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 - Ignition light a bit on when it shouldn't...

Hi all,

I have a 1972 UK B.

Very recently I've noticed the ignition light coming on very dimly when it shouldn't. The ignition light works as it should when starting the car etc, but then sometimes it comes on dimly and sometimes switches back off mid drive only for it to go back on dimly again.

Have checked fan belt & plug at the back of the alternator.

I've also got some voltage readings across the battery terminals (from one of those cigarette lighter type voltmeters, so accuracy may not be amazing - I fitted a 12V socket across the battery terminals for my tyre inflator)

There is not a massive difference between readings when it is running normally and when the light is dimly on. At about 1000-1500 rpm, I get 14.4V normally, and maybe 14.2-14.4V when the light is on. If I give it some more revs, voltage goes a bit higher up to 14.8V ish and light continues to be on at the same level of brightness.

Is it more likely to be an ignition side fault?
Nat

Most likely a diode in the alternator on the way out-
Almost everything else (wiring/brushes) would show as an output voltage drop
William Revit

The light acts like a pair of balance scales between the voltage being generated inside the alternator from the IND wire and the voltage from the ignition switch on the white wire. Those voltages are the same with the alternator charging correctly and good connections throughout, but if anything causes either of those voltage to rise or fall then the light will glow.

Diodes can cause this but usually have some effect on output voltage, and the light can glow brighter as more electrical loads are switched on, depending on which diodes and whether they go open-circuit or short-circuit. They can also cause changes in noise and heat but that's harder to spot.

If you have a test meter with probes measure the voltage on the white at the fusebox, and at the brown and brown/yellow at the alternator plug (still fitted but with the cap removed), all with respect to earth, and that will tell you more.

As a 72 you may well have two brown wires in the alt plug, one thick and one standard gauge, test both of them as well as the brown/yellow. If all three show the same as each other and as your previous readings then the alternator would appear to be working correctly.

Test the brown and white at the fusebox and if the brown is good but the white low it looks like bad connections in the ignition switch circuit.

If both brown and white are low it looks like bad connections on the starter solenoid, or in the brown to the ignition switch.
paulh4

I had similar on my midget, it was a recently fitted rebuilt unit.
Supplier replaced the unit and said diode was going dodgy.
richard b

Thanks everyone, some readings (with the ignition light dimly on):

White at Fusebox wrt to earth - 14.35-14.4V
Brown at Fusebox wrt earth- 14.4V

Thick brown wire from alternator wrt earth - 14.48-14.5V (I only have that one brown from the alternator, no additional standard gauge wire)

Brown-yellow from alternator wrt earth - 12.2V

The brightness of the light doesn't really seem to vary, even when I switch on headlights, indicators, fan etc.

Rats, is it the diode? The alternator seems to be some sort of aftermarket one which I don't recognise - I'm not sure where I'd buy replacement bits.

Second question - how urgently do I have to repair/replace it? Would you chance a 1hr 45min x2 there and back journey this weekend without getting stranded?

The B is my only car!



Nat

From the voltages on the wires at the alt it does look like diodes.

Your alt looks very similar to the attached which I think is a Lucas A127 and came to me on my roadster. They were rated at 45 to 65 amps or 70 amps with a stud output terminal which mine didn't have, but yours seems to. Replacing with a pukka 18ACR without the stud you will find the voltages a bit lower at idle with the lights on but still charging so nothing to worry about, they were all like that back in the day. Unless a high-output alt was installed because you run a lot of extra electrical equipment.

As the voltages on the brown and white at the fusebox are good you have nothing to worry about for the moment as far as the battery goes, the only ill-effect - at the moment! - is the glowing light. It could run for years like that, OTOH further faults could develop which might affect charging and not show on the light. You would see that from your cigar lighter voltmeter though.

If you replace the alt get one with pulley attached as they often aren't compatible unless you get the identical replacement. Size is also a factor as a bigger one will spin the alt slower and give less output at idle, they seem to be 2.5" and 2.75" but measuring them is tricky. Drawings show the pulley size is it's overall diameter, but with a belt on some sit flush with the OD and others can sit 5mm below the OD making it effectively smaller. If that wasn't enough there is also the question of right-hand and left-hand, the 4-cylinder MGB is right-hand with the two mounting ears at the top and the adjuster link at the bottom, but it's only a few minutes work to swap one over.

When the roadster one packed up I was struggling to be sure what I was getting online so went along to a local village motor spares shop and they had exactly the right one in stock. An amazing place, in the past they have had anything from a flasher unit for a Triumph TR4 to a specialised relay for a Lola kit-car. Slightly cheaper and saved the postage to boot. But one of the usual MGB suppliers should (!) have the correct one.



paulh4

Much appreciated!

I dislike being stranded, especially when further from home!

New alternator ordered from one of the usual suspects- I don't particularly have lots of extra electrical equipment - a radio and a USB outlet for a dashcam/phone charging at most.

Thanks for the tip on the version of the current one - I'll see if I can replace the diodes so that I have a spare.
Nat

Another tip from personal experience is that some replacement alts are fatter than others, and if you go from a slim to a fat you might find you can't get the belt over the pulleys, even if you leave the water pump until last which in theory is the easiest. Sometimes just removing the bolt that goes through the slotted link is enough, but I have also had to fit the belt with the two mounting ears very slack and bolt in the front one removed altogether as well.

