MG-Cars.net

Welcome to our resource for MG Car Information.

Recommendations

Parts

MG parts spares and accessories are available for MG T Series (TA, MG TB, MG TC, MG TD, MG TF), Magnette, MGA, Twin cam, MGB, MGBGT, MGC, MGC GT, MG Midget, Sprite and other MG models from British car spares company LBCarCo.

MG MGB Technical - Starter problems - starter motor not turning

Hi,

My previously very reliable'78 GT was laid up for the first half of 2013 and my attempts over the last few months to get it started have failed. At first I suspected a duff battery, but a replacement hasn't got it going. A volt meter suggests it's not the battery cables either. When I tried to start it in the summer I'd get the solenoid clicking, but even that won't work now. My last idea was to have the headlamps on and turn the ignition to see if the lights dim, but there's no effect.

I'm no expert (by any stretch of the imagination) but these symptoms are suggesting I need a new starter motor and solenoid. What do you guys think? Cheers.

A Jordan-Doherty

laying a car up usually creates some problems and makes existing ones worse, sounds like it's been laid up for a year now so you'll need to get it running and used to find the other wrinkles that'll be in it

very good idea to have a good battery, clamps, leads and earths

I think your car has a starter relay so I'd have a look at that to see if the connections (and its earth) are crudded up or damp

you might also be able to check the starter motor turns by turning it by hand using a spanner on the square drive
Nigel Atkins

Had the same symptoms on my '79 GT about 5 years ago and ended up replacing the starter solenoid. I now have similar symptoms on my '81 B; cleaning and lubricating the linkage between solenoid and starter cured it for a few weeks but now the solenoid just clicks with no response from the starter motor (although a second turn of the ignition key always starts the car). Yesterday I removed the starter motor/solenoid assembly from the GT and fitted it to the B - starts every time. The suspect starter motor is now on the bench awaiting a replacement solenoid from Moss.
Brian Shaw

This sounds v-e-r-y familiar!! Questions:

You turn the key to "ON" and what happens?
Do you have gauges, and accessories?

You turn the key to "START" and what happens?

Do you still have lights & horn? and no start?

My 79 roadster gave me pretty much the same issue this past summer and in the end it was a bad igntion switch assembly (the major culprit).

Other contributory factors were 14+ deteriorated connectors and a battery dated Dec.2000


Cheers

Gary
79 MGB
gary hansen

for a starter solenoid sometimes it's earth can be bad and need cleaning up

obviously all connections including the main lead to and from solenoid need to be like all electrical connections - clean, secure and protected
Nigel Atkins

The post script to my "Elelctrical Issue" saga is the car is still runing great and it starts up without a hitch. Thanx to Nigel, Paul, Les, Alan, et al for their valuable and well appreciated knowlege. Do keep us apprised of how you issue resolves

Cheers

Gary

79 MGB
gary hansen

Thanks for the advice guys. On turning the key half way, I get the ignition light, a bit of noise from the fuel pump and the fuel guage kicks into life. Turning it further produces nothing extra, not even a click from the solenoid, which I was at least getting a few months ago.

I shall have a look at the starter relay next and report back, probably at the weekend.

As for allowing it to stand for too long without attention, it's advice I've been given many times and always adhered to previously. In my defence, the birth of my son in February gave me other things to think about for a while. If I have another, I'll prioritise the car!
A Jordan-Doherty

sounds like your son is now old enough to help you with the car, get him to put a wet finger on the starter relay/solenoid as you turn the key and if he immediately fills his pants/nappy then you know you have a live connection there

probably best if you don't let his mother know about this though as they can be over protective
Nigel Atkins

Your symptoms are the same as I had with my 79. There were times it worked and suddenly while driving it just died. Then it would start after a while. Again, in the end the major culprit was the Ignition Switch Assembly.

Cheers

Gary

79 MGB
gary hansen

Just for example, it could be poor battery and/or starter terminals. Are they clean and tight? It could be the ignition switch as Gary said, or the relay, and it could even be the solenoid and starter, but that is the most unlikely cause in my opinion. Starting at the ignition switch, I would follow the circuit with a digital voltmeter. Shouldn't take long to find the culprit. Don't just go and buy new components on a whim. Find out what's wrong.
Mike Howlett

"My last idea was to have the headlamps on and turn the ignition to see if the lights dim, but there's no effect."

