Civic battery drain mystery

Here's the situation:

-'95 Civic DX hatch -d16z6 swap (w/ 120k on motor) -battery died--thought battery was just old and replaced it -new battery starts draining after a few days -headlights go dim (most of the time)--passenger side is brighter than driver's side -according to trickle charger, one day it's way down, one day, it's fine -just had alternator tested and it puts out 12v

SO... Can a failing alternator work intermittently or do they tend to just die?

If it's not the alternator, where else should I be looking? What kind of potential loose connections are there to seek out?

I'd much rather avoid a mechanic if I can.

Thoughts from the gurus???

Japanese cars often have bad grounds or ground cables that go bad....especially the older ones, tho your saying one headlight is brighter than the other is a "stumper".

I had a 280Z that would "cut-out" for what seemed like no real reason. That is, you would be driving along and as you changed gears the engine died. I finally traced it to a corroded ground for the fuel injection (by that time, the corrosion was getting so bad you could see it). I'm not saying this is your prob, but check the grounds and the negative cable coming off the battery.....as a start.

That's a good start. Thanks.

I had the same thing happen with an 89 Civic Si. Replaced battery but the new one died. Replaced the alternator with a rebuilt one from Parts America. Battery died again. I gave in and took it to the Honda dealer. They said the alternator was bad. I told them it had just been replaced. Turns out, I bought a bad new alternator.

I have since found out that this is not uncommon. A new Honda alternator did the trick for another 60,000 miles.

I'll bet the alt is not good.

the alt, when charging should put out 13.6v or so..

I had a similar problem with mom's D16Y8 swapped wagon. Tried replacing the alternator, which contains the voltage regulator, which didn't completely solve the problem. I then replaced the fuse box like thingy on the right front strut tower that contains another odd and little known device that controls part of that system that I have never seen go bad before and who's name escapes me after 3 cape codders...Id say try another alternator first (I may have a spare in the garage, you are free to check (under the big workbench in a box).

Cape Codders in PA. How poignant.

Thanks for the offer--I'll probably just get one new, anyway..

Oddly enough, I'm suspecting something in that fusebox, a) because the alternator was working at full capacity (whether it's always doing that is the $64k question) and b) the negative cable runs into it.

However, I'm as proficient a diagnostician as I am an impressionist painter.

Gad, I love old cars...

do you have a multi meter that can read amps? I have seen a faulty internal voltage regulator that worked perfectly when the engine was running, and then instantly began draining the battery when the engine was shut off.( this was on an industrial engine, not a honda). so if you have a multi meter, hook it up in series to your battery and positive cable (with the engine off and ign swithch off), and see how much amps the drain is pulling out of the battery. the amount of voltage drain should be low, less than an amp (the clock, stereo and other stuff that is still powered when the ign is off will cause some amperage drain, but it is normaly very little). If the amperage reading is more than that disconect all the wires on the alternator and check the meter again. If the drainage is still there start disconecting other stuff, its a proccese of elimination. there is a potential problem with method though, the fault could be intermitant and not acting up during testing. good luck!

HappyAndy is right on the money. Most DVOMs have a 10A test circuit which works well for this. Most times the 'parasitic draw' (radio, PCM memory, etc) is .25 amp or so. Connect the DVOM as described, if you have a draw first disconnect the small plug from the alternator (not the big battery lead). If the draw changes appreciably, I'd replace the alternator.

If the alternator plug does not change the draw, then reconnect it and remove fuses from the fuse block one at a time. If one changes the draw a lot, investigate that circuit and don't be surprised if there's a lot of different weirdo stuff on the same circuit. I had a weird one a while back, a Dakota pickup with a pretty strong (~1.2 A) draw which turned out to be the tachometer. Now just why in the hell does a tach need full time +12V? The funny thing: the tach worked just fine.

I'll make another "simple" suggestion. Are you using the correct alternator for the motor? I swapped a B18B into my B18A integra, and fried three alternators in as many years. Though the motors are REALLY similar, the alternators are different.

A good alternator should put out 14+ volts at 1500 rpm (engine). It has been found that many of the re-built alternators sold at the chains are defective. Maybe it has something to do with the China syndrome.

I use a simple 12v test light connected in series with the battery cable to find a draw.