Welcome to our resource for MG Car Information.



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 MGA - Extra fuses re-think and re-work.

I added two extra fuses to protect all the ignition components and the lighting a couple of weeks ago, quite a simple job and it seemed to work fine, but I did notice that with the headlights on, the lighting fuse did run slightly warm. I checked across the fuse with a meter and saw 0.01v drop, not a lot, but it got me thinking about the consequences of the fuse failing with fatigue and being plunged into total darkness; no lights at all! I felt I needed to do two things; relieve the pressure on that one lighting fuse and split the lighting circuits so that I don't lose both the head and the side lighting at the same time. So here is the re-worked fusing system just completed this morning. First of all, the extra two-fuse box I previously fitted is replaced with a Lucas 7FJ side entry 4-way fuse box. By a happy coincidence, the distance between the mounting holes on this fuse box exactly match the centres of the holes on one of the dashboard steady brackets, as you will see from the attached picture. So this is how I have wired it; the main power lead from the A1 terminal(brown/Blue) is disconnected from the ignition switch and connected to the feed side of the first fuse, the same side of that fuse (there is a double Lucar connector on both sides of each fuse) receives the wire from the ignition switch that goes to the common of the lighting switch, the other side of fuse one goes to the ignition switch via a new wire. The blue wire is disconnected from the lighting switch and is connected to the distribution side of fuse two, the red side/tail wire from the lighting switch goes to the distribution side of fuse three. A new blue and a new red wire are provided from the feed side of fuses two and three respectively to lighting switch terminals S1 and S2. The thick red foglamp wire is extended via a snap connector to the other distribution side lucar teminal of fuse three. The two red dash light wires are connected via a double snap connector to the distribution side of fuse four and finally the feed sides of fuses three and four are linked with a short piece of wire with a Lucar on each end.
So this is how the car wiring is protected: In the new fusebox, fuse No.1 (35A) supplies the ignition and fuel pump and the original fuse panel Ignition items; I have reduced this fuse to 25A so that it blows before its 35A supplier. Fuse No.2 (35A) supplies the headlamps. Fuse No.3 (25A or 35A if foglamps fitted) supplies the side and tail lamps and foglamps if fitted. Fuse No.4 (15A) supples the dashboard lamps and map light. Check out the picture to see where I have fitted the fusebox (cover removed).
Hope you haven't nodded off to sleep!

Lindsay Sampford


I often wonder how the one fuse copes with every thing
I also run the Kenlow and soon eo be fitted a brake fluid alarm on the original fuse. been ok so far but I do carry a spear fuse.


David swaine

Lindsay, I am also debating extra fuses, but really if you lose both headlights going fast on a dark lane, the sidelights are really no help. Trouble is, to fuse L/R separately is a lot more complicated b/c of the dip switch. I think you would need 4 fuses, one in each filament.
Art Pearse

You're right of course Art, but where do you stop with electrical security? You also have to think about the wiring TO the fuse. The closer to the source, the more effective the fuse is with regard to wire safety but more circuits will lose power if it blows. To protect individual circuits and maintain the same level of wiring protection would mean wiring everything individually right the way back to your fuse panel, so your standard MGA wiring harness would have to go and you'd have to wire the car from scratch! The left and right headlamps don't split until they get to the front panel, so if you place your fuses there, you have a lot of unprotected wiring in front of them.The same applies to the rear of the car. The fuses are there just in case, the best thing is to make sure that all your wiring and fittings are as secure as you can make them and not leave anything that can chafe or fall off and short out. But I have seen enough burnt-out wiring harnesses to know that these things do happen, so I feel happier knowing I have a bit of protection there.
Lindsay Sampford

Lindsay, We did a similar modification to our 1600 during the rebuilt. It now has i think seven fuses to seperate and help isolate problems. The relay is for the isolation switch for the power.

DJ Lake

Could someone who has added fuses show where they added them on the origional electrical schematic. This would be a big help for the rest of us thinking about doing the same. Thanks and have a good day!

