What would cause a car to not start when it's humid ?
 

OK guys, I have a puzzling problem with my '93 Civic DX :
Like clockwork, my car will fail to get electricity to the fuel pump when there is even the slightest bit of humidity out.
If there is even a single dark cloud on the horizon, the car will fail to start.


If i turn the key to the ACC position ( the second click of the key before the third click that starts the car ) the fuel pump does not come on. When i remove the main relay and trip it, the fuel pump comes on fine.

The main relay is a brand new unit.

I have replaced the fuel pump with a brand new pump


The distributor might seem to be the problem, since the car has very weak spark, but this doesn't explain the way the fuel pump acts by not coming on when i turn the key.

I am replacing the distributor this weekend regardless.

The only other thing i can think of would be a short in the ignition switch wiring. I posted about this problem on a honda forum about a week ago, but only got one response, so im posting here too just for the hell of it.

I was told that the ignition switch could be tested by checking the main dash lights that come one as you start the car.
They all check out fine, so i'm puzzled.

Dirty bad connections on the battery cables. Bad grounds. Bad connection in the fuse box. Sometimes what you are facing is one of the most difficult electrical troubleshooting problems possible. Intermittent loss of a connection.

In extreme cases I have run a test light circuit to the fuel pump power lead. If you have no light when the pump should have power you know it is in the power circuit to the pump. If you have a light but no pump you know it's in the ground circuit.

Once you establish which circuit is failing you can move the test light connection back alone the power circuit until you lose your power. This is if the problem is in the power circuit to the pump.

If it's the ground circuit then you need to use a DVOM to test for a voltage drop along that circuit.

Your bad connection will be where you lose the test light in the power circuit, or where you see a significant drop in voltage in the ground circuit.

Most common problem I have seen in Honda's is the ground wire from the engine to the chassis is either missing or has a bad connection, corrosion at either end. Could also be a bad ground at the fuse block. Use the DVOM (digital volt ohm meter) to check from the battery to the various places, like engine, body dash, steering column, etc.

Also pay close attention to the ignition switch. Not sure about Honda's but Nissan's were notorious for bad electrical parts of the ignition switch. So bad I can almost remember the Nissan part number, even though haven't worked on them in over 10 years.

I think it was a 48750 E7705 for the Z cars.

regards
Mech

My dad had starting problems on his Chevy pickup when it was humid outside. He'd take off the distributer and clean off the points and it would start. Eventually that didn't work anymore. Turns out he had a cracked coil. Most of the time it was fine, or at least good enough so that clean points would make up for it.

Thanks guys.
I suspect a short in the wiring in the steering column, just because problems always seem to appear in the hardest possible place to get to eh ?

I'm going to change the entire distributor this weekend, as well as the battery cables and grounds.

If you KNOW those parts are bad, or have reached the limit of their life expectancy then replacement is not a costly hit and miss guessing game.
Part of being a professional in business is to treat the customers money the same way you would treat your own.

A cracked distributor cap could be your problem, but before I replaced the distributor, it would have to have excess play in the shaft, or a bad ignition module, or other issues that made replacement the best solution. Bad ignition modules are usually heat related, but if a bad cap is the issue you should be able to see evidence of moisture inside the distributor.

Intermittent problems are seldom solved by throwing parts (read money) at the problem. Testing circuits for voltage drops starts at the battery terminals. Hook one end to the negative terminal (of your DVOM) and measure the voltage drop to parts of the engine and body of the car. If you see more than .25 volt (with systems operating) then you are on the track of the problem. Moisture can be added to the equation by spraying a mist of water on the car or engine with a hose, but don't drown it, just a light mist will do.

Lets say you have a bad fuel pump. If so you will show with a test light, which can be made out of two 20 pieces of wire, two alligator clips (preferably insulated) and a peanut light bulb. Hook the clips to the terminals on the fuel pump and if it does not come on and you have a strong light, then you KNOW it's the pump or corroded connections at the pump.

