Car Won't Stay On

[Rufus]

Hi everyone,

I'm having an issue keeping my car "on." For some reason, this morning, my car's desk does not light up when I put the key into position 2. The climate control turns on, but in the past, all the dash light would come on too. When I move the key to position 3, the car would start. But unlike before, when they key snaps back into position 2 (the normal driving position for the key), the car shuts off. If I keep the key in position 3, the car will continue to run after starting, but as soon as I let go of the key, the key snaps into position 2 (as it should) and the car shuts off.

It seems like there is something that is supposed to power up in position 2 that is no longer doing so. Perhaps the alternator?

It could also be an issue with my ignition key system, but since the radio turns on in position 1 properly and the climate control turns on properly in position 2 and the car starts properly in position 3, I think the ignition bore is probably ok.

Any thoughts?

Thanks.

 

[Old Guy]

Because your car starts up, my initial gut reaction would be that your ignition switch has failed. However, it can also be a problem with your fuel pump resistor. The ECU by-passes the fuel pump resistor during the engine cranking phase to speed up fuel pressure build up during starting. Once the engine starts and you release the key back to 2 the ECU opens the relay which by-passes the resistor. If the resistor has failed open then the engine will stop when the key is turned back to 2.

The ignition switch has 2 separate circuits in the run (2) position. The fact that the MIL light is not lighting up when you turn the key to 2 tells me that the circuit that supplies the ECU and other critical stuff is dead. During cranking, the ECU gets power from a different circuit which is why the engine starts when cranking; but, quits when you release the key to run. The service manual (you can get a .pdf copy via the Pime Wiki) has the test procedure for the ignition switch. You can sometimes fix the ignition switch by taking it apart and cleaning the contacts. However, its not the easiest thing to get to so you might just want to plan for replacement rather than repair. Not super expensive if you order from one of the on-line discounters.

 

[Rufus]

If I understand you correctly, you are suggesting that I test the ignition switch as described in Section 23-72 of the service manual. It looks like a continuity check for each of the ignition position. How is this done? Do you use an ohm meter and see if the two terminals are shorted to each other when moved into each of the key position?

And if I don't get the shorts (connections) that I expect based on the diagram/table provided, I can conclude that the ignition switch has gone bad.

I will probably replace it rather than trying to clean it. You mentioned on-line discounters. Do you have a specific site you recommend? Should I buy it from official Acura on-line parts site?

Lastly, I might be able to get a hand on an ohm meter (or buy one), but is there another way to test this without one?

Thanks.

 

[Old Guy]

Yes, the correct procedure is the one set out on 23-72 and an ohmmeter is the normal go-to solution for doing continuity testing. In a pinch, you can use a simple continuity tester made with a light bulb, a battery and some wire or get something like this:

https://www.princessauto.com/en/detail/continuity-tester/A-p8817512e

Probably as cheap or cheaper than trying to make your own continuity tester.

The critical connection for the ECU is the BAT to IG1 connection in position II. However, if any of the connections fail the test in the diagram on page 23-72, the switch is due for replacement. Also check fuses 29 and 13. If either of these is blown the engine will stop in the II position; but, if either of those is blown its an indication of a different problem which will likely be harder to fix than the ignition switch.

Buying from an authorized (official) parts reseller is a good idea. If you can find one close to you then you may be able to avoid shipping charges. I have used Delray, in part because they were the only US authorized seller that I found that would ship to Canada.

https://www.acuraoemparts.com/

Rockauto sells aftermarket copies of the ignition switch. So far, I have found Rockauto to be a reliable supplier. Lately I have purchased my NSX parts from THENSXshop and Amayama.com. Both are in Japan; but, its usually cheaper for me to order from them than the US because the brokerage and shipping fees are lower (go figure!). You just have to be prepared to wait 5 weeks for delivery. If you are in a hurry the dealership might be the way to go.

 

[Rufus]

I checked the continuity of the connections this morning. This is what I observed:

Regarding connections to ACC:
In position 0: all open (good)
In position 1: BAT-B should connect to ACC and they do (good)
In position 2: BAT-B and IG2-B should connect to ACC, but IG2-B does not (bad)
In position 3: all open (good)

I'm not sure what is powered by IG2-B, but no power is going to that node when key is in position 2. This makes me think that the ignition switch needs replacement.

This also seems to be aligned to what I am observing. In position 2, I am expecting the dash lights to come on, but they do not. Perhaps IG2-B powers that system.

Regarding connections to BAT-A:
In position 0: all open (good)
In position 1: no connections to BAT-A (good)
In position 2: IG1 and IG2-A should connect to BAT-A and they do (good)
In position 3: IG1 and ST should connect to BAT-A and IG2-A should disconnect. IG2-A does disconnect (good) and IG1 stays connected to BAT-A (good), but ST does not seem to connect with BAT-A as expected (bad). Not sure why. The car seems to start OK on position 3 and will even stay on if I force the key to stay in position 3. So not sure why there is an open between ST and BAT-A in position 3...