John Progess

Is this any good to you John?

Lindsay Sampford

John, just noticed an omission on the diagram. The red wire(41) from fuse No.4 to the panel dimmer should also connect to the red map light wire. I used a double snap connector at this point with the three red wires on it. The beauty of this arrangement is that it reduces those four wires that congest the light switch S1 teminal to just one wire!
Lindsay Sampford

I did mine with 8 fuses but put the fuse box in the engine compartment for ease of access, I don't worry about concours!!
Here is my wiring diagram on the modified 1600 original with wire changes in red.

Robert (Bob) Midget Turbo

Here is the view under the bonnet

Robert (Bob) Midget Turbo

Bob, I see that all your lights run through fuse No.1, what rating did you use here? I found that a 35A ran a little warm, which is one of the reasons why I split the lighting. If yours is a 35A, have you ever had it fail? I think that to go any higher than a 35A fuse would result in the wiring burning out, or at least being damaged, before the fuse blew. What do you think? The only thing that concerns me about my arrangement is that I have all the ignition switch dependant items running through a 35A fuse. Theoretically it should take it, but sometimes theory is not the same as fact!
Lindsay Sampford

"I found that a 35A ran a little warm, which is one of the reasons why I split the lighting"

Lindsay, what kind of floodlights are fitted to your car? Have a thought for us poor guys on standard fuse layout! That 35 amp fuse protects everything except the horn!

You guys are making me nervous!
Neil McGurk

Neil, if your wiring is bog standard the 35A fuse you are refering to protects the wipers, indicators/stop lamps, heater blower and fuel gauge. The other fuse, as you say, serves the horns. Everything else is unfused. If you don't believe me, pop both those fuses out and your car will start and run and the lights will work! This means, for instance, that if the live wire inside you number plate light falls off and touches frame while your lights are on, it could melt the insulation all the way back to the starter motor switch depending on how quickly you realize and switch the lights off. All British cars up until about the mid eighties (that I have looked at the wiring diagrams of) are wired this way. Many car fires are started by electrical faults; if you don't count those that are stolen and set fire to by morons with nothing better to do!
Lindsay Sampford

35 Amp Lindsey????????

Flippin heck I hope not.

The fuse rating is calculated by a very simple formula Voltage (V) X Amps (A) = Power (Watts)
So we know the Watts (Value of all our lamps) and the voltage (12 volts nominal) so calculating the amps is easy.
Headlights (main) 2 X 65Watts = 130W
Sidelights 4 X 5 Watts = 20W
Instrument panel say 4 at 2.5 Watts = 10W

Total power.. 130 + 20 + 10 + 160W

Power (W) divided by Volts (V) = Amps.

160 divided by 12 = 13.33 amps

Then we generally multiply by 1.5 to find an optimum fuse rating and wire size and that should give us about 20 amp fuse and circuit.

As you rightly say in a standard car the lighting circuit is not protected and any short (generally on the rheostat) will effectively set the car on fire!! But from a power point of view remember the dynamo on our cars is capable about 25 amps and thus you will realise that our total usage has to be below this otherwise we would suffer from a flat battery when driving on a wet night!!
Robert (Bob) Midget Turbo

On modern cars, when switching to main beam, both high and low are on simultaneously during the switch. Is this the case with our MGA foot switch?
Art Pearse

Sorry Lindsey, of course I meant everything that is fused, except the horn (being on the other fuse).
Neil McGurk

Bob, I wasn't thinking in terms of the fuse rupturing due to overload, but running close to its safe working current, which in the case of a "35A" fuse is 17amps. I was wondering if the continued warming and cooling, over a period of time, could weaken it so that it blew below the rated working current. I am on sealed beams and I calculated about 12.3A with all lights on.
I think the dashboard is probably the most dangerous area on the car; if a dasboard lamp holder falls out, there is a possibility of the lamp breaking and earthing out on something, this is why I have fused the dasboard lamps separately at 15A.
Lindsay Sampford

Thanks for the drawings, they are exactly what I need being a little electrical challanged as I am. Have a good day!