Belt size is also a factor here, and depending on what you have now with your after-market alt you might find you need to go up or down a size. Originally 18V before 1975 used GFB176 (GCB10900 in modern numbering) which is 900mm long going from an original alt to a universal replacement you might need a 10925. However if your alt is an A127 that is classed as a 'fat' alt which puts the shaft a little but further out from the engine so you might need a smaller belt now.
paulh4

Thanks. I hadn't considered alternator fatness - but I ordered a standard belt anyway as part of my order with the new alternator as I wasn't sure how old the current belt is.
Nat

You'd very likely have made it. Mine recently died on a road trip, battery lasted 3 hours.
Paul Walbran

New alternator fitted! (Slightly less straight forward then expected - I needed to source another nut and bolt for the lug which attaches to the slotted link - the one I ordered in line with the catalogue was too big for the alternator lug at 3/8" - and the previous A127 style alternator used a M8 bolt in a threaded lug! & then I found that the new alternator was a bit long and wouldn't fit in the space in front of the oil filter- it turned out that I had to rotate the oil filter housing head at the engine block)

Interestingly, with a 'stock' alternator and pulley, the standard 900mm belt was a little bit short for my liking - the alternator had to be not connected to the upper swing ears in order to get the belt over the pulley. A 925mm belt would be better next time - there was plenty of swing via the slotted link to accommodate that kind of belt length!



Nat

Glad you got it sorted.

I don't like to mention it, but the hose onto the thermostat housing looks a bit crusty.
Dave O'Neill 2

Ah yes - forgot to mention the adjuster ear can be threaded or not threaded!

Didn't have the oil filter fouling problem though, in my case it had been fitted too far the other way and when the engine rocked the hose fitting was hitting the inner wing - like that when it came to me and for probably 20 years after that. I started getting problems getting replacement filters to seal and discovered the housing had bowed from the impacts - which hadn't been making a noise. It was getting harder and harder to get it to seal each year, the attached is just lightly screwed on showing how bowed it had got, it had also put a small dent in the inner wing. I had the short A127 at that time so could position a replacement head towards the alt and angled the hose fitting to clear the wing.

My replacement alt fitted without fouling the filter, make sure yours hasn't moved too close to the inner wing.

And while looking at pictures of both those issues I was reminded of yet another which is that the alt pulley may not align correctly with the water pump and crank pulley and you may need to use shims/washers to pack it out - the front ear is fixed but the rear should have a sliding bush. That was more of an issue on the V8 though, with the roadster only the adjuster link was out of line which matters less, I still shimmed it though.



paulh4

Dave - good point, I hadn't noticed that! Another thing to add to the list of things to sort.

Paul, good tip re the hose fitting and the inner wing- will check tomorrow. The alternator has a sliding bush at the rear lug. All the pulleys seem to be aligned - alternator is putting out 14.4V across the battery, and there are no funny noises coming from the pulleys or belts.

However, in fitting the alternator I either disturbed something or had a fault occur by chance -

After about 30mins into a drive, I lost power briefly - revs dipped (not to zero, ignition light did not come on). Continued with foot on accelerator, power picked up and then dropped a bit again. Power loss was smooth, rev counter wasn't bouncing. No bucking like a fuel pump issue.

Since then, I've tried driving around, but haven't been able to recreate it.

I haven't been able to replicate it by waggling the wires to the coil (only changed 3 months ago), wire going to my electric distributor, or the wires going to the fuse box or starter relay.

Tomorrow, I'll break out the multimeter and check voltages, and try driving for 30mins to see what happens...
Nat

Anything round the coil (with the exception of the condenser if you had one) would usually cause the tach to jump up and down.

Fuel loss from one carb would cause loss of power and a regular misfiring beat in the exhaust until it sorted itself out again. Had it been the pump it would likely affect one carb first then the other then be total loss of power depending on how long the problem lasted.

Could be loss of HT to one or more cylinders but then that usually causes a backfire when it comes back on again.

Electronic ignition can cause some odd problems the only test for which is usually substitution. But if it's enough to cause loss of HT it would be all cylinders so total loss of power, and again likely back-firing when it came back.

Tricky unless you can reproduce it, and even then still tricky unless you can bring it on with your head under the bonnet!
paulh4

Alas, I spent yesterday checking the coil (spade connections all tight, resistance across the terminals fine), got out some contact cleaner & emery paper and gave a good go at the connectors to the fusebox, coil, starter relay (for good measure).

Drove around for about 30mins, no issues.

Then this morning started up and it was going great until about 15mins in when I got a sudden smooth loss of power (like a cut off), revs dropped to zero, before the ignition light illuminated as the engine stopped spinning. Indicators worked fine as I coasted to the side of the road.

Couldn't restart (ignition light went on, engine turned but no luck).

Waited for breakdown truck. Breakdown man sprayed some carb cleaner, no luck. Then pulled the lead from the no.1 plug and it started on 3 cylinders. Put the lead back in & it ran fine. He then followed me home . Could it be the electric distributor? What else should I be looking at...
Nat

For elimination purposes ignition light NOT coming on until the engine stopped spinning shows you have power from the ignition switch and the alternator is charging.

When he pulled the lead off No.1 did he check for a spark to the block? Or did it fire up straight away on 3 cylinders so he didn't bother?

When you turned on the ignition switch for that start did you notice if the fuel pump was chattering away? If so then it looks like fuel pump issues or power supply to it.

Without that it's pointing towards ignition HT. Could be electronic ignition, coil HT, distributor cap or rotor issues. You say 'electric distributor', does it have a fully electronic distributor such as a 123? Or an electronic ignition module added to the original 25D4 distributor?