I always had a very hard time when a customer accused me of a hit and miss repair with no regard to their cost. Proper diagnosis is not hard, usually just tedious. Wiring is a lot like plumbing and on older cars the diagnoses are fairly simple.

I had customers ask me why I replaced the water pump when I did the timing belts on a 84-89 300 ZX, especially when the water pump was not leaking. I told them if the water pump failed it would destroy the timing belt and possibly the engine. That was why we learned the hard way and replaced the cam seals, cranks seals, tensioner, and the timing belt as a part of the job. It was a systematic approach that gave them the best chance of not having to go into that system for the recommended 60k miles between replacements.
Seen to many bent-broken valves, holes in pistons, and pieces of piston skirts coming out in the oil drain pan.

regards
Mech

Thanks again Mech, I really appreciate the advice.

I will try the testing that you suggest.

I changed the fuel pump because it seemed to show all of the symptoms of a dying fuel pump that i had read about. Same with the main relay.

I had previously changed my cap and rotor, as well as installed new wires, but i still have weak spark, so i thought "what the hell", i might as well replace the whole thing.
At least that would hopefully help with the weak spark issue that i am having.

It's been raining today, and there is moisture in the air.

Is it a bad idea to change the distributor if there is dampness in the air still ?

( If extremely mild humidity can kill a car dead, i thought it might be a good idea to ask before i change the part with dampness still in the air. )

Actually since your issues are dampness related it would be best to see if you can resolve the weak spark issue in the same conditions when it occurs. Just be careful you don't get ZAPPED! :eek:

What is your battery voltage reading both running and with the ignition off?

Engine to chassis ground could be the weak spark. Check for a voltage drop from the positive battery terminal (part on the battery itself) to the connection where the big cable meets the starter. The check for a voltage drop from the engine itself to the negative battery terminal.

Voltage drops of over .25 volt indicate there is additional resistance in the circuit that is not supposed to be there, Just take the leads and go to the parts and touch them with the meter set on volts (not ohms, it could fry your meter or blow the fuse). Think of a voltage drop as a plugged plumbing pipe and you have low water pressure. Starting at the battery, as long as it reads about 12.5 volts (not running) and 13.8 to 14.2 running. Then go to the first place the main battery cables go where there is a connection. The voltage should remain the same. If the voltage drops you have probably isolated your problem especially if it is more than .25 volt. Unintentional added resistance in the circuit reduces to power available to the electrically driven components. Low battery voltage can really make things go crazy, all kinds of stuff not working properly. The battery may be the problem, or the pathways electricity need to do the job could be the problem. It's usually at the points where things are connected to other things, especially in areas where corrosion is a problem.

Keep me posted and I will see if I can walk you through it. Believe me I have been there and done that, including buying cars that were partially submerged and fixing them so they could go another 10 years, but that topic could take pages.

I have had people get mad at me for not tackling these types of jobs, because there is no way you can estimate the unknown, just plug away and isolate the area where the problem is. You start at the battery and work outwards, fuse block, grounds. Remember the engine to chassis ground which is usually just a wire that runs from the valve cover to the radiator support. If you don't see one the put one on. The volt meter will show you it needs one if you just touch the engine and a bolt on the support that is plated. Scratch the bolt head to make sure you have a good connection. If you see a voltage reading from the negative battery terminal to the engine itself, you know there is not a good ground, add the ground (even a simple wire with two alligator clips, or just the bare wire ends under two bolts and eliminate the voltage reading it might fix the weak spark.

I'm not going anywhere and there are others here who know just as much, so you have the advantage of our common experience. When I was 16 it took me week to figure out plug wires could be bad, the second time it took about 5 minutes. Stupid graphite core wires!

regards
Mech

Again and again thank you !
I'll try and do some testing this weekend unless its raining ( no garage )

Of course i will have new questions that pop up since i have a voltmeter, but have never done testing with it.
i'm not even sure how to conduct testing with it actually. ( I will do a quick review online to see how it is done.

