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 - Strange horn behaviour

Hi all,

I've a 1972 home market B roadster, with an electrical layout which seems to best match Diagram 11 on Advance Autowire's wiring diagrams.

My horns are doing something a bit odd - they have previously worked as they should, but recently I've noticed that the horns don't sound properly when the engine isn't running. Pressing the horn push results in a very a quiet sound like a momentary energisation but that's it.

When the engine is running, the horns sound as they should when the horn push is pressed.

I've checked everything else that runs on that purple circuit (interior lights, boot light & output of the cigar lighter wire (now connected to a USB charger)) and all is functioning as they should.

The horns aren't that old (new Lucas ones fitted in May last year).

Possibly related is that when the ignition is off - usually within a minute or so of switching it off, I sometimes get a very quiet & brief sound like a momentary energisation of the coils in the horns.

At first I thought it was me just accidently touching the horn push, but I found it sometimes happened after I had already got out of the car.

Any ideas on what is going on?

Nat

Horns are voltage sensitive and when the engine has not been running for a few minutes the system voltage will be lower then when the engine is running and the alternator charging. But the correct horns i.e. two of them should be plenty loud enough with the engine stopped.

It's more likely there is a bad connection in the circuit, could be either the 12v side or the earth side.

Assuming the interior light works what does that do when you sound the horn with the engine stopped? It will dip down a bit but not much, if it dips right down then there is a bad connection on the 12v side quite possibly somewhere in the fuse-box where there are several connections, if you have a voltmeter you can check voltages on the brown to purple fuse.

However on both mine it was the earth path that was the problem, my V8 came to me with a moped horn because that was all that would make any noise. The earth path is from the body via the front cross-member, steering rack, column UJ, steering column and steering wheel to the horn push and resistance in that path can build up, I had to fit relays and after that they were fine, more info on that here https://www.mgb-stuff.org.uk/electrics1text.htm#hornrelay

If you have a voltmeter you can do some tests at the off-side horn. There should be two purple wires on one horn spade and two purple/black on the other. With the wiring on the spade terminals and your meter connected between the purple connection and a good earth you should see 12v. Have someone sound the horn and it should still be close to 12v, if it drops down noticeably there is a bad connection back through the fuse-box.

With the meter on the purple/black connection you should again see 12v, but when someone sounds the horn that should drop right down to zero. If you still see some voltage there then there is a bad connection through the earth path via the steering wheel, a relay will get round most of those but it could be the horn push itself or the siring down to the horn, again a voltmeter will show which.
paulh4

The first things to check on anything electrical forward of the radiator, are the bullet connectors.
Pull them, clean them and remake. Easy way to check the earth return is to remove the purple/black wire from the horns and jump from the spade to a good earth.............wear ear defenders though.!!
Allan Reeling

No bullets for the horns in front of the radiator, the wires come along the inner wing with the lighting wires (which _do_ have bullets) terminating in spade connectors on the horns.
paulh4

I was relying on a distant memory Paul, as both my V8s have lost 75% of their bullets.
Allan Reeling

Thanks both.

First test - trying to sound the horn with engine off and interior light on. Light dips slightly, so 12V on that side seems ok.

Will break out the voltmeter next & will report back...
Nat

You say you've checked everything on the purple circuit, but you didn't mention the headlamp flash.

Is that working as it should, with/without engine running?
Dave O'Neill 2

Headlamp flash works as it should, either with engine off or running.
Nat

Are the horns even quieter when you have the headlamp flash stalk pulled?
paulh4

No noticeable change in volume level-i.e. still very quiet, brief noise (this is with engine off). I haven't tried engine on, high beam on, horn button pressed.

I'll get my volt meter out over the weekend.

I still get the weird momentary energisation horn noise with engine off, horn button not pressed and battery isolator off, shortly after switching off the engine...
Nat

I was focussed on the weak operation rather than the comment near the bottom of your original post about the brief sound after you had switched off the engine, which is something else altogether.

If it's only a click could it be something else? Have any relays been added to the car? There should only be the starter relay on a 72. I know certain eras of North American spec can make a click shortly after switch off but not UK spec.

What you could do is remove at least one of the connections from each horn and see if you get the same brief sound then after switching off then. If when reconnecting and disconnecting several times you only get the sound when the connections are made it would seem to be the horns ... but as to why would be another matter.

If it still happens when they are disconnected then it must be something else.
paulh4

Have you tried looking at the horn push Nat and how the carbon pencil makes contact with the slip ring?

I had a problem with mine quite a few years ago now and Paul Hunt helped me out. See attached from Paul's website.

https://www.mgb-stuff.org.uk/hp.htm

Andy
Andy Robinson

It's got the symptoms of a relay wired in wrong
If it has a relay--probably tricked by the live feed and earth switching and has the relay fitted and incorrectly wired into the earth(button) circuit
If it has a relay, I'd be tempted to ditch it and go back to original wiring.

willy
William Revit

Hi all,

Andy, that was one of the first things I checked out - pencil is making good contact. To ensure a continued good fit, I had previously wrapped a bit of insulating tape around the plastic sleeve to make a snugger fit with the hole through to the slip ring so it's less likely to jump out of position.