Just to be absolutely clear, are you saying that when you turn the key to crank, the lights don't dim?

If so, then the battery and it's connections are good, but power is not reaching the starter motor.

MGBs from 1970 had a starter relay on the inner wing. This should make an audible click, power the solenoid that makes an audible clonk that drowns out the relay, and the solenoid powers the motor which should drown out the solenoid.

So the first thing is to determine whether you are getting a relay click, or a solenoid clonk, or nothing.

If nothing then on the relay with the key turned to crank check you have 12v on the spade that carries the white/red wire, and 0v on the spade that carries the black wire.

If nothing on the white/red check back at the ignition switch and both sides of the multi-plug by the steering column. If you have 12v on the black the earth is bad.

If the relay clicks but the solenoid doesn't clonk check you have 12v on the spades that carry the brown and brown/white (or white/brown) wires.

If not on the brown the 12v feed to the relay is missing.

If on there but not on the other one then the relay is bad.

If on both then the relay is sending power down to the solenoid but either it's not reaching there or it's not operating for some reason. Do the same tests on the solenoid white/brown and the starter body as for the relay white/red and black.
Paul Hunt

Thanks for the advice everyone - much appreciated by one who isn't that technically savvy, but who is game enough to try to fix this without resorting to component shopping or expensive London mechanics.

Just to clarify my wording on using the headlamps - I can be clear that on cranking the engine, the lights do not dim.

Having had another listen in the engine bay (not easy when living on a busy main road!) I can report back that the starter relay switch is making the required clicking sound on cranking. I'm satisfied that the problem must be further on in the chain.

So what's my next move? I've hit the starter a couple of times with a broom handle, but that hasn't helped. I can't see how I could get a spanner in there or what I'd attach it to if I could.

I await further advice with my shiny new volt meter at the ready...
A Jordan-Doherty

see Paul's post below and his web site mgb-stuff (spanners)

sorry if your starter doesn't have the square I'm not good on remembering what parts are on which models 9or even what day of the week it is
Nigel Atkins

Nigel, the last year of use, of the starter motor that you describe that had the square shaft on the end, was '67. The '68 and later B used a pre-engaged starter with an integral solenoid. RAY
rjm RAY

Either I am confused about the problem, or you are confused about the terminology involved.

You said:
"Just to clarify my wording on using the headlamps - I can be clear that on cranking the engine, the lights do not dim."

I thought the problem was that the engine would not crank, so what you said does not clarify it to me.

Charley
C R Huff

RAY,
cheers for that but I fear it wont stick in my mind and it's a lot earlier than I thought

Charley,
you're technically ahead, me being non-technical knew what was meant first time, it is a non-start of engine, lights were put on to see if there was a draw on the battery from the starter components
Nigel Atkins

I'd try bypassing the starter relay first. Pull off and join together the wires from terminals 30 and 87 and see if the starter cranks then. It should be a round black one on the inner wing forward of the fusebox.


If it does the relay is prob duff, but do check the earth is good on the black wire and yr getting a switched live on the white/red.


If it still wont crank and the batt leads and engine earth lead is ok, the starter solenoid has packed up I would think.


Hope this helps.
SR Smith 1

Having checked all four of the wires to the starter relay switch, I've got just under 12v running through all of them when turning the key all the way. It's reading about 11.6v. Is that sufficient to engage the solenoid?

Apologies for the confusing terminology - you are dealing with a bit of an amateur here - thanks in no small part to this car having been extremely reliable over the 9 years I've owned it.

When I've said 'crank' previously, I've meant turning the key to the crank position. The engine itself hasn't cranked since about May!

Anyway, with the starter relay switch apparently functional, where should I look at next? Engine earth lead?
A Jordan-Doherty

Only three of the starter relay terminals should show 12v when the ignition key is turned to crank, not four. The fourth wire is the earth connection for the relay winding and should be at 0v. If that was at 12v the relay wouldn't operate, so as you say you can hear it clicking, I'm assuming that only three of the terminals do show 12v. Incidentally 11.6v is rather low if the starter is not being powered, should be more like 12.5 to 12.8v. That in itself is not likely to be the cause though.

