Battery went flat while driving - what's the problem?

[Rob Bell]

With a lovely weekend, I wanted to get my dear old MGF out on the road to enjoy good old fashioned roof-down motoring. I haven't really done this for the last year, principally because I live in the ULEZ zone, and a '95/'96 MGF is not compliant - and that is a £12.50 per day charge. It puts you off casual use - which I guess was the London mayor's intent...

After a year sitting outside the house in the elements with minimal use, my MGF was grumpy - which is unusual, because it is generally just amazingly reliable. Unless the battery is flat. Which was kinda the problem that I am starting this thread to discuss...

Several problems cropped up as I was going round to revive the car from its slumber. I had disconnected the battery to prevent battery drain during the period of disuse. Annoyingly, the bonnet release cable has seized. That was a problem not only in terms of battery re-connection, but also because I needed to pump up the hydragas system (despite replacement, re-gassed hydragas spheres from a well-known supplier) - again, for the third time. This needs investigating as well, but there are other things to check on first.

Opening the bonnet without the cable release is do-able - but for obvious reasons, I'll not detail the process here. However, because of modifications to my car, the usual approach was not possible, but I was able to do it. With the bonnet now opened, I lubricated the release mechanism and pumped up the suspension. And I also spent about 20 minutes clearing out all the accumulated leaf mulch...

With the battery re-connected, obviously the MG just started first time on the key, like it always does. However, some dash lights no longer work, including the indicator dashboard turn signal tell-tale lamps (the left one works intermittently), and that is despite the external lamps all working as they should - I suspect that there has been corrosion/dirty contact somewhere - and this may or may not be relevant to the subsequent flat battery issue.

Then there is the EPAS, which turned on its warning light. Often this leads to the steering being over-assisted, but there was no assistance to the steering. Effectively, the EPAS was switched off. The warning light and assistance would then come on and off intermittently - again, I suspect that this is possibly because of a dirty/oxidised connection somewhere because there appears to be some correlation with road conditions (bumps etc). Further investigation needed.

However, everything else was fine, and had a lovely short drive on the Saturday for an evening meal. One odd thing on that journey: the battery charge light came on - flickering dimly, but then going out if I increased engine speed. Then a bump in the road seemed to coincide with it going out. Hmm. Odd - another dirty connection? But everything was running fine, so I adopted the "ignore it until it becomes a problem" approach.

Yesterday (Sunday) I decided to drive my MG to see my dad in Wiltshire. Much longer trip, but I figured it would help shake out any remaining issues. I was, as it turns out, quite right that it would!

Beautiful sunny morning with clear blue skies. Little K-series engine burbling away, the first stop was to the petrol station.

First new problem: the fuel cap lock is amazingly stiff. It needs cleaning/ lubricating - any suggestions?

Filled up - and EPAS decided to start working again. Happy days! Good omen? One observation though: although brimming the tank and the engine at operating temperature, neither of the dashboard gauges were reading what they should - being a fraction lower on their respective scales. Odd. Poor earth somewhere?

However, after driving on the North Circular for around 15 minutes, I was enjoying basking in the sun and remembering why I loved my MG so much, I noticed that the SRS, then the EPAS and then the ABS warning lights were coming on in succession over a period of a couple of minutes. I had a very bad feeling about this.

On Project Shed a few years ago, I lost the alternator belt that had shredded itself, and that the shredded belt took out the alternator with rubber fragments. Charge light came on, but the car still ran. I finished the day's competition and thanks to very friendly co-competitors, I had borrowed battery packs to nurse the car back to my father's place so that I could replace both the Alternator and the drive belt. My observation from that trip was that the engine ECU (MEMS1.9) is the least sensitive ECU to low battery voltage: the other ECUs controlling things like ABS etc, throw in the towel first - and I reckoned (correctly) that this was what was happening here - a cascading series of ECU faults due to low voltage. Strangely, however, there was no battery charge warning light lit, which caused mild initial confusion, but with all the other signs, I was sure that battery charge was the problem.

I turned around immediately and, I kid you not, the engine stopped as the battery voltage finally dropped too low for MEMS just 100 metres away from my house - and I was lucky enough to have enough forward momentum to turn off the main road and coast straight into the same parking spot I had left 30 minutes earlier.

I LOVE my MG! It gave its all to get me home.

That is as far as I have got in terms of investigations, as I took another car to go see my dad. Next weekend will be the time to really start investigating and sorting out a number of issues that I have identified. It's a long list, of which the battery is just one

To me, this sounds like either a bad battery connection or a bad alternator. The fact that the engine stopped suggests bad alternator, as the alternator ought to be generating enough power to run the engine even in the absence of a battery.