I should add that my engine to chassis ground wire has been modified. I snapped the bolt that holds the valve cover on and so i have it attached to a something ( metal ) on the block itself and leading to the chassis.

That ground should be fine but it's easy to check it. With the car running one lead on the negative battery terminal the other on the engine metal, No voltage it is OK.

regards
Mech

Mech, this is the DMM I will be using.
I lent it out and no longer have the manual.
I'm not making sense of all the settings on this thing. It is currently set to ' 20 ' on the ' DCV ' section ( just below the ' back light ' button .
The leads are plugged in to the ' V ? mA ' and 'COM ' outlets
( the black lead being plugged into the ' COM ' outlet of course.
Digital Multimeter Manual

I just tried out the multimeter with the settings I had above and it shows 12 .1 which I assume is volts.
Not running of course since its sprinkling out here now.
It also shows 12.1 at the fuse box ( both wires coming from fuse box )


EDIT. I was doing it wrong .12.1 amps are now showing at the starter .

I am getting current at the pump.
Reading is 11.88 and the picture in the chiltons guide shows 12.25 .

When I try a reading with the main relay installed , versus jumped , it goes to 0.10 for s split second then nothing.

I have the probes inserted into the fuel pump power supply connector.

12.1 at the battery compared to my Maxima that has not moved in 24 hours reading 12.45. The chilton manual reading of 12.25 is higher than your battery voltage. I don't think you are seeing any problem though as the volatge should be sufficient down to less than 10. Possibly older battery?

You are not reading any voltage with the relay installed versus bypassed? If so then the car should not start with the relay installed but start with it bypassed? That would be either the relay or the power source to swtch the relay.

The relay should have two different size wires, possibly a total of 4 wires (maybe more if it is multifunction). Check the voltage across the smaller wires, then across the larger wires. A relay is just a magnetic coil that works a higher power connection through a lower power connection. You may be able to take the cover off of the relay and actuate it manually. The relay should have power to the high tension circuit constantly. When actuated it should have basically the same current going in and coming out of the high tension part. The low tension part will have a voltage drop, not sure how much but the resistiance of the coil that moves the contacts together requires some power.

The relay is probably energized by the ignition switch. Anyway, if I understand your post it seems like the relay is not doing the job. What you need to do is check to make sure the relay is getting good voltage, enough to make it work properly. If you find it has good voltage then the relay is your prime suspect. You can even install a bypass and a switch to power the pump in an emergency, as long as you use the same circuit you have fuse protection. If you make a completely new connection then use an in line fuse for protection. If you have replaced the realy and nothing changed, the odds are you have an issue with the power getting to the relay to activate the coil inside of the relay.

regards
Mech

The purpose of a relay is to allow a low power circuit to engage a high power circuit. A great example is the starter relay which connects the battery to the starter motor. The starter can draw as much as 200 + amps of power, which would melt your ignition switch if that much power went through the normal ignition swtich. Basically the coil creates a magnetic field wich draws the points together to create the higher power circuit.

regards
Mech

I'm in the process of reading your posts right now, and i'll edit this if i miss something that is mentioned above, but i just wanted to mention that i did a very thorough testing of the original main relay on the car and it checks out fine. All terminals were checked for continuity. I bought a brand new main relay and it isn't getting voltage to the pump either, but i'd bet it will check out fine.

Regarding the battery voltage, it has been run down a bit with the car not starting.

[QUOTE=Old Mechanic;272979]

You are not reading any voltage with the relay installed versus bypassed?


Correct.

When i install the probes in the plug going to the fuel pump ( the red wire going to the yellow wire and the black wire to the black ) Nothing happens when i have the main relay installed, but i get juice when i have the main relay off and the circuit tripped.

I'll try again tomorrow just in case i did something stupid.

If so then the car should not start with the relay installed but start with it bypassed?

The car does not start with the main relay off ( circuit tripped ) and the fuel pump running. I have tried this several times.