If you can get your meter on the spade connector on the end of the white/brown wire at the relay, and that shows 12v, the problem is almost certainly down at the starter.

Check the spade terminal on the solenoid that the other end of that wire is connected to, and see if you have 12v there. There should also be a white/light-green wire on the solenoid, that is for the coil ballast bypass, and won't show 12v until the solenoid has operated i.e. clonked. The white/brown is usually thicker than the majority of the other wires on the car, including the white/light-green.

If you do have 12v on the solenoid spade that has the white/brown wire, then the circuit through the pull-in winding of the solenoid and the motor itself is open-circuit. When you first send 12v to the solenoid current flows through two windings. A very low resistance (i.e. high current) one that is in series with the motor - the pull-in winding, and a higher resistance one that goes directly to earth - the hold-in winding. When the solenoid operates, its contacts connect 12v directly to the motor, which bypasses the pull-in winding as well as powering the motor. This leaves current flowing through just the hold-in winding, which as its name suggests keeps the solenoid operated and the motor spinning until you release the key. The current through the pull-in winding of the solenoid is such that it also should cause a slight drop in the system voltage, hence a very slight dimming of the ignition warning lamp.

But if you do have 12v on the solenoid spade, but the solenoid isn't operating, then you are almost certainly going to have to get the motor off anyway.
Paul Hunt

Strange, I'm definitely getting a reading of about 11.6v on all four wires, including the earth. I don't know why that would be. Any thoughts?

I guess the next job is to get down to the solenoid and test those wires. Any advice on how? I don't see a way to them from above so I'm guessing it's a job for ramps.
A Jordan-Doherty

I had this issue, turned out it was the starter relay. Try removing the brown wire and the solenoid wire and touch them together. If it turns its the relay.
A P New

I've charged the battery up a bit more and I'm now getting 12.5v or thereabouts at all four of the starter relay switch wires.

I've also tried the suggestion above of touching the brown and solenoid wires together and that has not led to any further effect. I've measured 12.6v across those wires when connected.

Given the lack of further noises, should I be looking at the solenoid next?
A Jordan-Doherty

If you really do have 12v on the black earth wire then the relay can't be operating because the earth is bad. But then it wouldn't click, and neither would you have 12v on the white/brown going down to the solenoid. So I suspect your test method is not correct somewhere.

As AP says if you bridge the brown and the white/brown, and the starter cranks normally, then the problem is that the relay contacts aren't closing, but that could be because there isn't 12v across the relay winding, as well as the relay possibly being bad.

But if, as you say, nothing happens when you bridge the brown and the white/brown, then either the brown doesn't have 12v as it should, or that 12v isn't reaching the solenoid winding for some reason.

Ignore the earth wire for the moment, and connect your meter between the white/brown at the starter relay and clean metal on the engine somewhere. Ordinarily you shouldn't see any voltage there. Now turn the key to crank and you should see 12v, so what do you see under both those conditions.
Paul Hunt

for future reference and in case you have unsuccessful attempts at starting whilst you resolve this problem

a fully charged battery will help you in fault finding whereas a battery in low charge could hinder progress

so if you do a lot of churning then top up the battery charge
Nigel Atkins

I suppose it's possible I'm doing something completely wrong with my testing. Can someone advise the correct procedure for testing?
A Jordan-Doherty

Most tests are done with one probe of the meter on the terminal being tested and the other on a good earth point nearby.

Occasionally you might want to measure between two different earth points, for example the engine and the body, when testing the engine earth strap while cranking.

Also you might want to check that a particular terminal is at earth, in which case you would have one probe on the earth wire being tested and the other on a known 12v point such as the brown terminals on the fusebox.

However any circuit being checked needs to be carrying it's normal current, as only then will bad connection or open-circuit faults show up. An example here is the black earth connection on the starter relay. Normally it should always be at zero volts, but using a meter in the above test without having the relay powered from the ignition switch, could well indicate that the earth is good. However the current drawn by the meter is so small compared to the current drawn by the relay, you won't 'see' on the meter a fault bad enough to cause non-operation of the relay. Similarly testing connections with an ohmmeter is not a reliable indication, as bad connections by their very nature can be variable, and get worse the greater the current that is trying to flow.