But either way, since no position brings power to IG2-B, I think the switch needs replacement.

What do you think? Am I on the right track?

 

[Old Guy]

I checked the continuity of the connections this morning. This is what I observed:

Regarding connections to ACC:
In position 0: all open (good)
In position 1: BAT-B should connect to ACC and they do (good)
In position 2: BAT-B and IG2-B should connect to ACC, but IG2-B does not (bad)
In position 3: all open (good)

I'm not sure what is powered by IG2-B, but no power is going to that node when key is in position 2. This makes me think that the ignition switch needs replacement.

This also seems to be aligned to what I am observing. In position 2, I am expecting the dash lights to come on, but they do not. Perhaps IG2-B powers that system.

Regarding connections to BAT-A:
In position 0: all open (good)
In position 1: no connections to BAT-A (good)
In position 2: IG1 and IG2-A should connect to BAT-A and they do (good)
In position 3: IG1 and ST should connect to BAT-A and IG2-A should disconnect. IG2-A does disconnect (good) and IG1 stays connected to BAT-A (good), but ST does not seem to connect with BAT-A as expected (bad). Not sure why. The car seems to start OK on position 3 and will even stay on if I force the key to stay in position 3. So not sure why there is an open between ST and BAT-A in position 3...

But either way, since no position brings power to IG2-B, I think the switch needs replacement.

What do you think? Am I on the right track?

Your test results are very confusing. However, to compound the confusion, I have discovered by checking my NSX Electrical Trouble Shooting manual that the BAT-A to IG2-A (white wire) switch function is not used for anything on the NSX, despite being included in the test procedure. This helps

View attachment 161425

explain this diagram which only shows 4 switch outputs from the ignition switch.

As shown on the attached diagram, in switch position 3, the BAT-A connection to ST (blk/wht wire) provides power to the starter motor circuit. Since your starter motor was engaging that connection had to be working prior to the testing. So. assuming you were doing the test with the switch still installed in the car, four possibilities:
- the ST wire connection at the plug or the switch was damaged in the disconnection process
- your tester is not making a good connection with the ST wire in the plug
- position 3 is spring loaded and perhaps your were not turning the key 'all the way' for position 3 (enough to complete IG! to BAT-A but not complete ST to BAT-A)
- The ST to BAT-A function of the switch just happened to fail at this time

The BAT-A to IG1 connection is the critical connection for operation of the ECU and related mission critical stuff. Your test indicates that BAT-A to IG1 is connected in both position 2 and position 3. This suggests that if the car can get started in position 3 that it should remain running in position 2. So we have test results that say that the car should not start (no ST connection in position 3); but, if you did get it started it should continue to run in position 2. So the test results don't jive with the problem condition. You could go back and repeat your tests to see if there was a testing problem; however, this is an ignition switch and when they approach their end of life their operation can become erratic leading to inconsistent test results. I normally dislike replacement in the absence of diagnosis; but, since your car is a 1995 and particularly if it is a high use car, I might be inclined to treat the ignition switch as a consumable and just replace it. The inconsistent test results are consistent with a switch that is behaving erratically which is what they do at their end of life.

 

[Rufus]

My fault. After reading your post and looking at your diagram, I was trying to figure out why my readings were so off. Then I noticed the "View from wire side" note next to the plug diagram! I measured everything mirrored!

So doing a quick translation, my continuity check of BAT_B and ACC is still valid since those pin positions are mirrored from each other.
That is also the case with BAT_A connection to IG1. And that connection was also good.
So basically the two valid continuity check I made both show that the ignition switch is working.

I need to remeasure BAT_B with IG2_B and BAT_A with ST when I get home tonight.

But according to your post, the critical connection is between BAT_A and IG1. But that connection seemed "OK." Unless the resistance is a factor. When I measured BAT_B with ACC, the resistance was 0.2 ohms - basically a solid short. But when I measured BAT_A with IG1, the resistance was ~200 k-ohms. It's not open, but there is resistance... Perhaps the ~200 k-ohms is an indication that something is damaged? Should I also be seeing near 0 ohm for this connection too?

Also, do you know what IG2_B is used for? Your diagram gets cut off and I cannot see where it's going.

My car is a 95 with under 40k miles. I would say that it's fairly low mileage and would normally not expect the ignition switch to go out so soon. But perhaps the fact that the car is 24 years old make any failure a possibility.

Let me know your thoughts on the ~200 k-ohm measurement between BAT_A and IG1. Thanks.

 

[Old Guy]

But according to your post, the critical connection is between BAT_A and IG1. But that connection seemed "OK." Unless the resistance is a factor. When I measured BAT_B with ACC, the resistance was 0.2 ohms - basically a solid short. But when I measured BAT_A with IG1, the resistance was ~200 k-ohms. It's not open, but there is resistance... Perhaps the ~200 k-ohms is an indication that something is damaged? Should I also be seeing near 0 ohm for this connection too?