The weird thing however is that the charge light that transiently lit the previous day was conspicuous by its absence. Odd. However, the alternator will need a charge light to actually charge the battery, so there is a chance that there is a bad connector to the dash, that led to the non-operation of the alternator that in turn has led to dead battery while driving. I don't know if this is the answer yet, but it is one of the scenarios in my mind.

Plan of action:
1. Get the bonnet open again (cable still seized - not sure if I can clean this up and getting it working again, or whether it needs replacement? I'll need to remove the bumper to look at this properly, but of course I need to open the bonnet to do that!)
2. Before removing the battery, turn on ignition to see whether the battery light comes up on the initial dashboard light check - if it fails to come on, then I think that'll be the answer.
3. Charge the battery
4. Re-fit battery, and check battery voltage at rest and with engine running - the former should be 12V on a charged battery, and 14V (give or take) and a charging battery with a functioning alternator. If the voltage remains at 12V with the engine running, then the fault lies with the alternator.

That should get me a long way on the diagnostic route. I'll likely need to remove and attempt to clean all the dashboard multi-plugs too (all the external lights and turn signals are all working as they should, so I think a dash connector is the most likely issue). Hopefully this is something simple, but if I need to replace the original to the car (27 year old) alternator, I do have a spare in the garage

Any other thoughts on what the cause might be or things to check? More updates to come as I work through this problem

[mgtfnut]

Just a thought. My TF has lived outside all it's life and seems to be fine as I can run a trickle charge to it.
Last year I was concerned that the ecu indicated battery voltage rarely crept above +/- 13v when on the move. I checked the main suspects whilst changing the starter motor, and noticed the +ve cable, and the earth cable on the other side of the block to the rear boot bulkhead has corrosion of the thick bare copper cable to the point when wiggled, loads of individual strands broke! I ordered tinned strand cable to replace the original but never got round to lying on a wet road
I only got round recently to making a new -ve battery to body cable complete with glue lined heat shrink.
The earth arrangement at the rear with a fuse? looks very dubious between the block and rear bulkhead as well.

Worth checking for corrosion and voltage drops?

[Rob Bell]

Thanks Jerry - yes, I think there are quite a lot of electrical connectors to look at!

Now I've finished the kid's learn to drive MGZR, I now have a lock up to keep my F in, so once I've got everything working again (and it gets a fresh MoT), I'll be storing in a dry garage from here on in

[Chris Tideswell]

It does sound like a dead alternator, not sue if it is the same on an early F but on the TF the complete absence of the battery/charge light means a dead alternator. When my TF’s alternator packed up I didn’t notice the light but the car drove without issues right until the battery ran out of electrons.

[Deepfat]

Just a thought have you checked behind the dash or elsewhere for rodent damage? the pesky blighters seem to love insulation.

[Rob Bell]

It does sound like a dead alternator, not sue if it is the same on an early F but on the TF the complete absence of the battery/charge light means a dead alternator. When my TF’s alternator packed up I didn’t notice the light but the car drove without issues right until the battery ran out of electrons.

The alternator on the MGF is an Italian Magneti Marelli, not a German Bosch - but it has made it to 27 years and perhaps the rectifier has given up the ghost - and interesting that your alternator failed with the absence of the charge lamp. Perhaps that was what the dimly glowing charge light the day before was telling me??

Just a thought have you checked behind the dash or elsewhere for rodent damage? the pesky blighters seem to love insulation.

Good thought - I don't think so, but with some time over the weekend, I'll be inspecting the dashboard wiring and connectors, and if one of our little murine friends have been having the midnight munchies, I'll likely find out then!

[mowog73]

One thought Rob. With the car running, if you do not find alternator voltage at the battery, yet alternator voltage at the alternator, it may be that the fusible link has given up the ghost. This had me scratching my head for the longest time until I looked at the wiring diagrams on Dieter's website.

[Rob Bell]

Well... this turned out to be a MUCH bigger job than I was expecting. It's taken me days, but I now have a charging battery.

Thanks for that thought Mark - a really good one actually, but happily it wasn't the fusible link. And I haven't changed the alternator either.

Are we sitting comfortably? Then I'll begin the story...

[Rob Bell]

Day one.