When you power a relay with a bad earth, 12v comes through the winding onto the earth terminal, only some of the voltage is applied to the relay winding, and the rest to the bad connection. With your meter probe between the winding and the bad earth i.e. on the relay earth spade, you will see all or some of the 12v, depending on how bad the fault is, instead of zero volts. So when you said you could see 12v on the earth terminal of the relay when you had the key turned to crank, that tells me the earth was very bad and so the relay could not be operating.

Not for nothing is it known as 'electrickery', I've spent a lifetime playing with it.
Paul Hunt

Thanks for the explanation. You learn something new every day. I've learned that I knew even less about electrics than I thought, which wasn't much to begin with. Using something I hope is your method I get:

Brown wire - 12.6v

On turning the key to crank I get:

Red/white - 0v at the switch and 12v at the cable
Brown/white - 2.5v at the switch and 0v at the cable
Black - 0v at the switch and 12v at the cable

I've given all these values as I'm not sure I understand when you talk about the 'spade' whether you're talking about the bit on the end of the wire or the bit sticking out of the switch. Apologies for my ignorance.

I also tried to measure between the engine block and the body, but could get no reading at all on any point I measured when turning the key to crank.
A Jordan-Doherty

"I've learned that I knew even less about electrics than I thought"

Ah yes, Donald Rumsfeld's 'unknown unknowns' :o)

To me the spade is the flat terminal that is part of the relay and solenoid, for example, and what is on the end of the wire is the spade connector, but that's just my convention, others may vary.

To that end you will have to explain 'switch' and 'cable', in your voltage measurements.

With the key turned to crank you should have 12v on both the white/red wire connector and its spade on the relay. Likewise 12v on white/brown (and brown) wire connector and its spade, but 0v on the black earth wire connector and relay spade.

Note that I have mentioned both spades and the wire connectors this time, you can get bad connections between the two, which will give different voltages between the two, when they should always be the same.

If you get no voltage when measuring between an engine earth and a body earth, then that indicates the engine earth strap is not completely disconnected. But even when it is there are usually enough alternative paths via heater, choke and accelerator cables to allow the engine to crank, albeit slowly, and with said cables usually smoking!

You need two good earths to see any voltage that might be there of course, so no voltage could mean just that, or it could mean one or both of the earths is not good. the way to test any earth point you are proposing to use is to connect your meter between that and a known 12v point, i.e. the brown at the fusebox, and you should see 12v. If you don't, then either the earth you have chosen is not suitable, or possibly there is no 12v on your brown!
Paul Hunt

About time we started calling a spade a spade...

My switch is your spade.
My cable is your spade connector.

I think I'm going to have to let this rest for a few days until I can get a spare pair of hands to help. I'll hopefully pick this up again next week with something that resembles a reliable set of readings.
A Jordan-Doherty

OK, so are you taking the spade connectors off the spades to do the voltage tests?

That's no good, as I said earlier any circuit tested has to be carrying it's normal design current, which means all the connections have to be in place.

It would explain the white/red readings, but there are still oddities on the other two.

If you aren't taking them off, then all the connections to the relay are bad, as the voltage should be the same on the spade as on the spade connector.
Paul Hunt

Ok, let's get back to basics here.



Firstly ditch the multimeter, get yourself a good old fashioned 12volt test light.



First check, test lamp to a good earth, touch the brown wire on pin 30. Should be a nice bright light now = the battery supply to the relay is ok.


Second, clip the test lamp earth to the black wire at the relay and the other to the brown wire, light lights brightly=relay earth is ok.

Third, test lamp earth to a good earth, other lead to the white/red, switch to crank, if the light lights = the wiring is good from the switch to the relay.


Fourth, detach the brown/white from pin 87. Test lamp earth lead to relay pin 87, switch to crank, if the light lights = the relay is ok.


Fifth, attach the test lamp to earth, probe the brown/white spade terminal at the starter solenoid, switch to crank, light lights = wiring is ok from the relay to the starter.


If all that checks out ok, your fault must be the starter, the battery main lead or a poor engine earth lead.



SR Smith 1

"Should be a nice bright light now"

That's the point of a meter, how bright a bulb is is subjective, a meter gives you absolute values. A test-lamp is fine for a go/no go test, or if the connection is bad enough to make the bulb significantly dim. But a test lamp bulb takes a lot less current than a solenoid, and a resistance fault bad enough to affect a solenoid won't affect a typical test-lamp bulb much at all. And if you have an LED test lamp not at all.

