Ping Mrcheerful

Does it go flat as quickly after a proper external charge?

You seem to have done things correctly including noting that many cars take a while to go to 'sleep' these days and that 20mA is about normal quiescent current. With a good fully charged battery it shouldn't be a problem for several weeks. So either something is taking *a lot* more current - or the car charging system isn't charging the battery properly.

Have you measured the battery voltage with the engine running? Preferably with a low battery - it could be it simply isn't charging the battery fast enough, possibly down to a slightly iffy battery. Under the same conditions check for voltage drop between alternator output and battery and alternator body to battery negative.

One fairly common reason for a battery discharging for no apparent reason is a short circuit diode in the alternator rectifier. But this is unlikely to be intermittent so would show up as a constant drain on the battery of about 5 amps.

You can try leaving the meter in circuit overnight set to min-max & then check it in the morning, obviously reset it after 20 minutes so you know it was quiescent & check it before you unlock the car.

Dave Plowman (News) was thinking very hard :

It has built in diagnostics and you can set it to monitor all sorts of inputs, one of which is the charge voltage as you drive down the road. It has never varied between max and minimum load from 13.8v with the engine running. Even whilst towing and power the extra lights and the fridge in the caravan through the night it stays at 13.8v.

The old battery was not that old, about two years old and one of the better ones and to the car makers spec. - but I have replaced it anyway and with one of a larger size, as recommended by other members of the cars club. It was completely flat when I took the old one off, but now it has been given a full charge it seems it is fine.

My own thoughts are that it is sometimes not going into sleep mode and that is the cause, but it is so difficult to catch it in this state.

Just what I thought. 13.8v will take a long time to charge a low battery. You need more like 14.4v.

Snag is a heavily discharged battery can have altered 'chemistry' which the car charging system doesn't work properly with. But a decent external charger might well recover it to somewhere near full capacity.

And of course most decent max/min DVMs go to sleep themselves...

I've got a couple of decent min-max meters - but both do an auto power down after a short time. So not long enough to log something overnight. I do have a mains benchtop one which doesn't, though.

Dave Plowman (News) explained :

I have a min/max meter, but it would maybe take weeks of logging to actually catch it misbehaving.

I just thought there might be something made for the job, as it seems many do suffer the same problem. All it needs is an indication of the the slight load.

I think your tests are on the right lines and what the others have said is all valid.

Is the replacement battery you have fitted a good quality new one of the correct type for the vehicle? and is it giving the same problems? (you haven't actually said)

The drains you mention could not flatten a good correctly charged battery overnight, so if we assume that the drains are being measured correctly then that leaves battery and/or charging.

Battery is best checked by substitution with a new one (I have only ever come across a couple of faulty new batteries (cheap makes) in many hundreds and never had a faulty top brand battery at all (bosch or similar)

Charging is best checked by results during and after use. 13.8 is absolute lower end of charging, 14 - 15 is far more normal.

In your case I would put on a fully charged battery and leave the car for as long as you reasonably can and see what the battery voltage has dropped to, it should stay over 12.6 for weeks. If it isn't then something is pulling more than you are measuring. disconnect stuff till you isolate it. Does it have a tracker on board?

If it stays up fine with the car out of use then it would appear that there is a charging fault. Probably replacement alternator is needed.

You haven't said what vehicle it is, perhaps there is something model dependent that is common.

Mrcheerful laid this down on his screen :

The old one was a 60aH 600cca Varta and its spec. was as per manufacturer and about two years old. It charges up fine off the car, and holds its voltage was over 13v off charge for 24 hours.

The replacement is a new Bosch 72aH 680cca fitted yesterday morning. The extra capacity was recommended by other owners due to the mass of equipment fitted on the car. Obviously I have not had it long enough to find out if it suddenly goes flat yet.

Agreed, but at this stage I think the drain is only happening sometimes. I fancy something is not going to low power mode just once in a while.

Did that yesterday.

I have always considered 13.8 to be spot on, but perhaps I am wrong there. What is puzzling is the fact that it can go for days/weeks/months without any problems, then it will suddenly have a completely flat battery. Flat as in not enough charge left to even pull the starter solenoid in and this can happen overnight. It never has any middle ground of just barely cranking it over enough to start it, it really is all or nothing.

All the driver can leave on accidently is the parking lights, which would be noticeable in the dark garage, or the interior lights/glove box/boot light etc.. The latter all go off via a battery saver relay after leaving the car undisturbed for 20 minutes and I have checked this works correctly.

Even if 13.8 is not giving it a full charge, it doesn't explain the suddenly completely flat battery.

Normal is for it crank over quickly and start/run at the first pull. It is not the starter, the battery really is flat as confirmed by the dash doing an instrument reset and the lights also being dead.

It now hovers around the 14.06v mark on a full load at a little over tick over. Its a 100amp output BTW.

As I originally asked - how would an auto electrician go about tracking down a discharge which just happened sometimes? Is there an instrument able to pass cranking amps, but still able to measure down to a few milliamps?

Thanks for the help!

the cranking current would exceed any (practical) current meter abilities, but you wouldn't need to monitor that, just the side loads are the bits to really look at.

the slightly bad connection on the alternator could easily be the cause of your problems.

an exanple of the model specific faults is: nissan micra certain models: Battery flattening problem due to software programme which continuously rechecks that everything is off, using power in the process. Can be re-programmed not to and problem disappears

!!!!!

Mrcheerful

Well a tinytag will do it but they're £80 so unless you really want to know...

I suspect you've found it then. That sounds like the right charge voltage

You use an autoranging multimeter or datalogger to measure the voltage drop on the earth wire, if you don't want to know about the start current then you use it on the feeds to the battery box

None I know of - you usually use a clamp type to measure high current and they ain't accurate for small amounts. An easier way would be to measure voltage drop on the battery earth connection - a very sensitive device will show low current flow while not being damaged by high. Maplin used to do a kit for this. I've got one on the SD1 feeding a tri-colour LED - shows amber with no current flow, green on charge and red on discharge. But can be used to drive a centre zero meter too. But not sure how you'd log from it.

I've got the circuit somewhere if you'd like a copy.

That's also about the figure a very flat batt will show on a car with engine running after been jump started but won't be putting any juice back into the batt. If it's too flat a modern(ish) vehicles charging system won't attempt to charge to save damaging alternator.

Dave Plowman (News) formulated the question :

I'm really not too bothered about logging, I was just looking for something basic like an LED to light if the current (or rather voltage drop across the main feed) should exceed a set amount.

Yes please, if its not too much trouble.

I am though looking at designing something myself, like an FET op/amp comparator circuit to connect in parallel with the main feed and turning an LED on when the current exceeds a set level.

No problem - what file type would you like? I'm not using a PC to scan it so can only use the universal ones. TIFF, JPEG, GIF? And is your address valid?

on 22/06/2009, Dave Plowman (News) supposed :

jpeg would probably be easiest for me. My email is valid, providing you remove the NOSPAM - thanks.

Have you checked the usual suspect - a boot or bonnet bulb staying on even tho shut (switch broken). This might explain intermittant - sometimes its off, sometimes not.

Sent.