'Next time' if you have a timing light you can do some diagnosis while it still won't start. In this case the original neon type that fits between the plug lead and the plug would be best, in my experience the modern type with a 12v supply and a pickup that clips onto a plug lead may not work with the reduced battery voltage when cranking. OTOH if the engine has only just stopped so it's voltage is higher it may be enough.

Alternatively there are various HT testers around which you can watch while cranking. Some need the plug lead to be removed which could disturb things but this https://www.gunson.co.uk/product/77077/Spark-Indicator only has to be touched onto the outside of each lead in turn while cranking and it will flash is there is HT. At least with RHD you stand a chance of being able to lean into the cabin with the door open and operate the key while peering into the engine compartment with your other hand holding the gadget onto the leads!
paulh4

He had a screwdriver from the lead onto the block- I think he saw a spark.

Fuel pump wasn't making any weird noises for that start.

By electric distributor, its one of those ones by accuspark where they supply the whole thing including the main distributor body.

I have changed the rotor arm recently (decent ones from the distributor dr)

Thanks for that link- I also have a couple of those inline things that plug inbetween the top of the plug and the lead to see spark strength.
Nat

OK, I _think_ that is a 'conventional' distributor with a magnetic electronic trigger in place of points and condenser.

If it has a black and a red wire coming out of the distributor where does the red wire go?

What type of coil is it - any markings?

Can you measure the resistance between the two spade terminals, remove the wiring from one of them.

What tach do you have - does it say RVInnnn or RVCnnnn on the front? A 72 would originally have had an RVI, 73 changed to RVC. There can be issues with electronic ignition and the RVI tachometer.

All this is clutching at straws a bit given how relatively infrequent the problem is. Given that the tach only slowly drops as the engine spins down (as I understand it) the distributor does seem to be powering the coil as it should even after after the power loss.
paulh4

Hi Paul,

The red wire from the distributor goes to the same side as the white on the harness and to the -ve marked terminal on the coil.

The coil is a '12 volt High Performance' one - GCL110HP - Lucas

Resistance between the 2 terminals with wiring removed from one measured 3.3V

I have a RVC tach - I think the PO changed it when he put the electric ignition in.

I drove it again, and after about 25mins, it died again. I put one of those inline spark plug things on & it started up straight away.

Spark at No. 1 immediately after engine died
https://youtube.com/shorts/-zlk2VRAz3s

Then at home I did the following:

Spark at No. 1 & 2
https://youtube.com/shorts/c1mE93jK3Iw

Spark at No. 3 & 4
https://youtube.com/shorts/1gUP08BQOUs

Have I got the connectors the wrong way round onto the coil spades? (and have I toasted my distributor/coil?)
Nat

Nat--

"Then this morning started up and it was going great until about 15mins in when I got a sudden smooth loss of power (like a cut off), revs dropped to zero, before the ignition light illuminated as the engine stopped spinning. Indicators worked fine as I coasted to the side of the road."

Reading this I notice you say revs dropped to zero---I'm assuming you mean the tacho went to zero as you were motoring along.
Most likely this would mean the signal from the distributor has failed.

Then-
"The red wire from the distributor goes to the same side as the white on the harness and to the -ve marked terminal on the coil". ------WRONG

I'd say you've got your coil wiring around the wrong way and the dist. module is getting feedback/splatter from the coil--It 'might' have toasted the module or because it restarts after sitting you could be lucky and it's just overheating---- Depending on how long it's been incorrectly wired up there could a lot of rubbish/dust in the dist.cap that needs cleaning out as well---?

The wiring should be--White power wire from the wiring loom and the red from the dist.module onto the +POS terminal of the coil and the black from the dist,module onto the -NEG terminal of the coil

willy

William Revit

Thanks Willy,

It seemed like the Rev reading decreased in line with the engine slowing as it lost power rather than it still motoring whilst the gauge said zero if that makes sense.

I must have accidentally swapped the wiring positions when I changed the coil about 3 months ago due to a wobbly terminal. I'll have a closer look under the distributor cap for debris, but didn't notice anything untoward when I changed my rotor arm last week
Nat

The coil wiring does seem to be reversed but I don't think it is the cause of your problems. Reversed wiring does reduce HT output very slightly but unless the HT pulses are very weak for some other reason you wouldn't notice it. As coil current is switched on and off for each HT pulse the electronics are subjected to the same high voltages that you see as arcing at points and it is designed to be immune from it. But always best to get things the right way round.

With an electronic ignition module the instructions do usually say to put the red wire on the coil +ve along with the white. But on 72 and earlier modules it is the current pulses through the coil and white wire that drives the tach reading and with the red wire on the same terminal the additional signals from the electronics can confuse the tach and in these cases the red wire should be moved to the white terminal on the fusebox to separate the electronics and coil currents. On 73 and later models the white at the coil comes direct from the fusebox anyway. But I don't think that is causing your problems either.

Having recently changed the coil you do need to check you have the correct one, suppliers get very confused about which coil should be used when, and if you have the incorrect coil for your model then it may be overheating. For a chrome bumper MGB the coil should measure between 2.5 and 3.5 ohms between the spade terminals with the wiring removed from one of the terminals at least. If it measures 1.5 ohms or less it could be the wrong coil.

But this is where it gets even more complicated! Some electronic ignition modules are what is called 'variable dwell' which only passes current through the coil for just long enough to generate a good HT spark. Points and other modules are 'fixed dwell' and pass current for a much longer time which just generates wasted heat - the coil can get too hot to keep a hand on in summer. With variable dwell much shorter current pulses generating much less heat and the coil remains cool to the touch.