John Progess

Became worried about lack of fuses from previous posting and planned to sort this in the winter. I was considering using the existing lighting circuits to just power relays switching power to new lines to the lights. There has some debate on this on the spridget site as the frogeye shares the same economical use of fuses. But of course this would become an obvious deviation from originality.
Graham M V

Graham, I think that relays are a bit of an over-complication. My extra four fuses protect everything, it took me about three hours to do, and quite a lot of that was "thinking time". It required no hole drilling and would be reversable in under an hour. The existing fuse panel stays and the new fuse panel is out of sight but easily accessable. See first post of this thread and the diagram I posted later.
Lindsay Sampford

IMHO, far from being "a bit of an over-complication". relays are the only sane way to wire anything with a heavy current draw. Pulling near 15A through the entire lighting circuit looses a LOT of light, not just from the headlamps but also the rears, panel, heater blower, etc., overloads the aging switchgear, and requires fuses far heavier than would be required if the main loads were taken off prior to the feed to the lamp switch. I have found that it is simply not possible to get voltage drops down to anything reasonable while still running the main lamps through the OE wiring and switches. The improvement in rear lamps alone is more than ample payback for doing the job correctly, as anyone who has followed one of these cars on a rainy/foggy night can tell you. It is not necessary to relay the rears, as just removing the headlamp load gives most of the benefit.

This is a representative table of light output versus voltage AT THE LAMP. Measure it! Engine running at 2500, system hot and stable.
10.5V : 510 lumens
11.0V : 597 lumens
11.5V : 695 lumens
12.0V : 803 lumens
12.5V : 923 lumens
12.8V : 1000 lumens <--Rated output voltage
13.0V : 1054 lumens
13.5V : 1198 lumens
14.0V : 1356 lumens <--Rated life voltage
14.5V : 1528 lumens

I find that even on a system with all new connections and everything "good", there will be a drop on the order of 1V from the Alt/Gen to the bulb, and that is around 20-25% of light output, worse if everything else is not optimum. With relayed systems I can usually reduce the drop to < 0.25V.

Here is a summary of measured changes by cleaning up and relaying, and adding fuses to an MGB recently:
Old/new voltages:
Headlamps 10.5/13.8 , Taillights 10.9/12.8 , Ignition 12.5/14
Brightness change from wiring revamp:
Headlamp 250%, tail lamps 170% Ignition power available: approx 150% this feeds brake lights, wipers, heater blower, etc. Fan power change is about the same as Ignition.

Do note that on MGA or any car with the 2 unit control box, the headlamp (relay) power feed needs to come off the A1 terminal of the box, since there is an internal winding meant to increase voltage output when main lamps are on. Use this to feed the headlamp relays via fuses, and the OE wires to activate the relays.

Email & ask me for "electrobabble" for more details.

FR Millmore

It's true what you say FRM, but an MGA is a simple car, that's what make it fun to own, maintain and drive. Relays are fine when they are new, but they too can suffer with light, dirty or coroded contacts, having spent my working career maintaining analogue telecom equipment I am well aquainted with relays. I will stick with my "dim" headlights, if they do go out on one of those foggy nights I stand a better chance of getting them working again if I haven't got to consider whether or not the fault is in the relay.
Lindsay Sampford

Why does installing a relay give you less voltage drop? Do you use thicker wire from the relay to the end user? I thought relays were just to save load on the switches etc.
Art Pearse

You've got it Art. You can run a shorter route and if you like, use thicker wire. The relay can be anywhere along the route to the lamp or whatever and is controlled by lighter gauge wiring. The long and the short is that the wire to your lamp doesn't have deviate to where the operator of the switch is sitting. The controlling circuit can be fused at a low current, whilst the the supply is fused at a higher current. It is good sound electrical engineering but it does add extra stuff which can go wrong.
Lindsay Sampford

The primary ingredient for less voltage drop is a larger wire. The relay does nothing by itself without a new fat power supply wire to bypass the original smaller harness wire. If you can run larger wires for the entire original circuit you don't need the relay. The only other thing a relay offers is to bypass an original switch that might not be up to the task of carrying increased current load, like uprated headlights or add-on driving lights.