Paul,

Measured some voltages:
From OS horn -
purple-black to earth = 12.45V, when horn push pressed = 4.2V

purple to earth = 12.4V, when horn push pressed = 10.4V

purple-black connector at the steering column to earth - 12.4V then 11 ishV when horn push pressed

Willy,
No extra relays added to the car, its all rather standard up there!

In the meantime, I've tried unplugging the OS horn to try and chase down the source of the brief sound. Will see how that pans out. (In the meantime, with only one horn connected, the remaining connected one sounds nice and loud when the engine is off and the horn push is pressed).

cheers
Nat

"wrapped a bit of insulating tape around the plastic sleeve to make a snugger fit with the hole"

The pencil needs to be free in the hole in the wheel boss. It is supposed to have the insulated 'fixed' end (as in the attached) facing the horn push which means the whole pencil moves up and down as the horn push is operated. It needs to be this way round because the sprung contact being longer can touch the body of the horn push and sound when it's not supposed to. If there is any resistance to movement in the wheel boss hole it can stick and not sound or make a poor contact. There should be a spring inside the pencil that keeps both ends pressed against the back of the slip-ring and the horn push contact when the horn push is in position.

"the remaining connected one sounds nice and loud when the engine is off and the horn push is pressed"

That does show bad connections - enough current available to power one horn but not both.

"purple to earth = 12.4V, when horn push pressed = 10.4V"

Some resistance in the 12v supply, do the same test on the purple spades at the fusebox. If you see much the same voltage check the brown side, and if that shows 12v then the fuse-holder connections are bad, as well as the ones you can see there is one each side of the fuse on the back.

"purple-black to earth = 12.45V, when horn push pressed = 4.2V"

Bad earth. Measure where the steering wheel connects to the column. Should be zero released or operated, if it shows 4.2v when the horn push is operated the column earth is bad.

If it still shows zero then the horn push contact or pencil contacts to horn or back of the slip-ring are bad.

Somewhere between 4.2v and 0v show both sides have some problems!

"purple-black connector at the steering column to earth - 12.4V then 11 ishV when horn push pressed"

This doesn't compute - it should show 12v with the horn push released and ideally 0v when pushed from the earth via the horn contact. Your comment higher up says at the horns it shows 4.2v when the horn push is operated, if it was 11v at the wheel then even one horn wouldn't sound and you couldn't possibly have a lower voltage at the horns as the 12v comes through the horns themselves. This does assume the horn circuit is wired as per the factory manuals. Because you have bad connections voltages could well be varying from one horn push operation to the next of course, which is why it needs multiple operations to see if there _are_ variations.



paulh4

I may have measured something wrong - will try again!

Re pencil - I forgot to say, a previous PO has replaced the steering wheel and boss to one of those motolita style ones, so the contact is within the boss itself. I have tested that separately, and there is adequate electrical continuity through the boss itself when it is pressed.




Nat

As that stands there you should be getting 12v(roughly) at that green wire where it's soldered on there.and 0v when the button is pressed.
If it's not going to 0v then there is an issue with either the column earth, the horn button or the earthing of the button housing to the steering wheel hub.
To prove the column earth just short the green wire or soldered joint to the steering wheel hub with a jumper wire or screwdriver and the horn should operate---If not there is a column earth issue-----If it does sound properly at that ,then it's either the button contacts or the button housing not earthed to the hub properly.
William Revit

Resistance is not a good test for bad connections as by their very nature bad connections are likely to vary depending on how much current is passed through them, which is why they should always be checked under their normal operating conditions.

Willy's test is with both horns connected and the engine not having run for a few minutes.

As before if you put your voltmeter between the steering wheel nut and a good earth elsewhere you should see 0v whether that green wire is bridged to the nut or not. If you see any voltage when bridged then the column earth is high resistance.

You can get away with the odd volt lost somewhere but from your previous tests as well as losing 4v somewhere in the earth path you are also losing 2v in the 12v path through the fusebox, so the horns are trying to operate on half the design voltage.

paulh4

Thanks all,

Some more test results via a good earth:

Earth to steering wheel nut = 0V
Earth to Green wire at horn= 12.7V not pressed/2.4V pressed

Short green wire to hub - some sparks, and not sounding properly.

Earth to front purple wire at fuse box = 12.5V not pressed/10.8V pressed

Earth to front brown wire at fuse box = 12.4V not pressed/11.3V pressed.

Earth to rear brown wire at fuse box = 12.45V not presssed/11.3V pressed.

(I couldn't reach the rear purple at the fuse box.)

Nat

"Short green wire to hub - some sparks, and not sounding properly."

Sounds like bad column earth or how the green wire attaches to the slip-ring, brush and purple-black wire under the column cowl. If you remove the cowl (four screws) and earth the purple/black where it connects to the horn brush do they sound well then?

