to stop. I don't have exhaustive experience; but, my experience has been that if the fuel pump starts up it will continue to run and not just stop due to failure. My experience has been that fuel pump failure usually materializes as a failure to start.

The igniter or bad connection / ground connection on the igniter could cause this problem. But, so could a bad connection on the +12v supply to the ignition coils or the +12v supply to the fuel injectors or bad connections at a lot of other spots. Checking the condition of the battery posts is the easy thing to do. Also, carry out a general visual inspection for damaged wiring. Other than that it may be a case of waiting for the next occurrence and observing the dash lights to see if they give you any clue as to what might be going on.

Prime member NsaneXr had a persistent random problem with a hesitation caused by momentary loss of electrical power. He hooked up a small event recorder to measure and record voltages at various points in the car to track down the source of his problem which turned out to be an intermittent short circuit on a connection in the alternator.

 

[Wavey1]

Thanks for your extensive analysis Old Guy! I thought you might jump in :smile: Some responses:

"If the MIL came on and stayed on until you turned the key off and then the car restarted normally with the MIL staying off, that is odd." Yes that's exactly what happened. Cruising along at about 70 MPH in 5th, engine loses power like it was switched off concurrent with check engine light coming on, all gauges functioning normally, check engine light stays on until ignition switch turned off, then on restart everything functioned perfectly normally, including the other idiot lights going through normal sequence. Odd indeed. The ONLY symptoms were complete loss of engine power, and the check engine light on. The everything functioning completely normally, immediately, on restart.

"Did you notice whether the warning lights happened to go through the self test sequence just after the engine quit" No they did not, just the check engine light on until the ignition switched off.

"If that occurred, then one of the first places I would look is the condition of the battery post clamps." I will check that but I'm nearly certain they're fine; I'm pretty anal about that stuff. The positive cable is Redbird's later OEM Acura part.

"The ignition switch could also cause this problem." Yes, I may see if I can recreate the problem by fiddling with the key and the connector while the car is running. But a failure there wouldn't cause ONLY the check engine light to come on would it? Wouldn't all the dash lights come on?

I will start at the battery and check/clean all the connections and grounds you suggested and see if I catch anything. I'm also going to replace the battery given that it's at 5.5 years; was planning on that soon anyway.

Good advice here, thanks!

 

[Old Guy]

Responses are in bold. The section of the 1991 service manual starting at page 11-27 gives you some things that you might look at and poke and wiggle trying to induce a reappearance of the problem. Unfortunately, I suspect that it may be a case of waiting for the problem to reappear because your description of the symptoms does not provide a finger pointing at one obvious probable cause.

You have my sympathy. Intermittent faults can be like a giant Sword of Damocles waiting to drop on you as you are speeding up on the on-ramp.

Thanks for your extensive analysis Old Guy! I thought you might jump in :smile: Some responses:

"If the MIL came on and stayed on until you turned the key off and then the car restarted normally with the MIL staying off, that is odd." Yes that's exactly what happened. Cruising along at about 70 MPH in 5th, engine loses power like it was switched off concurrent with check engine light coming on, all gauges functioning normally, check engine light stays on until ignition switch turned off, then on restart everything functioned perfectly normally, including the other idiot lights going through normal sequence. Odd indeed. The ONLY symptoms were complete loss of engine power, and the check engine light on. The everything functioning completely normally, immediately, on restart.

The service manual for the pre OBD II ECUs has this wonderful rather obscure disclaimer in a footnote to the error code table:
"The Check Engine light may come on, indicating a system problem, when, in fact, there is a poor or intermittent electrical connection. First, check the electrical connections, clean or repair connections if necessary."
I read into this that an intermittent condition could cause the CEL to come on without an error code. The footnote does not elaborate on whether the light may go out by itself. Perhaps if there is no error code the CEL resets on restart? Its possible you had an intermittent fault that killed the engine and caused the CEL to light. The part of this scenario that does not hang together for me is that if the problem was truly intermittent, why didn't the engine restart itself when power was restored as the car was coasting down in speed (unless you had disengaged the clutch quickly prior to power being restored). Unfortunately, if the problem resolves itself its pretty much a wait until next time.

On page 11-27 of the fuel section of the service manual, it addresses the problem where the CEL comes on and stays on with no stored codes. However, this section seems to be for the scenario where the light stays on all the time. Still there are some things in the trouble shooting flow chart that it might be useful to have a look at as possible sources of trouble.


"The ignition switch could also cause this problem." Yes, I may see if I can recreate the problem by fiddling with the key and the connector while the car is running. But a failure there wouldn't cause ONLY the check engine light to come on would it? Wouldn't all the dash lights come on?