That would be either the relay or the power source to swtch the relay.

With the testing that i just did that showed that the main relay is fine, im guessing it's the power source to the relay.


The relay is probably energized by the ignition switch.

So it indeed could be the ignition switch causing my problems. I used to have a wad of keys hanging on my ignition key and a guy warned me that this might be what the problem stemmed from, since it stresses the inside of the locking mechanism.

Anyway, if I understand your post it seems like the relay is not doing the job. What you need to do is check to make sure the relay is getting good voltage, enough to make it work properly. If you find it has good voltage then the relay is your prime suspect. You can even install a bypass and a switch to power the pump in an emergency, as long as you use the same circuit you have fuse protection. If you make a completely new connection then use an in line fuse for protection. If you have replaced the realy and nothing changed, the odds are you have an issue with the power getting to the relay to activate the coil inside of the relay.


regards
Mech
.

Basically that means you are not providing the power to energize the relay. This can be the power supply or the ground circuit to the relay. Use your volt meter to check the power supply and to check the ground circuit. Do this with the relay connected.

From the battery positive terminal to the positive terminal on the relay. If you read any significant voltage there is resistance in the circuit that should not be there. If you read no voltage it may indicate a ground problem or the circuit is OK. Remember voltage drop indicates the proper battery voltage is not getting to the point where it can activate the relay.

On the negative side connect to the negative battery terminal and to the negative side of the relay (on the part that is powered by the ignition switch, not the fuel pump power supply. Use a jumper wire with alligator clips if you can not reach both points without one. I usually use a jumper wire to the negative so it can not short out.

This is where a bad ground can be your issue, or a bad connection (probably a corroded plug in the harness). By going from the battery to the ground connection on the relay, your are checking the whole circuit. If either test shows a voltage drop, then you have to go to different portions of the circuit and isolate the bad connection.

The ignition switch would be my focus point in isolating the problem (assuming you find the place where the voltage drops occurs). The ignition switch is part of power supply.
The ground circuit is usually a black wire (but not always, best to get a diagram of the circuit-found on the internet) and print out the part you need.

Sorry if I missed any spelling-spell check is not working tonight.

regards
Mech

What a frekin headache !

I still don't get how that the slightest bit of humidity can kill a car when no moisture gets into the car at all.

( And if there is a problem...why does it only happen when its humid ??? )

Thanks for all the help.

"Circuit tripped" means you have made a connection that works without the relay in the circuit and the fuel pump works?

That can cure the problem temporarily but remember the fuel pump will run continuously, even with the key out of the ignition switch. Don.t leave it that way or it will kill your battery, but it should make the car drivable.

I have seen one Mercedes with 4 ignition switches replaced. The woman owned 100 vending machines and kept all of the keys hanging from the ignition switch. Once she separated all the vending machine keys from the key she used to start the car her problems ended.

regards
Mech

Oxidation in a connection is caused by moisture. As the connection gets worse failure will occur when the moisture content is at it's highest. The oxidation works like a sponge absorbing moisture which makes the resistance higher since pure water is an insulator.

While it is a headache, once you get the process down pat, you will never be in a position of allowing a similar issue to cause nearly the same amount of frustration. I have had some real nightmares, and it included a car my friend had rebuilt (collision damage) that would not start cold. After two dealerships wasted hundreds of dollars with throwing parts at the car in hopes of fixing it, he was going to junk it for parts, after selling it for $5000 and taking it back and refunding the customers money. He had a junked car out back that had the part needed to fix it sitting there. It just took me 14 hours to figure out which part it was, and 5 minutes to replace it.

regards
Mech

When you find the cause, it might be outside the passenger compartment, but in either case, humidity inside a car is basically the same as outside. You will get it figured out and learn that that car can't beat you that easily. The most important knowledge is not knowledge at all, just boneheaded perseverance. "I will kick that problems arse". The real question is how bad will it beat me up before it surrenders!