Having a flat battery and a bonnet that doesn't open is a bit of a pain, so the first task of working out why the alternator wasn't doing its thing wasn't to look at the alternator (although I could confirm that the alternator belt was as fit as a fiddle just by peering into the wheel arch), but rather the first thing was to get the bonnet open and to get the remote cable release working reliably again.

Again, I'll not post the process of releasing the bonnet latch here, and while a practiced thief could potentially pop the bonnet pretty quickly, it took me a thick end of an hour to get into the bonnet space thanks to other modifications to the car. That'll teach me to mess with me motor...

Once the bonnet was open, I could have another look at the latch mechanism. I had sprayed everything in oil on the previous attempt - but I figured where it comes to penetrating oil and corroded mechanisms, more IS more. I also sprayed the inner cable and got my son to pull and push back the boot compartment release cable. The cable started moving, and the release mechanism of the latch started to free up. The release mechanism is still far from perfect: the return spring is insufficient to return the latch back to the lock position without also pushing the release cable back in, but at least the release mechanism is now working as intended.

With that out of the way, I removed the battery and put it on charge overnight.

[Rob Bell]

Day two.

I replaced the battery and fired up the car - the engine bursting to life instantly on the first turn of the key - the way a K-series always does.

The bulb check was as before, with notable absence of the ignition charge light, oil pressure light and ABS light, in addition to absence of the main beam indicator and turn signal lamp (although all non-dashboard functions were normal, bar the battery charging).

To confirm what I already knew, I checked the battery voltage with ignition off and with the engine running. On a freshly charged battery, the resting voltage was 12.6V, and with the engine running, 12.6V.

Alternator not charging.

I have a spare alternator, but before attempting to swap this, the lack of other warning lamps on the dash made me suspicious that there was likely dirty connections to the dashboard - and if the alternator doesn't have a charge light, it won't charge the battery.

I figured I would start with the dashboard first. As this thread isn't really a how-to, I'll not document the process of removing the steering wheel and instrument pack here, but it is pretty straight forward and I've done this many times.

Having pulled out the instrument pack, I removed the three main wiring connections. I happen to have a fourth - an extra bulb that serves as a dashboard warning of low coolant level (Brown and Gammons coolant level sensor - the original and still going strong).

Looking at the terminals, all looked fine. I cleaned away any oxidation I could find, and sprayed the terminals in contact cleaner (which reminds me - I need to buy another can - no prizes for guessing why...)

Putting all the connectors back into the instrument pack, and there was no change - everything was as before. Strange.

I couldn't work it out. So I looked at the back of the instrument pack to inspect the printed circuit, removing and cleaning the bulb connectors and checking that the bulbs hadn't all blown simultaneously. None of the dash bulbs had blown.

One thing did stick out - looking at the rear of the instrument pack to the right, there was a nick in the circuit, partly severing it. By the time I looked at it and prodded it a bit, the track was completely severed. This track I believe serves as a common earth for the instruments - and likely explained why the fuel tank was not reading completely full despite me having filled it up that previous evening.

I then had a think: what was going to be the best way to repair this damage? I was about to solder the track back, but then realised that I had a spare printed circuit on a spare VVC instrument pack I had lying there as I had taken it out to see whether the spare instrument pack behaved the same way as the one in the car had - it did - so the damage to the printed circuit was not the cause, but it needed to be replaced, and the problem would appear to be in one of the wiring harnesses or its connections... To repair the original instrument pack, I simply removed the old printed circuit by removing all the bulbs and screws -

And then swapped over the "new" printed circuit to the original instrument pack -

[Rob Bell]

By now, you would have thought that this would have fixed the problem. A fresh printed circuit and cleaned up terminal connectors. But no. the problem persisted - although now the fuel tank DID read full. So that was one problem sorted.

I then had a long look at the subharness between the instrument pack and the fuseboard and the main loom connector. I think I had routed this
section of wiring badly when I replaced the main body wiring loom harness when I swapped this over to retrofit ABS to my MGF. My bad: the sub harness had some creases in the wires' insulation, but no obvious breaks. Hmm. Nothing too obvious here - so there was nothing for it, but a methodical checking of circuit continuities, referencing back to the wiring diagrams (I have a lovely, period A3 book, but these diagrams are also on-line on Dieter's website).

There are, as mentioned, three main connectors - the two where I was suffering issues are connector 228 and 227. The pin order I worked out from the diagrams - and from which I could work out continuity.

Oddly, a number of circuits when straight to ground with bulbs removed and plug 228 connected... This was becoming seriously strange.