The only sure way you can tell if you have fixed or variable dwell is with a dwell meter (if not run the engine for an hour or so and feel the coil). If you have fixed dwell then the dwell reading will change very little as you rev the engine, staying at about 50 to 60 degrees depending on distributor type. By contrast variable dwell can be as low as 10 degrees at idle only getting up towards 60 degrees as the engine is revved hard, see here https://www.youtube.com/watch?v=1P0BYeDV4Ts Late North American spec with variable dwell electronic ignition from the factory had a coil resistance of 0.8 ohms - and if you get a shock from one of those you really know about it.

paulh4

Thanks,

hmmm.

Reading across the coil terminals yesterday (without wires connected on one side) came out at 3.3 ohms

Does reversing the terminals on the coil speed up breakdown of the coil?
Nat

Just noticed your post with the Youtubes. Because the video is digital and the HT pulses only very brief it can miss flashes which does seem to be the case in the first one less so with the other two. Idle does seem to be a bit lumpy though, which idle speed do you have? Only out of interest it's not likely to be a factor in this unless there is an irregular beat in the exhaust.

As far as retro-fitting an RVC tach goes it depends on how it was done. If done as per factory with an additional wire on the coil -ve to the tach then it still acts as a check on the ignition LT circuit, and if anything in that fails it will instantly drop to zero while the engine is still spinning or bounce around with a misfire. But if done differently it may not show a problem there. With electronic ignition now having a black wire replacing the original white/black points wire you could have three, two or one (the electronic ignition black wire) on the same coil terminal.

And now it's time to determine if you have a 72 model or a 73! 73 models started in August 72 so it's quite easy to have a 73 model first registered in 1972, as mine was. The change-over point in chassis numbers is 294251 roadster and 296001 GT so if your chassis number is later than that it's a 73 model which would have had an RVC from the factory.

3.3 ohm coil is correct so that's one thing that can be removed from the equation.

Reversing polarity shouldn't make any noticeable difference especially to breakdown, it's one of the decisions that has to be made when converting a car from positive earth to negative. Both reversing the coil at the same time and leaving it as it was has a very minor effect on HT strength - one effects the auto-transformer effect and the other the polarity of the voltage at the plugs but as I say neither should affect running unless there are other issues. Even than it's more likely to affect cold starts than run normally for extended periods then suddenly cut out, and especially as intermittent as it seems to be.

Correct the polarity and see how it goes. If the problem never happens again then all to the good :o)

But if it happens again and won't start then use your plug testers again ... and hope it doesn't fire up as soon as you have fitted them!
paulh4

Paul
My thinking was that with the coil hooked up backwards there might have been some feedback overheating the module as both windings in the coil are connected to the _NEG terminal and maybe some flaring inside the cap from the reverse polarity creating heat as well-
Hopefully it's that simple for Nat. and correcting the polarity fixes it.
Cheers
willy
William Revit

Thanks both.

Yeah, it was idling a bit lumpily, probably because it was cold and I hadn't pulled out enough choke- it normally idles somewhere around 700 rpm very comfortably once warmed up.

I think the RVC tach is wired correctly in my earlier (pre Aug 1972 chassis numbered car) -indeed, Paul, you were able to detect a slight tach dropping intermittently from my video indicating a LT fault on my previous thread here

(https://mg-cars.org.uk/cgi-bin/or17?runprog=mgbbs&mode=archiveth&archiveyear=71_2025.dat&access=&subject=71&subjectar=71&source=T&thread=202509281058562384939)

with the wobbly coil spade terminal (and the resistance varying about across the coil terminals).

See attached picture for how my tach appears to be wired up (along with the distributor and coil and now the white from the harness and red from the distributor now correctly connected to the +ve terminal of the coil and the White-Black and PO earths now on the -ve) - not shown in the diagram is the white to the ignition light

I haven't had a chance to run the car today with coil terminals connections swapped, but did have a look under the distributor cap - see attached photo - not sure if it is overly dusty - maybe a bit.

I asked accuspark if their distributors are variable dwell and they are, and powerspark (which also looks very similar) say that they are also variable on their website.






Nat

Wiring looks correct, I see you have a 'kill switch' fitted which makes a nifty immobiliser on this type of tach.

All you can do now is see how it goes with the coil connections correct. I'm sure you will then leave it alone, but me being me I'd reverse the coil connections again and see if it reoccurred :o) If so then running a reversed coil with this type of ignition is definitely to be avoided.
paulh4

Okay, had a chance to do an extended test run this evening with the coil connected up correctly.

Several circuits of very local roads - so no chance to go above 30mph, but managed a run for about 1hr and 20mins continuously without any issues. However, shortly after that the problem seemed to reappear, though not as bad - it was more a momentary power loss, and the tach did not have a chance to drop fully to zero before the engine recovered and the power seemed to return to normal. Stopped and then left it idling for a bit with some increases in revs to see if I'd get any power loss, but couldn't replicate the fault.

Coil felt warm, but not much more than the oil cooler hose. Left it for about an hour and a half, and then tried driving it again - about 5-6mins on, the power dropped again (not to zero) before recovering momentarily and then dropping again. This happened a few times - but I kept my foot on the accelerator at the same level and didn't ease off. Coil didn't feel particularly warm.

Nat

OK, still there then. As at the outset I can't see it being the ignition LT side i.e. voltage supply, electronic ignition or coil primary or the tach would drop to zero while the engine was still spinning down.

I have my doubts about HT as well as that usually causes backfires from unburnt fuel in the exhaust.

That leaves fuel but I can't see how that would stop and restart as seems to be the case. However there is one possibility and that is a significant vacuum leak that comes and goes, but it would have to be a big one. Servo if fitted? Servo hose? That could be blocked off at the manifold port for elimination purposes as the servo gives negligible assistance, it was optional for a number of years when the rest of the barking system was standard.