Horn push probably OK, but ideally you would earth the spring on the side of the horn push to the hub so you can test that green wire with the button pressed (fiddly it's true).

The next three statements make me think the fusebox is not wired up correctly. All the brown wires should connect to the front of the fusebox and all the purple wires connect to the back. Power comes in on one of the brown wires and daisy chains off to one or two others. If you only have two brown wires then it's a 72 model but if the have three then it's a 73. I suspect you have a 73 as normally the two brown wires on a 72 would be in the same spade connector (but those wires may have been cut and reterminated), the extra brown on a 73 has its own spade connector. Power goes through the fuse to three purple wires (both models) again two wires in one spade connector and the third in a second.

If a brown wire and a purple wire have been transposed then some brown circuits will fused when they shouldn't (well, from the factory at least) and some purple circuits not fused when they should be.

The voltages there are not too bad, some additional voltage will always be lost in the wire down to the horns, and spades and fuse-holders get tarnished over the years.
paulh4

Sorry, I was a bit ambiguous with my front/rear terminology.

Purple wires are at the back of the fusebox (away from front of the car), brown wires are at the front of the fusebox). By 'rear' purple and brown, I meant inner connections closest to the inner wing rather than the more easily accessible ones closest to the engine!


Nat

Understood :o)

Also interesting to see that even though there are only two browns and two purples they are in individual (and original) spade connectors.
paulh4

Incidentally, is that Black Tulip?
paulh4

Its midnight blue, a colour change by a PO as part of a restoration.

There are actually 2 purple wires going to the one outboard connector- partially hidden in the previous photo.

By earthing the purple-black at the cowl - the horn sounds great!

The voltage across the purple-black to the green wire with the horn button earthed to the hub is terrible when the horn push is pressed: 0.55V!
Nat

It sounds like the actual steering wheel/shaft isn't getting an earth.
Maybe try a jumper wire from a good earth to the steering wheel and see what happens.
William Revit

0.55v isn't that bad, especially compared to the 4.2v you got at the horns when the other end of that wire was measured. But if I have read correctly your meter was connected between the green inside the hub and the purple/black at the cowl so is only measuring the resistance of the slip-ring and brush i.e. not the horn button or column.

Ideally do that test again but with one side of the meter somewhere on the column under the cowl and the other on a good earth away from the column and that will give you the column resistance.
paulh4

Some more results:

Jumper from good earth to various locations: steering wheel; column under the cowl; contact of horn push that contacts the steering wheel hub - no improvement to sound.

Voltage across good earth to steering column 0V not pressed /1.4V pressed

Across good earth to steering wheel nut 0V not pushed/1.4V pushed

Across good earth to horn push contact 0V/12.4V


Nat

Are you earthing the horn button housing to the steering wheel when you're doing your testing----
William Revit

You said before "By earthing the purple-black at the cowl - the horn sounds great!"

But earthing the column and steering wheel nut then pushing the horn button there was 'no improvement'. I assume it made some noise but poor? If it made no noise at all then as Willy perhaps the earthing spring on the horn push body wasn't making good contact with the wheel hub.

But if that was earthed then a poor sound when earthing the column or steering wheel nut has to be the horn push or wheel hub that is the problem.

Try the same earth that made the horn sound great connected to the earthing spring on the body of the horn push - the 'single spring' in the attached which was my MotoLita - push the button. Then try that same earth on the green wire.

If neither make a good sound then the problem is the connection between the green wire and the purple/black via the slip-ring and brush.

But if the green wire is good and the earthing spring poor then it's the horn push contact itself.



paulh4

Thanks both,

Yeah, I'm earthing the single spring of the horn push (as shown in Paul's photo) to the steering hub when doing the tests (except the one where I connected the good earth directly to the spring).

By no improvement, I meant that there was a sound, but it's poor rather than no sound.

Trying the good earth to the spring and then to the green, it looks like the horn push itself is a bit suspect - unfortunately it doesn't look easily disassemble-able (and re-assemble-able)
Nat

Unlike the MG horn push which can be dismantled to it's individual components and more importantly reassembled my MotoLita had spire clips on plastic pegs so risk of breakage is high.
paulh4

I installed a relay and the horns now are even and loud. It takes some rewiring but the relay is located near the horns with a direct connection to the fuse block. The original circuit is adequate for firing the relay.
Glenn Mallory

I put my relays in the cabin by the column for protection as the V8 was a daily driver used in all weathers including snow and salt.

I interrupted the purple/black there and put the one from the horn push on one side of the relay winding and the one to the horns on one of the contacts. Then tapped into the purple for the other contact and picked up an earth for the other side of the winding from where the relay was mounted.

That overcame a high-resistance column earth and was good for 30 years, it probably would a dodgy horn contact as well but there is more likelihood of that getting worse over time.
paulh4

This thread was discussed between 26/01/2026 and 21/02/2026

MG MGB Technical index

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