You are likely correct. The ignition switch has two circuits IG! and IG2 so they can fail independently. However, the ECU and the cluster gauges and warning lights are both on the same IG1 circuit so an intermittent failure should affect both systems. However, the self test on start up for the ABS, TCS and SRS lights is controlled by the ABS, TCS and SRS modules, not the gauge cluster. I believe that the ABS and TCS modules are off of the same IG1 circuit; but, if the power interruption was brief enough it might not have triggered a restart on those modules.

 

[Wavey1]

I love it when people do my homework for me :biggrin:

Thanks for the additional info Old Guy, and my thoughts exactly.

 

[Wavey1]

Ran through a few things today. First, I was mistaken about the battery cable - that is Redbird's GROUND cable that I replaced. The positive cable is original and there was a bit of corrosion at the post. I thoroughly cleaned everything (would really like to avoid replacing that) and replaced the battery today so that's all good. BTW, on another recent thread a Group 24F was recommended because of higher CCA than the Group 35 batteries. I did that and the car does crank and start quite a bit faster. About 2 extra pounds but not a big deal as it's low and in front.

Before I replaced the battery, I ran the codes again and this time I got a code 47 which indicates a fuel pump issue. I'm thinking I didn't have the jumper in full contact the first time. I cleared the code before the battery replacement.

Ignition switch appears to be nice and tight - didn't have any issue jiggling the key around.

I also cleaned the connectors and grounds at the igniter and the alternator, but the all looked fine.

I'm thinking the old battery (5.5 years) and the corrosion may have been the culprit. I'll drive it and see if the problem reoccurs.

Questions: My understanding is that the fuel pump relay is incorporated in the main relay (my main relay has been replaced btw). But the diagnostic pages indicate that there is a separate fuel pump relay. See page 11-101 of the manual. It looks like it's near the main relay, and from the parts diagrams it looks like it's part number 39797-SE0-003, "Power Relay Assembly." Is this a separate fuel pump relay from the one in the main relay?

Further down in the diagnostics is a Fuel Pump Resistor, the finned part on the firewall at the passenger side. Is this part commonly know to fail?

The relay and resistor are cheap enough I may replace them for good measure.

Now, the fuel pump. The OEM part is about $400. Anayama in Japan doesn't have the OEM available, and almost all other sources are various aftermarket brands, most in the $50-$100 range but are not the complete fuel pump kit and require some fitment. However I ran across a source that has the complete Denso part that from the photo and description looks to be identical to the complete OEM pump assembly with the fuel level mechanism, filter, seal, everything, for a ridiculously low price. The listing states it's '96-'01 only but the OEM parts sources show all years are the same. I think I'll order one on the off chance it is bonafide, but if I never need it its so cheap it doesn't matter. Normally I really like to use OEM parts wherever possible.

Comments welcomed.

 

[Old Guy]

You are correct about the fuel pump relay; but, in Honda/NSX speak, the fuel pump relay is not the relay that turns the fuel pump on and off. That is, as you note, carried out by one of the two relays that are inside the main FI relay. I have attached a diagram of the fuel pump circuit. As you can see, there is a fuel pump relay; but, the purpose of that relay is to short circuit the fuel pump resistor when the engine switches into Vtec mode. This increases the voltage to the fuel pump making it run faster.

Error code 47 is a bit of a mystery. As far as I can figure the ECU is monitoring the voltage on the fuel pump and if the pump doesn't get the expected voltage (high or low depending on engine operating condition?) it generates error code 47. That is my guess from reading the trouble procedure since the '91 service manual is not explicit about what triggers the code. So, a problem with the fuel pump relay or the fuel pump resistor might trigger error 47. A bad connection might also trigger error 47 which may have caused your engine quit so it is worthwhile checking both components.

I wouldn't just replace if things look OK. As far as I know the fuel pump resistor is fairly reliable (its only on when the engine is in Vtec mode). You do read posts about people by-passing the resistor; but, that is because their fuel pump is getting old and failing to make adequate fuel pressure. By-passing the fuel pump resistor applies full voltage to the pump which can provide a temporary fix for low pressure as long as you are not operating in Vtec mode.

Rock Auto sells a Denso pump for the NSX for around $100. You are correct that it is not the complete fuel pump / sender assembly. The filter sock is available as a separate piece for less than $10. Those are the only two pieces you really need if it is just a pump issue.

 

[Wavey1]

Thanks again Old Guy. Yes, I worked through those diagrams and was thinking the same thing. The Rockauto Denso with filter, and probably the gasket - was thinking of that too, should be a direct replacement. The other complete pump source turned out to be NA anyway. I'll go through the troubleshooting for the relay and resistor next, seems simple enough. Thanks, appreciate your brainstorming!

 

[Wavey1]

Did about 60 miles today and no issues. I'm hoping the old battery with a corroded positive terminal was the issue... we'll see.