If you still have the old coil I'd be inclined to put that back for the moment. I realise that was giving you momentary misfires before but that is better then cutting-out altogether. If it's something other than the coil then you will probably experience both problems!

Until it happens for long enough to stop altogether and not restart so you could use your spark testers I can only think of one other option at the moment. That is to use one of the old-style neon timing lights positioned in the cabin with long leads connected to a plug lead or ideally the coil king lead, and that would flash all the time HT was present. The same thing with a powered light would be possible but mine has a curly leads which don't pull out very far and you would need to keep the trigger operated all the time.
paulh4

Thanks Paul,

I actually had a spare new coil, so had fitted that with the connections the right way round, so I don't think its the coil.

I don't have a servo fitted, but do have vacuum advance on the distributor. Would it be worth disconnecting that and driving around to see if maybe theres some intermittent dodgy connection at the breaker plate?

Is is this kind of neon timing light you mean?

https://share.google/AHMnsX7DIU1roXhtt

Also, this issue only started after I changed out the alternator. I note that the red wire from the electronic distributor is quite close to the back of both the old (A127) and new (ACR 16/17) alternators. Would it be possible that the new alternator is affecting the red wire with some sort of magnetic field issue?






Nat

OK re the coil.

For distributor vacuum I doubt it would allow enough air in to prevent combustion, try just pulling the tube off the port on the rear carb with the engine running to see what I mean. There will be a difference, but it shouldn't be much. You could also remove that tube and block off the port altogether just to see what happens. It will affect part throttle acceleration though - if vacuum advance is working correctly.

The test for that is with the tube removed from the carb suck on the tube, you should not be able to pull any air through. If you can suck hard enough with the distributor car removed you should be able to see the 'points' plate that carries the electronic module twisting back and for, and with the car refitted and the engine running it should speed up the idle.

The usual check for vacuum leaks is to spray propane/butane from an (unlit!) blow torch (or carb cleaner) round all the joints between carbs, manifold and head with the engine idling. If the engine note changes at all it indicates a vacuum leak.

It's very difficult to visualise how the alternator could be causing this problem, unless it's earthing the coil HT lead somehow. Anything that stopped the electronic ignition module sending regular current pulses through the coil each time a plug is supposed to fire would show up instantly on the tach.

Yes that does look like the neon timing light I mentioned.

Another idea that comes to mind is an exhaust gas monitor. If it goes rich when the problem occurs then the plugs aren't igniting the mixture, but if it goes weak then there is no mixture to ignite. However with something like the Gunson's Gastester it takes a finite time to respond, and unless you are driving on a flat road at a steady speed with a fixed throttle it's likely to be varying anyway. Easy enough for a passenger to watch that in a GT, less so in a roadster unless they are willing to be shut in the boot :o)
paulh4

Thanks Paul. Not sure I could persuade someone to hang out in the boot of a roadster!

I disconnected the vacuum advance hose from the carbs, and gave it a hard suck which rotated the plate.

For fun, I measured the resistance of my HT leads (out of the car and disconnected from the distributor cap)- they all varied a bit, but the king lead came out at 8.3k ohms and the others somewhere between 5-10k ohms

An interesting development though - in case that red distributor wire was too close to the alternator and affecting it somehow, I clipped the cable tie from it and moved the wire. Removing the cable tie gave more degrees of freedom for me to manipulate that wire more. I had the engine idling and then manipulated the wire more including giving it relatively firm tugs.

It wasn't 100% conclusive, but I did a manage to cause the engine to die, and also stumble a couple of times, but not every time I tugged or manipulated the wire...

The red wire runs straight into the hall sensor in the distributor so I wasn't able to get a resistance reading across that length of wire.

The lack of replicating the issue consistently is a bit of a head scratcher though.
Nat

Maybe because that red wire was a bit tight it has broken inside from movement-or has a bad connection in the end terminal that attaches to the coil. Can you try and pull on the wire in sections while it's idling and see if the breakage can be tracked to a particular area. maybe stick a new terminal on it-
I've been sort of stepping back from this to be fair to Paul who always has good advice and doesn't need me being a pain on the side. and-
Having said that i agree it has got the symptoms of a fuel blockage somewhere---If it runs around town ok but fails with a bit more speed--Do you have a fuel filter on it---how long since it's been changed. Maybe pull the fuel line off one of the carburettors into a bottle/container and check that you've got a good constant flow that doesn't drop off after letting it pump for a minute or so--A blockage usually starts off strong for a few seconds then drops right off-
I doubt it's the ign. module--I've fitted quite a few Accuspark modules and never had a failure, except one and that was a wiring issue where there wasn't any/much slack in the wires in the dist and the movement of the adv. plate had fractured the wire inside it's insulation.

Exhaust gas monitor------
Yes they show rich and lean--but
If the engine has a missfire it won't show rich from unburnt fuel
The monitor measures the value of the burnt gasses coming from the exhaust--If there is an electrical misfire and unburnt gas in the exhaust the meter will read -lean- as there is no burnt gas for it to get a reading from---I know it sounds a bit back to front to what we think but that's the result from electrical missfires.
Looking forward to see what you find--away over the weekend so will check back Tuesday.
William Revit

No pain in the side Willy.

Those leads are fine, unless the coil lead goes very much higher when pulled about or twisted. Are they silicone cored or the original carbon string? The latter do go high resistance over time as the carbon dust seems to migrate, but silicone cored should be stable for decades and many tens of thousands of miles. Basic Hot Wires are fine at around £15, you could spend ten times that and gain nothing on our cars.