If you are installing more than one device that needs more power, and those devices are physically close together (like driving lights or fog lights close to the headlights), then one new fat power supply wire and some relays could reduce the number of new fat wires you need to make the new devices work. I now have an information web page dedicated to the installation of relays for headlights, fog lights, and driving lights. See here:

I would not recommend the increased complexity of a relay installation to bypass a faulty or inferior replacement switch, like some of the new brake light pressure switches for instance. The proper solution is to buy a good quality switch and keep the car simple.

For my MGA I did long ago pop for the $10 halogen headlamp bulbs, but I have not installed relays (and I get on quite well at night). A couple years ago I switched from generator to alternator, which is a personal decision depending on your view of originality requirement. The primary advantage of the alternator is not more power (unless you install more power hungry accessories), but keeping voltage up with engine at low speed. The alternator will do more for bright lights at idle speed than everything else you could do combined. Incidentally, I had no problem with the original generator (or dim bulbs at idle speed). The alternator was a reaction to poor quality replacement control boxes.

I do use a few miniature relays for control logic, like 4-way flashers and rebuilding the original MGA 1500 turn signal relay box.

Last year I did install one new fat power feed wire and a power relay for an air horn with instant-on air pump motor (because the relay came with the new horn). The relay is (was) triggered by the original wire to the horn button. Within a few days of installation the new power relay crapped out and left the horn dead. On the spot I swapped some wire connectors to connect the new fat power wire and the original horn button wire directly to the air horn pump. It works with perhaps 10% less loudness, still much louder than the original dual Windtone horns. I never bothered to replace that failed relay, so now I'm back to having one new fat wire and no power relays in the car (and I still like the simplicity). If I would run one more fat wire between the air pump and the horn button the maximum loudness would be restored, and there would be no purpose in life for the relay.

One thing to keep in mind is the large number of new wire connections that come along with installing relays (and additional fuses). Every new wire and relay and fuse and wire end terminal is a potential item for future maintenance problems. If you would do any of this, do be sure that all of the new electrical connections are secure and the new harness wires properly strain relieved with rubber grommets anywhere the wires pass through a bulkhead. This will be your future maintenance problems you are going to create or avoid with the quality of the installation. To that end, I HATE LUCAR CONNECTORS, love screw post terminals.
Barney Gaylord

Sounds like your career with open relays got to you!
Today's automotive relays fail so rarely that I've never had to change one, except on a few late MGB where the IGN relay is overloaded because it doesn't have a separate fan relay. Even then it's the wire ends that cause the trouble.
Fit the relays in sockets and carry a spare for the price of a sparkplug or two - beats hell out of running into trees, tracing/repairing wires and switches beside the road in a rainstorm, or trying to get your car out from under a following SUV. Or are those the simple things that "make it fun to own, maintain and drive"?
A clever lad like yourself should be able to figure out that a simple jumper lead across two relay socket terminals will cause the lights to shine, even if the rest of the lighting circuit is toastado. (Relay socket terminals are super easy checkpoints with a test light!)

From the aforementioned "Electrobabble":
"For example, on a typical car, such as an MGB, the headlamps are at the end of about 20 feet of wire, involving 20 or more connections and two expensive switches. These switches operate near their load limit, are frequently in bad condition, and are sometimes difficult to obtain. Each circuit going through a switch has at least 4 contact points: two where the wires attach, and two inside the switch. Every connection and contact causes a power loss, which is transformed to heat, which in turn degrades the connections still further. The total power loss in a headlight circuit is frequently as much as 2 or 3 volts. A portion of this may be due to undersized wiring in the power circuit."