"Fourth, detach the brown/white from pin 87. Test lamp earth lead to relay pin 87, switch to crank, if the light lights = the relay is ok."

Again removing a wire from a terminal and testing the terminal is not a valid test. If the relay has a high-resistance contact it will be limiting the current to the solenoid, but the resistance required to affect a solenoid is much less than that required to affect a bulb.

Other than that you are recommending the same test points as have been recommended before.
Paul Hunt

Hate to disagree with you Paul, but I do this sort of thing 5 days a week as a living, mostly on classics.


In my experience a meter can confuse because it can show battery voltage is present but not current, ie amps, which is the stuff that makes things work. Many is the time I've looked at things not working which someone has condemned with a meter showing volts present, you attach a proper test lamp and the light is either very dim or won't light at all, as you're "working" the circuit, which can highlight a poor connection, dud switch etc.


An average 12v relay will draw about .25 of an amp, a 21 watt test lamp draws about 1.2 amp a multimeter draws nothing.


And you say yourself............


"Again removing a wire from a terminal and testing the terminal is not a valid test. If the relay has a high-resistance contact it will be limiting the current to the solenoid, but the resistance required to affect a solenoid is much less than that required to affect a bulb."

Current, ie amps, being the important bit. Of course it's a valid test, it shows the relay switches current and it's simple to compare it with the current at term 30, ie the power into the relay.Plus you know as well as I do, by joining 30/87 bypasses the relay anyway.



And it's a simple test to chase down low current using a lamp, just do a back to back across the battery or fuse box against the circuit you're working on.



SR Smith 1

if like me you don't fully understand electrics then the simple test bulb can sort the simple yes/no/not-as-bright as it should be situations better than a test meter

I think if you don't fully understand it then a test meter can complicate things

what ever test equipment you use you need to test the test equipment before each test to prevent false results

with a bulb you could have burnt it out without realising it or with a meter have it on the wrong setting - or like my cheap test meter it goes intermittently faulty so can't be trusted

I made up a test bulb by soldering two wires to the bulb with spade terminals to the wires and made up a selection of wired fittings that can be used with the the test bulb or as by themselves as bypasses or joins including a long twin wire that can be connected to battery clips so that I can do direct to battery tests anywhere on the car (e.g. reverse lights to potentially save crawling under the car to the switch connections - Sod's Law I still had to)
Nigel Atkins

Exactly what I was trying to point out Nigel and I use these tests nearly everyday on classics.


Cars of our era are reasonably basic, you really can sort out most things with a 21 watt bulbholder with two long bits of wire attached.


Great for finding short circuits as well, saves blowing loads of fuses, just wire it across the fuse that's blowing. It'll be nice and bright if a short to ground is present, go round disconnecting things until the filament is just glowing, ie passing low current.If you make the wires long enough you can see the bulb even if you're working on the interior or at the rear.
SR Smith 1

"a meter can confuse because it can show battery voltage is present but not current, ie amps, which is the stuff that makes things work."

That's exactly why I say all connections need to be made and the circuit being used - or attempting to be used - the usual way. You use the components of the circuit to draw what current it can, and the meter will show you if there is a bad connection back towards the supply.

If the relay has a high resistance contact, or the brown wire feeding the relay has, then a voltmeter will show a drop in voltage - *if* the solenoid is connected to the relay contact. The solenoid on a pre-engaged starter - as in this case - will try to take about 60 amps when the relay is first operated. It needs a minimum of about 40 amps to operate reliably. As little as half an ohm will be enough to prevent the solenoid operating, by reducing the current to about 20 amps. In doing so it will reduce the voltage across the solenoid to about 4 volts, and this is what you would measure with a meter if the fault is as I say in the relay or the brown wire feeding it. My contrast a typical test lamp of 1 or 2 watts will only take about 200mA, which through a fault resistance of half an amp will only drop the voltage by a tenth of a volt. You would never see that reduction in brightness on a test-lamp if the load of the solenoid has been removed from the relay. You *would* see this reduction in brightness on a test-lamp with the solenoid wire connected. But with components that take less current, and small but significant bad connections, the effect of the bad connection is much easier to see on a meter than it is on a test-lamp.