The wires to the trigger module can fracture inside the insulation as described with the continual twisting back and fore of the points plate as you move the throttle pedal, but it's rare. The original distributor wires are very flexible for this reason, including the points/condenser wire on a 45D4. However there is still the theory that if the module stops switching the coil on and off the tach will drop instantly. You could confirm that by holding the engine at 2 or 3 k and removing the red and black wires one at a time and seeing what happens to the tach.

However! Another thought is because the module is not a physical contact that opens and closes like points the electronics could still be switching but not going fully open or fully closed when they are supposed to. A smaller voltage or current variation at the coil could be enough to trigger the tach but not enough to produce sufficient HT to fire the plugs.

If you can continue to reproduce the problem by waggling the wires then do that with a voltmeter connected to the -ve terminal (black wire) and when idling you should see a steady voltage which is actually a representation the dwell angle. I would expect it to be between about 7v and 14v depending on whether it is a variable dwell module or not. If it is then the voltage will be high at idle and go down as you rev the engine like in that YouTube link of a dwell reading I posted on 17th, but not below about 6v. If it's fixed dwell then it will vary much less and be about 7v all the time.

If the engine cuts out when pulling on wires with the engine held at a steady speed then watch the meter carefully to see if the voltage goes either up or down. If it does either then the module or its wires are at fault. The reasoning behind that is if the module is not fully 'closing' and drawing less current through the coil that it should then the voltage will go up. If it's not fully 'opening' and still passing current though the coil when there should be no current then it will go down.

If the voltage doesn't change then it's back to HT or fuel. The SU pump should deliver a minimum of 2 pints per minute in a continuous series of pulses with minimal bubbling. If it can deliver half that in half the time that should be a good enough test.
paulh4

BTW, I've done this kind of testing with analogue meters which average fluctuating voltages (and currents) well. I have found digital instruments can give different results depending on their internal design, which may do the same averaging, or may not. In some cases the reading will jump all over the place and never settle.

Measuring coil current can be another way of looking at this problem, if the current goes up when the engine stops firing the 'switch' is failing to open, and if it goes down it is failing to close.

They both assume that the 12v supply on the white wire is a steady 14v or so while it is spinning.
paulh4

Nat, I've followed this thread with interest as I had a similar problem last year. The car would run absolutely fine and then suddenly die. Then if you left it a few minutes it would start and run fine and then die again. I checked everything, reterminated all the LT wires (as they looked like the could be suspect after 50 years) and even changed the fuel pump for good measure.

The car is a 76 roadster with a 45D distributor with an Accuspark electronic module. Like you I checked all the wiring in the distributor to ensure there was no chafing of the wires on the baseplate, making sure that the cable tie was keeping the wires away from the rotor and ensuring that the baseplate was able to move freely with the vacuum advance.

There is not a lot of room for the wires to go through the body of the distributor and into the ignition module and it was there that the problem was. There was an intermittent break in the red wire, which was causing the issue. Given the construction of the module, I was not able to rewire it and so ended up just changing it.

It's best to take the distributor out to change the module, which I did, ensuring that the wires going into it were not too tight and had sufficient movement for vacuum the advance mechanism. I retimed the engine and It's been fine ever since.

Andy

Andy Robinson

Andy - did your tach suddenly drop to zero when you lost power or keep registering as the engine was spinning down like Nat's?
paulh4

It's a bit difficult to remember Paul if I'm honest, as when this problem started the car was in the garage, just warming the engine up in readiness for a run out and oil change. As is the case with all intermittent faults, it was difficult to locate and I did initially think that the problem was fuel related and it was just running out of petrol, as mostly I was getting a good spark, but that was of course after the engine had been left for a while.

When the fault was actually there and I sat in the car to start the engine, I do recall seeing the tach bounce like Nat's.

I did wonder if the problem was heat related as these Accuspark ignition modules rely on white silicone heat sink on the base to help dissipate heat. In the end it was intermittent break in the red wire that was the root cause of the problem.

Andy
Andy Robinson

From my understanding Nat's only drops slowly as the engine spins down "It seemed like the Rev reading decreased in line with the engine slowing as it lost power"

Tach suddenly dropping to zero while the engine is still spinning down, or bouncing with a misfire, does indicate an ignition LT problem including the ignition module wires.

Incidentally those 'under the cap' modules still need a good earth connection between the points plate and the body, done originally with a cloth covered tinsel conductor wire which is very flexible. It's that earth that is connected and disconnected to the coil -ve to generate the HT.
paulh4

Thanks for everyone's contributions, it's really appreciated!

With the red wire now 'free', I've tried manipulating/twisting/pulling it section by section, but I would say I've not been consistently able to replicate the fault. Quite often despite pulling and pushing, there seems to be no effect!

I don't have a fuel filter, this is the flow delivered by the pump:

https://youtube.com/shorts/rJMVFG0hAyY

My container (500ml) filled up quite quickly so I had to shut it off to prevent spilling too much!

King lead is one of those silicone cored ones - resistance didn't vary despite be twisting and pulling on it in various directions. I do have a spare which I could swap out as another option.

I couldn't quite do the revving to 2-3k rpm then pulling the black and red wires off at the distributor without an assistant, but here is what happened at idle (sorry about the slightly dodgy camera work, it was with a web cam that didn't quite have the right focal length!)