I checked for shorts in the subloom, but could not find any - except between two green wires, which I found very odd. Both carried +12V from the fuse box, but why should they be connected together at the loom? I decided to take a punt and buy a replacement second hand instrument pack sub-loom and put the interior back together for the night (so as to not loose any screws or other bitty things!)

[Rob Bell]

Day three.

The "new" instrument pack sub-loom had arrived, and I excitedly unboxed it and checked circuit continuity. To my disappointment and general confusion, there is still circuit continuity between the pair of green wires?!?

I felt somewhat dejected, but decided to replace the original sub loom with this one - you know, because why not?

In Ground Hog Day style, the dashboard was stripped down once again and the instrument pack removed, new-to-me harness plugged in and...

The ignition charge light came on! Yay! Also, so did the other indicator and also the main beam indicator - dimly at first, but it became brighter after a while (for what reason, I can only guess?)

The oil pressure and ABS light however remained unlit. Odd.

Back to the charge light! Would it extinguish if I started the engine? It blinked off, and then came back on. Interesting. Voltmeter confirmed non-charging battery (now 12.55V). If I revved the engine, the light would extinguish, but then defiantly come back on again when the rpm dropped to idle.

Faulty alternator then?

[Rob Bell]

I don't find working on MGFs too difficult (not compared to current generation cars anyway - far too little space, far too much packed into a small area), but even I get bored of lifting the back of the hood, removing the roll hoops, taking out the parcel shelf carpet an insulation and then un-bolting the inspection cover.

My first surprise was the discovery of a wire floating in space that hilariously I've never noticed before - because I have never had a charging fault on this car before!

This is the cable for the airconditioning air compressor that my car does not have. Okay, relax - the alternator only has two connections - a great big brown power cable and the ignition charge lamp cable (which is much more dainty).

If I was going to remove the alternator, I had to remove these cables - and besides, it would be a good idea to ensure that these had not corroded.

On removing the charge lamp terminal connector, from the alternator, the connector fell off the wire!

Well, that would not have helped my cause - and is another potential cause for a non-functioning charge lamp!

Does anyone know what this type of connector is called?

I'd like to order a replacement, but for now, I have crimped on a simple loop connector which seems to be working satisfactorily.

The charge light still came on with the engine running at idle, so I still had the original charging problem.

I figured I'd just swap over the alternator. But I had figured without taking into account my 4-2-1 exhaust manifold. However fabricated it for me did a really nice job, but Dave (for that was his name) did not think about how to remove the alternator 20 years later. The issue is that the upper bolt cannot be withdrawn far enough to drop the alternator: I would need to remove the exhaust manifold. And looking at the primary pipe design, getting to the bolts (not studs!) holding the manifold in place was going to be a nightmare. Plus, I did not have a new exhaust manifold gasket to hand.

So rather than muck around with the alternator, I decided to check the connection of the brown cable to the starter solenoid. Getting to this is a little challenging, so I didn't remove the retaining nut completely, and decided to reseat it. But I did re-start the engine at this point.

What the...

The charge light went out - and now the battery was charging at 13.88V!

[Rob Bell]

So a long story short, the problem was a bad connection. Or rather, a bunch of bad connections. Some of those bad connections directly contributed to the non-charging battery, whereas others were likely bystanders, just waiting in the wings to become a future cause of a non-charging battery.

It's only taken three days over a period of a week to get that far - and I still have two misbehaving warning lamps. Fortunately, the ABS unit is working normally, with no recorded faults on pscan and normal live data values. So this warning light no-show is likely a bad connection too - and is particularly strange, as it was clearly working before as it heralded the impending flat battery! I feel a bit exhausted now - so will wait and see whether it decides to wake itself up before investigating further, but the more immediate need is for the MoT, and a non-functioning ABS light is unlikely to cause issues.

[Rob Bell]

Just an up-date on this thread - just in case this is of any use to someone in the future.

The problem re-emerged - the charge light once again came on and failed to go out. This took some while to get around to repairing, but when I got down to it, I found yet more problems. Perhaps the most significant being that at some point in the past - unbeknown to me - some one at a garage or workshop had attacked the engine loom. No idea when this happened - but some long while ago I suspect. As a consequence of this loom assult, I found two problems: (1) someone had replaced the brass nut on the starter solenoid power stud with a steel flanged nut that had corroded, and (2) the alternator to starter cable had been cut and the cable then repaired by a spliced-in cable extension with a new cable of a fraction of the 16cm2 cable that was original (maybe 6cm2?). All of this was out of sight (and out of mind!) beneath two layers of flexible cable protection. I only found this out when attempting to remove the corroded nut from the starter solenoid main power cable post...