[QUOTE=Old Mechanic;272987]"Circuit tripped" means you have made a connection that works without the relay in the circuit and the fuel pump works?

Maybe i used the wrong wording. I place a jumper wire in the #5 and #7 terminals as shown in the Chilton guide. I hear the fuel pump stay on constantly but the car still will not start ( and there is spark, though weak ).

That can cure the problem temporarily but remember the fuel pump will run continuously, even with the key out of the ignition switch.

That's odd. I have to have the car turned to ACC to hear the pump stay on.

Don.t leave it that way or it will kill your battery, but it should make the car drivable.

As stated above, the car still will not start as odd as that is.

I have seen one Mercedes with 4 ignition switches replaced. The woman owned 100 vending machines and kept all of the keys hanging from the ignition switch. Once she separated all the vending machine keys from the key she used to start the car her problems ended.

I have since stopped using the wad of keys. ( around six months ago actually ) There is play in the switch, but not that much. Nothing that i would think of as 'loose'.

regards
Mech

The loose part is the mechanical portion of the ignition steering column lock. The electrical part (at least on Nissan's) was a separate switch mounted to the back of the steering column lock.

The weak spark may also be related to the ignition switch, so that points to the switch as the most likely suspect. In some cases when the mechanical lock gets sloppy (as you described) it MAY cause the connections in the electrical portion to not line up properly.

Most ignition circuits use battery voltage for starting then use some resistance to keep the voltage lower when the engine is running. Is the spark stronger when (and if) it is running assuming you can get it running.

When you have two separate things that are not working properly, then the most likely cause is the source of voltage to those two separate things. One would be the ignition switch, the other would be battery and or grounds. That not an absolute statement but a most probable cause. It's why these things can be so damn frustrating.

regards
Mech

I would not worry about the pump running with the switch on accessory and not when the key is out of the switch. I may be there is another relay powering up instead of the pump circuit going directly to the battery. My knowledge of Honda's is not that specific so for the time being don't get to concerned with that.

I have seen a leaky windshield dump water on the ECU and fire up the pump and the ground circuit to the injectors in a 300 ZX. When the car was towed to us 4 gallons of fuel had been pumped into the engine and it was liquid locked with fuel in the cylinders and the cranckcase was full of gas!

regards
Mech

I had the car running the other day and tested the spark.

It was identical to when the car will not start.

The battery cable was very corroded once i pulled back the rubber casing.
I replaced it and thought this would possibly be the cause of the problem, but the car just seemed to rev stronger.
I bought a piece of thick gauge wire and im going to ground the hell out of this thing !

I asked on the Honda forums about testing ignition switches, and they said that i should see the warning light come on and stay on for a second or two when i turn the key. Also, if i pull up the emergency brake, the light is supposed to stay on.
Are you aware of any other tests for the switch ? ( it passed the signal light testing )

I'll deal with this tomorrow. Thanks for all the help Mech.

The more I think about it the more I wonder why the pump would work when the ignition switch was in the accessory position. Honda forums usually have a couple of real pros who have more Honda specific knowledge than myself.

I would check the voltage at the positive terminal of the ignition coil, it should be close to battery voltage under the same conditions.

Also if you have the car running check the voltage at the battery to see if the charging system is working properly.

If you want to get the electrical portion of the ignition switch and it is not too expensive it might be a good shot to replace it. We had them in stock at my old Nissan shop so we could just plug in a new one and see if it solved the problems.

Or, you could check for voltage drops at the ignition switch (remove bottom cover of the steering column). A voltage drop is almost certainly the best evidence of finding the real source of your problem, since ignitions switches are not supposed to have any resistance to create a voltage drop.

You have determined that the fuel pump is not the only problem and the spark is weak. I think you have already replaced the distributor and plugs. The old plugs were cruddy and had a large gap. How about the ignition wires and the coil?