Pulling the red wire off (there was a secondary energisation when I touched the connector back on to the terminal at the coil):

https://youtu.be/Ke3hgHOcmbk

Pulling the black wire off:

https://youtu.be/d02abLiZ2P0

This is the voltage from the black wire at idle and some revs to get an idea of dwell:

https://youtube.com/shorts/0D1KpkwCsBg

This one shows the voltage from the black with the red wire wobbled - I'm not sure its very conclusive- sorry about the camera work
https://youtube.com/shorts/0II55GqWkmU

Here's another with the red wire wobbled, but with the tach in view. In this clip, it looks like the tach drops a bit faster...
https://youtu.be/Se11GH8jjqw

I tried doing the same but measuring the current, but couldn't recreate the fault after much fiddling.

I'm going to the Practical Classics & Restoration Show at the NEC tomorrow; if Accuspark have a stand there, I could see about buying another unit - the worst I guess is that I end up with a spare for the boot if that isn't the issue...

Nat

Oops, a couple of the videos didn't upload correctly:

2nd video (red wire pulled off)
https://youtu.be/B5GhWU3PcbU

last video (red wire wobbled)
https://youtu.be/vKXW66goVM0
Nat

All good stuff Nat.

Nothing wrong with that fuel delivery I can see.

The 'wires pulled off' videos do seem to show the sudden drop to zero that I would expect, as you say red wire dropping faster.

However 'voltage at black wire at idle ...' has higher voltages that I'd expect. A idle I thought the 13.nn might be indicating it was a variable dwell module, but to go up to 14.5v when revved is more like I would expect to see on the coil +ve i.e. red wire i.e. is the alternator output voltage.

Can you double-check that one, and do the other coil terminal/module wire as well?
paulh4

Doing the same test again on the coil/module:

Voltage at the black:
https://youtube.com/shorts/uW-YZz7akj8

Voltage at the red:
https://youtube.com/shorts/LJ7QKFfZNqI?feature=share

Nat

Thanks Nat, I appreciate it is a faff doing all these tests when neither of us know what the result will be :o)

A DMM periodically samples an instantaneous value on the leads then processes it before displaying instead of displaying the average that an analogue meter does. When the electronic module is 'open' the coil isn't drawing any current so nominally both terminals will show the same voltage i.e. battery voltage and when it is 'closed' the -ve will show 0v.

The interesting thing is that whilst the red voltage is pretty stable the black starts off lower than the red then goes higher as the revs are increased i.e. higher than battery voltage. There are all sorts of factors in play here including the back emf that you see as a spark at points (suppressed in the electronic module in a similar way to the condenser with points) and the design of the electronics in the meter. You can get an instantaneous 200v on the coil -ve when points open hence the spark.

The upshot is that short of an oscilloscope it is going to be difficult to see what is changing if you only get a very short break in ignition when waggling the wires as the meter will take a moment to display the new condition.

I think it's back to the neon timing light to hopefully tell whether HT is present or not when the problem happens when driving.
paulh4

Thanks Paul, doing all this testing is less faffy than being on the roadside waiting for a breakdown truck!

Did you expect to see the voltage to change when I was wobbling the red wire causing the stumbles & when I actually managed to stop the engine? I'm not sure if the tach behaviour there was actual sudden dropping or just spooling down!

I can have another go of trying to recreate fault but with the MM measuring current.

Accuspark were at the NEC yesterday & I picked up a whole replacement electronic distributor at a discount. The guy confirmed it had variable dwell.

I found a neon timing light on ebay, and that will be arriving this week...

Shall I change out the king lead? I'm not sure it will do anything but I have one to hand.
Nat

I must confess I didn't look closely enough at the 'red wire wobbled' the first time. Viewing it again the voltage seems to go up from below 14v to above 14v with each drop of the tach. I can't tell much from the tach itself as at idle it will drop to zero pretty quickly anyway, but those voltage changes combined with tach flicks could well be indicating that the module is intermittently failing to connect an earth to the negative to pass current through the coil in preparation for the next HT pulse.

Giving they both happen when the red wire is being waggled I do feel that confirms the module and it's wires are suspect. As you now have another identical distributor I would put that on and see what that does. I wouldn't do any wire waggling with the new one, just drive it as far as you dare and see what happens. Unfortunately there is a very faint possibility that the wire waggling cut-out is a different fault to the driving cut-out and non-starting problem, stranger things have happened! I'd have the neon timing light connected as well.

I wouldn't swap the king lead yet, don't want to make too many changes.
paulh4

Thanks, next step is to familiarise myself with how to correctly swap out a distributor and do the timing etc (handily with the neon timing light that will arrive this week) !
Nat

As you are changing distributors it's probably easier to leave the clamp plate attached to the engine.

But before you do that note what the timing is set to now. With a neon light you will need to do it in low ambient light levels as they aren't very bright, it helps to put a little white paint into the notch in the pulley (as in the attached). TDC is the top pointer on the cover and they are 5 degrees apart BTDC.

Timing for your engine is 10 degrees static (but I don't know if electronic ignition can be static timed, hence getting it approximately right to start with), strobe is 13 degrees at 600rpm. However 600rpm is a bit low so 19 degrees (the bottom pointer is 20 degrees) at 900rpm might be easier. Remove the vacuum pipe from the rear carb and plug the port.

Connect the light in series with No.1 plug lead, start the engine, and note where the pulley notch is in relation to the pointers and the rpm on the tach.

Switch off, slacken the long clamp bolt that goes across the clamp plate, and at least one of the screws attaching the plate to the block.

Note where the vacuum capsule is pointing, wiggle and twist the 'old' (a relative term ...) distributor out and put the new one in with the vacuum capsule pointing in the same direction as best you can judge.