Also, do you know what IG2_B is used for? Your diagram gets cut off and I cannot see where it's going.

My car is a 95 with under 40k miles. I would say that it's fairly low mileage and would normally not expect the ignition switch to go out so soon. But perhaps the fact that the car is 24 years old make any failure a possibility.

Let me know your thoughts on the ~200 k-ohm measurement between BAT_A and IG1. Thanks.

200k ohms qualifies as a open circuit for this particular test! 200k ohm across the connection would limit the current flow through the switch to 70 micro amps at 14 volts which is definitely capable of powering absolutely nothing! For future reference, resistance on a functioning switch contact pair should be less than 0.1 ohms after correcting for test lead resistance. Your switch definitely requires replacement.

IG2_B supplies a bunch of things. Starting at page 23-52 in the service manual you will find a complete breakdown of the 12v DC power distribution system for the car. Useful for future diagnostic forays.

I would not expect 40 k on a 1995 to represent high use for an ignition switch. However, did the car sit unused for any period of time? Long periods of non use can result in oxides forming on the switch contacts leading to poor connection. With regular use these oxides get scraped off. Since your problem seems to have materialized all of a sudden, I don't think it is an oxide formation problem. Every equipment expected life data set has a distribution of values and perhaps yours just happened to draw the short straw!

 

[Rufus]

I redid the measurements last night - keeping better note of the resistance value.

All but one connection seem to be functioning correctly for each of the key position - all showing 0.1 ohm. The only issue I see is the connection between BAT-A and IG1 (and also IG2-A which is not used in the NSX) while in position 2. That connection shows a resistance measurement between 60k-200k ohms. It does vary each time I click to position 2 within that range. When I move it to position 3, the connection between BAT-A and IG1 becomes good again (0.1 ohm). So this seem very much inline with what I have observed - the issue is only for position 2. Everything seems to work properly otherwise, including starting the car in position 3. But once key returns to position 2, IG1 is not making a good connection with BAT-A and system shuts down.

So I will be ordering a new ignition switch today to swap it out. Thanks for your help.

 

[Old Guy]

I think you have pretty much confirmed that your engine quit problem is related to the ignition switch and the failing BAT_A to IG1 connection.

If you are at all inclined to fiddle with stuff, after you get your new switch installed try taking apart the old switch. The operation of the contacts will be quite obvious and it might be obvious why the BAT_A to IG1 connection is failing. Others on Prime have reported resolving switch problems by taking apart the switch and cleaning and lubricating the contacts and sometimes bending the wiper to apply more contact pressure (the latter is probably a higher risk venture). Clean-up of the switch might give you a working back-up, not that you are likely to ever need it.

 

[Rufus]

Looking up the part information on the NSX Wiki

https://www.nsxprime.com/wiki/Ignition_Switch

The part # is: 35130-SL0-A01

But when I go to the Acura sites, the part they seem to sell is 35130-SL0-003

Looks like the same part. Wondering if anyone can confirm from experience.

 

[Rufus]

Checking into this, looks like Acura updated its part #'s. A01 and 003 are the same part.

Ordered it from Delray since you had good experience with them and they were $30 cheaper than my local dealership (same delivery time).

Hope this solves it!

 

[Rufus]

Well, same delivery time did not end up working out. I got an email from Delray saying that the part is coming from a CA warehouse and that it will take 3-5 extra days to ship (on top of their original 3-5 days estimate). Not sure why shipping from CA would double the shipping time. So I waited and when I saw that the part hasn't shipped after 2 weeks, I called Delray. It seemed that the part hadn't actually been ordered yet. After finally ordering it, it took two days for them to get it and two days for them to get it to me. Overall, it took 3 weeks to get this ignition switch, but it seems like they could have gotten this to me by their original promised timeline of 3-5 business days if they had actually ordered the part for me...

Oh well, at least I got the part and it fixed the problem. Before I installed the new switch, I checked the continuity of all the switch connections and they were all good. Once installed, the dash lights came on in position 2 and car started with no issue. I drove it for about an hour last night after the install and I had no issue with the start, keeping the car running and the repaired EPS unit (which I was finally able to test) was good too. Finally, everything back to how it should be.

Let me say, though that getting those two screws off was a challenge. I had to contort my body just to see what was going on under there. But well worth the effort when the car started.
[MENTION=26435]Old Guy[/MENTION], thanks for all your help with this. When I get a chance, I will open up my old switch to see what it looks like and why position 2 was not making a good connection. I'll post photos if there is anything important to share.

 

[Old Guy]

That is just poor customer service. I had used Delray in the past; but, perhaps I will drop them to the bottom of my list and try Amayama or the NSXshop first.

Excellent news that the switch did the trick. That would have been a double burn to wait 3 weeks and then find out that the problem was someplace else.