This picture is NOT from my car, and shows a large brown cable that is conspicuously not seen on my car (this connects the starter to the alternator - the main power cable from the battery to the starter is hidden beneath the black plastic conduit) -


The story plays out thus...

The battery charge light came back on. I was initially able to get this circuit working once again by wiggling the alternator to starter cable connection. Funny how this type of "repair" never really lasts that long on cars...

Therefore, when the battery charge warning light came back on, it was not a surprise and I thought I knew what I needed to do. Having recently replaced the starter motor on Project Shed, I thought this would be a quick 15 minute job. Hahaha - this was one of those 15 minute jobs that took around 4 hours...

The first problem encountered was that the retaining nut on the starter power stud would not undo. Because I had fitted a DIY Rover 820 air box for my K&N air filter many moons ago, access from the boot to the starter connections (by far the most convenient access point to the starter) was extremely limited. I get my hand and arm anywhere close to the starter motor required me to remove the engine ECU and move this to one side. Ah, the joys of modifying cars... However, I was eventually able to get a 15mm socket onto the terminal nut, but the nut simply would not undo - I did not want anything off or break the starter motor (which, afterall, was working fine!). It would later turn out that the nut was steel rather than the brass it should have been - and I think a combination of dirt and corrosion basically made it very difficult to remove. I sprayed oodles of penetrant spray over the nut and terminal stud and left that to stew for a while.

Because my MG was parked a little way down the road from my house, I thought I would try the "wiggle the cable terminal" trick again to see whether I could get the car started in order to drive it a little closer to my house. That's when the terminal at the end of the alternator to starter cable snapped off. Great.

The alternator-starter cable was a little short to get a new loop connector crimped on easily, and the cable was shrouded in cable protection tubing (not quite sure what to call it - it isn't typical semi-flexible plastic conduit, more like over-sized heat shrink and a fabric-like tube slid over the power cable - I hope you get what a I mean). To address this, I pulled on the cable to try to get better access to the end of it and pulled off the two cable conduit layers to discover a quite a dramatic drop in cable size from the aforementioned 16cm2 to something well under 10cm2 (6cm2 may be?) This really can't be ideal: the step in resistance brought about by the smaller diameter cable must surely limit current flow when the alternator is meant to be charging the battery? Luckily, it hadn't caused a problem until now. Moreover, the wrapped cable tape over the crimped connection connecting the two mismatched-sized cables was loose/damaged the crimp and copper cables had gone green thanks to oxidation. That can't be ideal either...

The task of simply cleaning up the connector to the starter solenoid power post had just snow-balled big time.

Luckily I did have some brown 16cm2 power cable (purchased when I re-positioned the battery on Project Shed), plus a 16cm2 loop/ battery connector and a 16cm2 copper crimp ferrule. This was good, but the real problem was location - the lack of length of the existing cable and the general lack of access when some fool had fitted their own air filter solution and retro-fitted a TF rear bulkhead cross brace meant difficulties accessing the area around the back of the engine and gearbox. I have some nice scratches to show for my efforts, but at least I could see what I was doing

Sadly, no photos to show and luckily I did not launch my YouTube career filming myself, as quite a lot of choice words were used in the process of attempting to crimp on the new cable extension using a tool that, while easy to use, is large and rather bulky. But I got there in the end...

Because of the time spent mucking around making the cable repair and the need for multiple cups of tea to vent frustration, I had plenty of opportunities to repeatedly douse the stubborn terminal nut in penetrating oil and plenty of soak time. Eventually, the nut started to turn, but was incredibly hard to turn, and the terminal stud itself became loose, moving perhaps 15degrees in either direction. I was worried that I would need to replace this starter motor too...

But fortunately not. Because I had not thrown away the original, broken Project Shed starter motor from 10 years ago (never throw anything away!), I had a spare, correct, brass nut and spring washer - these were used to retain the main battery cable and the alternator charge cable to the starter solenoid, and this nut, in contrast to the corroded steel one that I had just removed, went on beautifully. I was then able to jump start the car (remember that my poor MG had previously been running without an operating charge circuit) and the MG just purred, and the charge light finally extinguished. Great! And even better, I had not ruined the started motor!!!

Hopefully this particular case is not a common scenario, but hopefully the thread as a whole provides some helpful clues and pointers to look for if the battery charge light remains stubbornly lit! Occam's Razor suggests that a single fault, but Hickham's dictum suggests that many faults can co-exist, especially as our cars slowly age!