The distributor probably won't have gone all the way in, so slowly rotate the rotor arm until you can feel it drop further in. Make sure the flange on the distributor body is hard up against the clamp plate, it will only do this in one position of the rotor. However you may feel it's starting to engage with the rotor 180 degrees away from the correct position but it won't fully seat, keep turning until it does fully seat.

That should be close enough to get the engine started and running to check the timing, lightly nip up the clamp bolt.

Probably easier to fit the old cap onto the new distributor at this point rather than swapping leads over ... and hope it isn't the cap that is the problem! But if you do decide to use the new cap swap the leads over one at a time.

With the timing light connected as before go for a start. Hopefully the light will show the pulley notch in about the right place, with the clamp bolt loose enough you should be able to twist the distributor body clockwise to advance the timing and anti-clockwise to retard. You should find that as you advance the engine speed will go up and as you retard it will go down, so you will need to check what the tach is showing as well as where the crank pully notch is in relation to the pointers.

Good luck :o)



paulh4

Sorry, have been away- update so far!

Prior to fitting the new distributor, I tried running with the neon timing light - the wires were only just long enough, but I managed to get the light in the cabin under the dash. I had a bit of a disco situation with the flashing red light, and did a few short local trips- but couldn't get the fault to reappear except once:

Driving along at a steady speed (20mph ish), revs at 2000 rpm, before rpms suddenly fell away (ignition light did not illuminate & car recovered quickly). It was daytime, but I think the timing light continued to flash.

Today, I've swapped out the old distributor with a new one - thanks for the instructions Paul, they were very helpful.

Interestingly, the original setting at idle of 600 rpm (without vacuum advance plugged) was about 3 deg BTDC which is a lot lower value than the workshop recommendations.

I tried fitting the new distributor with the vacuum capsule in the same position, but the timing was way off. I rotated the distributor more, and have ended 6 ish deg BTDC. Revs at idle are a little higher at about 700rpm, but on the test drive, varied sometimes it was 750 at one set of lights, 1000 rpm at another...


(top picture of each photo is before, and bottom is after new distributor installation)

I used the existing cap and leads & have reinstalled the timing light under my dash.

A test drive didn't reveal the fault, so I think it might be a case of driving it around, which is not so reassuring - so might stick to local journeys in the meantime!

I now have the old potentially faulty unit - are there any bench tests I can do to see if this really was the cause?






Nat

Good feedback, thanks. Probably needs a longer cut-out to be sure what the timing light is doing.

I know that when people first change from points to electronic that they can find the timing has moved quite a bit and that is because the electronic trigger 'firing' point isn't in the same place on the mounting plate as where the points opened. However I'm surprised that has happened using ostensibly a duplicate distributor from the same supplier.

The firing order is anti-clockwise, and your before and after vac capsule position shows the after having been _retarded_ compared to the before. Have you got the leads one position round compared to before? I can't tell where the leads go but the left one off the cap definitely goes in a different direction but that could just be because the whole body including cap is quite a few degrees different to before. Alternatively the timing may have been very advanced with the new distributor compared to the old.

Unless you rig up a bench test rig something like the attached there isn't a lot you can do with the 'old' distributor ... except perhaps put it in another car :o) I made that rig when I was experimenting with springs, the variable-speed drill drove the distributor through a right angle drive and I made a degree wheel to fit on top so I could read off the timing with a light. In that picture I was using it to set dwell on the bench which is easier than on-car.

As far as 'what now' goes I don't think there is much more you can do except drive it around with the neon light fitted - assuming you can get the petrol!



paulh4

BTW wires only just reaching noted but if you can get the light in the coil lead instead of a plug lead it will flash four times faster and be easier to spot if any are missed.
paulh4

Yeah, not sure why the timing was so different. Perhaps the vernier adjusters were different settings on each. Setting the fuel mixture & the idle is on my list of things to do as I reckon it is running quite rich.

The leads are in the same position before and after (I just moved one a bit for a clearer photo for the vacuum capsule position).

I'll see if I can get the timing light terminals into coil lead and the corresponding port on distributor cap without the end falling out (the light leads are kind of push on the end of a spark plug/push into a lead boot)

Will drive around and see what happens - no petrol shortages noticed yet, though I use the higher octane pricier stuff which may help.

Depending on how it goes & if the electronic module was the issue, I might buy a spare replacement module for the 'old' distributor and chuck it in the boot.


Thanks for your help!
Nat

Good point about the vernier! If I've had the distributor out I pre-set it in the middle to give equal adjustment either side for 'on the road' tweaks.

No shortages here either, just long queues at busy times with people topping-up. I went at 9am on a Sunday morning, all grades available on all pumps and one other car for 12 pumps.
paulh4

Okay, an update, in a month I've done 400 miles with a variety of distances and roads types with the new distributor. So far (fingers crossed), the fault has not reappeared.

So it appears it was the electronic distributor at fault all along. Have got a replacement module to put on the old distributor to keep in the boot as a spare. Symptoms were a bit confusing, but got there in the end.

Thanks for all your help!
Nat

Good to hear. That's the problem with electronic ignition - very difficult to diagnose and often substitution is the only way. So despite their apparent reputation for 'reliability' one still needs to carry a spare!
paulh4

Incidentally! I would recommend that you refit the previous distributor with the replaced module to confirm that it is a 'worker', and keep the 400 mile verified unit as the spare!
paulh4

Gotcha! A way of avoiding a(what would be a very irritating) dormant failure!
Nat

This thread was discussed between 03/03/2026 and 19/05/2026

MG MGB Technical index

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