I have a 97 Accord LX, auto transmission with the 2.2L 4 cylinder engine. I've been getting a lot of no starts recently (nothing, no noise, no cranking).

Here's what usually happens. The car will start from cold every time. However, when it gets warm, and it has sat for a while (maybe 10 to 15 minutes), it won't start. When I turn the key to the start position the panel lights will stay illuminated but there is dead silence. It doesn't do this every time but it is becoming more frequent.

I'm suspecting it's the solenoid on the starter so I have begun to carry around a multimeter so I can look for power to the black/white solenoid wire when the problem reoccurs. My wife is driving the car locally so I have her primed to call me if she gets a no start. I started this routine a couple of days ago, and would you know it, the car has started to behave itself. However, I did catch one no-start of sorts. Here's what I found:

(1) I first tested battery voltage and got 12 point something.
(2) I then placed the multimeter negative wire on battery negative and the multimeter positive wire on the solenoid black/white wire.
(3) I instructed the wife to attempt a start.
(4) The voltage readout went to approximately 7.8 volts. However, I only saw this for about a couple of seconds because the solenoid eventually clicked and the car started. The car just decide to start, this time with a slight delay. I don't believe it's behavior had anything to do with my holding the multimeter on the terminal.

I'm waiting for more opportunities to test the solenoid voltage and am hoping for a true no-start (not a hesitation) so I can get a steady readout of the voltage. It will also give me an opportunity to whack the starter once I've finished the test to see if I get a start.

As I wait for the problem to reoccur I do have a question regarding the expected voltage at the solenoid. I was assuming I would get 12V and was initially surprised to only see 7.8V. However, since the solenoid is a coil I can see how there would be a voltage drop across the coil when energized. Does anyone have any background on what the voltage should settle at? If it should be closer to 12V is there some other area where I could be getting a loss of voltage? It might be that the solenoid simply won't pull in at such a low voltage. I'd hate to change the starter and still have the same problem.

BTW my battery is OK and the car will turn over well when it decides to start. Also, on regular starts I generally see the solenoid voltage run to the 7 point something range during the brief moment the key is turned. I am using a digital multimeter so there may be a lag in the voltage readout.

Thanks

 

Sounds like either a weak battery or a bad connection somewhere in the (+) battery cable. Look for dirty/corroded/loose connections, maybe between the terminal & the cable itself.

The starter is trying to draw current, that draws down the voltage. Seems like 7.8v is pretty low, so if you don't find anything flaky in the wires then get your battery checked someplace like AutoZone.

The behavior in your 2nd paragraph suggests it's not a bad battery, but an intermittent bad connection.

 

Funny you mention this, my car was doing the same several years back. Seemed to only occur at gas stations or quick trips to the grocery store. Turn the key and nothing. Probably happened to me 7-8 times over the course of a year. I did nothing to fix it but is hasnt done it again for 100k miles. I am with you on the solenoid, has to be the problem. GL and keep us posted.

 

Well, it happened again yesterday. My wife called me from a local parking lot and I popped over with my meter and hammer.

First, I checked battery voltage - 12.65V. Then I instructed my wife to turn and hold the key briefly in the start position. I got 7.6V at the solenoid and 11.65V at the battery. I then gave the solenoid a whack with the hammer shaft. The solenoid clicked (plunger pulling in) and the car started.

After the start, the check engine light came on. I'm thinking the system took exception to having the key held in the start position for several seconds. I'm going to take the car to a local Autozone to have the code interpreted - maybe later today or tomorrow.

While it looks like I'm dealing with a bad solenoid I'm a little concerned that I'm only getting 7.6V at the terminal. Current is obviously flowing through the solenoid but the plunger is not being pulled in. I'm clearly getting a volt drop (due to the current) somewhere upstream. According to my Haynes manual there is the Electrical Load Detector Unit, some fuses, the ignition switch, and the gear position switch. You would think that if all of these produced little or no resistance I would get 11.65V at the solenoid.

I have to satisfy myself that 7.6V is normal at the solenoid before I change the starter assembly. I will investigate more over the weekend - check O/C voltage + upstream wiring.

 

The smaller wire should get 12V. The source of the starter signal is the ignition switch. I am not sure what you are using for your ground when you test for voltage. I typically use a valve cover bolt, since they aren't usually corroded.

If you still get ~8V, I would disconnect the battery and clean all ends of both battery cables.

Most parts stores do a free battery/alternator/starter test.

 

I'm using the battery negative post. I doubt the battery cables are at fault since they are able to pass the high amperage to run the starter motor with total ease. However, I will be checking the various cables over the weekend, especially in the ignition switch circuit. More to come.

 

The low voltage at the starter while the battery voltage is OK - that suggests a bad cable connection. If it was the battery that was bad, then the battery voltage would match 7.6v.

Did you get 7.6v at the terminal of the BIG wire at the starter? Or the little signal wire? It might be that the little starter-signal wire has a bad connection?

Don't forget to check bad connections within the battery cables. I mean the connection between the end-lug and the wire itself. I've even seen corroded & almost broken cables in the middle, covered up by nice clean plastic insulation.

It started when you hit the starter. That's classic for a bad starter, or it's solenoid. Many people change the 2 together as a unit. If you don't want to do it that way, then dismantle the starter & solenoid to decide whether it's worn-out brushes in the motor vs. worn-out contacts in the solenoid.

The CEL is likely caused by the low voltage, not by holding the key too long.

 

"If it was the battery that was bad, then the battery voltage would match 7.6v." - Agreed, that's why I'm not suspecting the battery.

"Did you get 7.6v at the terminal of the BIG wire at the starter? Or the little signal wire?" - The signal wire from the ignition switch.

"It might be that the little starter-signal wire has a bad connection?" - Agreed. I plan to check the wiring from the battery, through the ELDU, the various fuses, the ignition switch, the transmission lock-out switch and through to the solenoid.

"It started when you hit the starter. That's classic for a bad starter, or it's solenoid." - Agreed. If it wasn't for the 7.6V I'd have a new starter/solenoid installed by now.

"Many people change the 2 together as a unit" - I will definitely go that way if I decide that the problem lies in that area. I'm not in the mood to take the old unit apart and fiddle with it.

"The CEL is likely caused by the low voltage, not by holding the key too long." - The code was an open or short circuit, or poor electrical connection, in the crankshaft position sensor. I've disconnected the battery and will reconnect later and see if the code crops back up.

More to come as I delve further.

 

Well, I eventually installed an exchange starter/solenoid. I couldn't account for all the anomalies I found but here was my thinking:

(1) During one of my no-start sequences I rapped the solenoid and it clicked and the car started. This was pretty much what I was hanging my hat on.

(2) The open circuit voltage at the black/white solenoid wire (detached from the solenoid) was full battery voltage with the key turned to the start position.

(3) After a while I realized that during a normal start sequence the voltage at the black/white solenoid wire would only climb to the battery voltage expected with the starter motor engaged. For a good battery that would be in the 9V to 9.5V range. I was getting less because my battery is a little long in the tooth. Granted, for the few milliseconds when the solenoid receives the signal, and before the starter motor engages, the voltage should be closer to normal battery voltage. However, my multimeter was not quick enough to pick this up.

The only observation I could not figure out was why the voltage at the black/white solenoid wire was so low when the solenoid was stuck and the starter motor was not engaged. I recall seeing a high battery voltage but a low solenoid voltage under these conditions.

I've had the new starter/solenoid on for 6 days now and have done quite a few trips without any problems. I'm keeping my fingers crossed that the problem won't rear its ugly head again.

 

Thanks for coming back with the new information. Hope it's fixed.