See if the coil wire has a good spark compared to the plug wires themselves. If so then its in the cap and plug wires. If not it's probably the coil or low voltage to the coil.

regards
Mech

Make sure both positive and negative battery terminals are clean (cables removed) and clean the inside of both cables where they contact the terminals on the battery. Also make sure the other ends of both battery cables are clean and tightly connected to their respective components.

regards
Mech

Today it rained and rained ...and rained, so I didn't do much of anything.

I tried to test the distributor, but rain was actually hitting the distributor ( no garage remember ?) So I quit for the day.

Off the wall question here : I have a SuperMID that has one wire connected to an injector. I removed it just as a test and it did nothing, but could something like that cause any sort of problems ? It looked like possibly a slight oxidation on the bare wire of the injector where i spliced in the SuperMID wire. I had it covered in electrical tape and I may just be seeing glue ( it was raining and i didn't look very hard. )

Today i tested the alternator, specifically the ICM.
It checks out fine with over 11.5 volts.

( battery is getting weak from the testing )

The car still fails to start with the main relay jumped. I can hear the fuel pump kick on and stay on, but the car just revs.

11.5 volts indicates a dead alternator. Running off the battery. Charge the battery if you can, get it over 12.3 volts after being off the charger for 15-30 minutes.

Did you test the alternator with the engine running? It should be putting out over 13.5 volts depending on the state of charge of the battery. Try jump starting the car with another good battery, but DO NOT run the engine or REV the engine on the good battery car.

regards
Mech

Some Other Dave " posted this on another thread related to this one here :

http://ecomodder.com/forum/showthrea...n-19273-4.html
Quote :
The MAP sensor should not be able to affect the inside of the distributor. It sounds like you may have some kind of bad contact in your ignition system? Double- and triple-check the plugs and wires, cap and rotor.

If the stuff you saw in the cap was reddish-colored, I would tell you to get a new distributor. The "red dust of death" is a known failure mode for those...

-soD "

Well guess what i found lots of under my leak cover ? Yup ! Reddish dust.

I have a brand spanking new distributor, complete with cap and rotor ready to install.
I paid $115.00 - much cheaper than a rebuilt unit from autozone .

BTW, I had just had the coil, rotor and cap replaced around 600 miles ago due to the weak spark issue.
It makes me wonder if the guys at the shop were just throwing parts at it.

( I'm glad i'm doing the work myself this time. )

Well, if I understand the drift of this thread well, it does seem you like me might be throwing parts at a problem that is deeper than it seems. I'm just an amateur, but in my current situation with my poorly running car the ignition system seemed to be the cause but now it seems to be the symptom. For a time the ignition problems seemed to be a symptom of a MAP sensor problem. I almost bought a new $95 MAP sensor, but I restrained myself, and did more testing. After testing, now the MAP sensor problems seem like a symptom of an ECU failure. I will test that tomorrow, hopefully, with a known-good ECU. If it solves the problem, maybe all is well.

Testing is cool.

And I love this quote from Old Mechanic: "The most important knowledge is not knowledge at all, just boneheaded perseverance."

I hadn't checked the alternator yet.
Sorry.Correction from earlier post.
Sorry. Correction : I ran a check on the ICM, not the ECU.

So how much of my problem could be attributed to the distributor ?
It still doesn't explain the main relay not working if the problem was in the distributor.
I did continuity testing on all the prongs of the main relay, and they all check out fine. The solder also looks fine.

Bump

Change your distributor cap. They get microscopic cracks in them and get moisture in side. My nissan sentra had that, my friends ford aspire had it. It's $15 at the most.

I don't understand. You already replaced the distributor cap once, maybe twice, yes? You have checked the alternator out?

The distributor cap, as well as rotor and coil have been replaced. The cap has been replaced twice.
Alternator is next on the list to check.

Good luck. I am still struggling and beginning to suspect my own alternator. Here are a couple tests I will try tomorrow: How to Troubleshoot a Honda Civic Alternator | eHow.com. And the ever helpful honda-tech moderator with a characteristic short and sweet recommendation: Honda-Tech - View Single Post - Alternator testing.