I'm completely with you on using a 12v bulb *in place of* a blowing fuse, to find a short-circuit on a fused circuit, and have recommended that procedure a number of times.

By now I'm sure most people will have zoned out. And seeing as how you raised the point, I've been working on this stuff for nearly 50 years, a large chunk of that professionally.

Paul Hunt

You're right Paul, think we're "zoning out" ourselves here!



I always use a 21 watt bulb in my test lamp on older vehicles which draws approx 1.6 amps, a good load which "works" the circuit.



Hope this guy has sorted out his problem with all these tips!


SR Smith 1

My head is spinning a bit with all this advice. Something about there being too many doctors springs to mind... though I really do appreciate all the help.

I haven't resolved the problem yet, but I feel like I'm finally at the crease having realised I was messing up with my testing method (well spotted Paul). Now with the spade connector attached to the spade when I test I'm getting the following readings:

Brown - 12.2v
Black - 0v
White/brown - 12.2v
Red/white - 11.4v

I'll try the bulb test at the weekend and report back, but is there anything more I could be doing with my meter in the meantime? Cheers.
A Jordan-Doherty

OK, relay and brown feed are fine then - up to a point, as the solenoid may not be drawing any current for other reasons.

Next test is to see if you have that 12.2v on the solenoid spade that the white/brown is connected to. If so, then the problem is in the starter so it will have to come out. If not, then there is a break in the white/brown back towards the relay. Going way back to the beginning of time I see you have a 78, which means it should have the solenoid with two spades, with the white/brown on one and a white/light-green on the other. Originally these spades were different sizes so could not be mixed up, but rebuilt ones I have sen have them the same size. So if there is the slightest possibility that someone may have taken the wires off these spades, and put them back incorrectly, swap them over (if they are the same sizes) and try cranking.

You could remove the brown and the white/brown from the relay and tap them together a few times, and see if it sparks. If it does then there is a complete circuit in the solenoid, but it is complicated by there being a pull-in and a hold-in winding there. It needs both to operate, then the pull-in is bypassed, and it will stay operated with just the hold-in. But it won't operate if the pull-in is faulty, although the hold-in winding on its own will result in sparks.

If you do end up replacing the starter then bench-test it before fitting it. You will need to connect a battery to the solenoid stud, bridged to the operate spade for the reasons I mention above, and the body of the starter for it to work. If you only connect the battery between the operate stud and the body it will spark but not work. You can do this test with jump leads, but ordinary wire will also work as until the starter is turning the engine it doesn't take that much current.
Paul Hunt

After spending a bit of time with my new bulb tester and SR Smith 1's suggested five tests, I came up with a lit lamp to each of the first four. I then somehow contorted my oversized hands past the steering column down to the solenoid (which I had not touched up to that point) and cleaned up the one wire I could reach with a dab of WD40 until it showed up as the green/white wire. I couldn't reach the one next to it, but assuming it might be the brown/white I attached the probe to it and got a load of sparks, which I hadn't expected.

I then tried the key in the ignition again and........... the engine turned! A few revs of the starter motor and the engine roared into life. The only downside is that it woke my boy up half way through his morning nap and he's now a right little grump.

Any thoughts on what was happening here? Just a bad connection?

Couldn't have traced this down without all your help - so a big thanks from me.
A Jordan-Doherty

Maybe the probe was touching the battery cable stud as well and so was trying to operate the solenoid, that does take a lot of current and will generate sparks with a moving connection.

It does sound like it was a bad connection of the brown/white to its spade, but I would advise taking it off, cleaning the spade and perhaps pinching the spade very slightly if it isn't a firm fit, or it could well happen again when you least want it.
Paul Hunt

Great you've fixed it!


We must all pay homage to the mighty piece of equipment that is the mighty test lamp lol



Sounds like a loose connection/dirty terminal at the solenoid exciter wire.


Well done, go have a beer.
SR Smith 1

And buy your son's favour back with CBeebies!

Glad that this story had a happy ending - I have shamelessly lurked my way through it and have learned some things, so thank you for that.
Piers Colver

This thread was discussed between 08/01/2014 and 29/01/2014

MG MGB Technical index

This thread is from the archive. The Live MG MGB Technical BBS is active now.