The secret to relays is short load wire paths with adequately sized wire and minimal number of connections. The relay contacts are the only load switch point, and they are far better able to handle high currents than any common dash switch. Much higher contact pressures, better materials, and very high actuation speeds make the difference. Compare the contacts in a Bosch 30A power relay with the column mounted dipswitch contacts on any late MG. The older footmount dip switches are more substantial, but they suffer from moisture problems (especially if used year round) that cause them to operate very slowly, which eventually kills the contacts. I used to replace the foot mount switch on my big truck at least once a year, until I got tired of having no lights for a while every time I switched beam. I fitted a dipping relay from a junkyard VW, operated by a momentary contact microswitch I could touch from the steering wheel. The relay was probably 15 years old when I got it, and I used it for another half million mostly night time miles, and it's still good. And, the lights were noticeably better!

FR Millmore

Well I tried to "edit", in response to Barney, but I guess I'm too slow. It disappeared and I can't take the time to redo now - more later I promise, or is that a threat?

FR Millmore

FRM, how far is Illinois from Pennsylvania, is there any chance of getting you and Barney in "The Ring" and us watching it on YouTube? It's alright, I'm smiling as I type this! I can see arguements for and against relays. Certainly if I was starting from scratch and building my own wiring harness, I would seriously consider including relays in the plan, but adding relays to the stock harness might be a bit messy. I shall do some thinking about "relaying" the headlight circuit at least. FRM, do you replace the existing wires down to the front of the car with heavier ones or use relays to shorten the route from battery to lamp. It's only about six feet at most from the "battery post" on the starter switch to the farthest headlight, so it might be possible to use the existing wiring with a little modification, is this what you do? Hang on a minute, I think you are persuading me! Isn't that what this forum is all about!
Lindsay Sampford

I have to disagree with FRM on this one, the relay has as many joints as the original switch two connections on to the relay base (2 connections onto the switch) and both have a contact inside. The contact pressure on the inside of any switch or toggle switch is the same as or more than any relay!
Where the relay circuit can gain is by having wiring simply for the headlights instead of carrying all the current for sidelights headlights etc. Voltage drop is a function of current voltage and resistance. The resistance of a wire is extremely minimal so any volt drop is due to poor connections. A volt drop of 2 or 3 volts over and above a similar relay circuit suggested by FRM would mean someone had a seriously bad connection somewhere.

However relays can be a slight advantage but if you are serious about brighter lamps at the rear then fit 10 watt bulbs instead of 5 watt!!

The MGA is a simple car and works well, I am with Barney on this one it does not need relays.
Robert (Bob) Midget Turbo

Bob, I do like the idea of just having a relay set-up for the headlamps. It would be dead easy to do. Heavy duty wire for the headlamp power from the A1 terminal via a fuse down to the bonnet latch cross member. Two relays mounted underneath the cross member, one for dip and one for main beam connected to the existing headlamp snap connectors via soldered tails with bullets. Simple and, in case of a relay failure, instantly reversable. For the little it would cost, I think I'm going to give it a try.
Lindsay Sampford

Oh dear what have I started by mentioning the "R" word? But it's a good debate and interesting to learn the pros & cons.
Graham M V

Hi Bob and Graham, have ordered the parts for the headlamp relay job, just under 17 including postage for relays, wire and everything I need. I'll give it a try and let you know what I think. Should have it done before the weekend. If I don't like it it'll be easy-peasy to put it back how it was!
Lindsay Sampford

Lindsay, Good luck - Be interested to hear how you get on. Graham
Graham M V

500 miles from here to Barney - was thinking I might drive over there and show him, if he'll publish the results! Bring your camera. Now that I am wifeless and to be childless after the upcoming weekend, I could use distraction, might start doing group seminars.

Since the day has gone by, I see that you have figured it out - basically what I do on a retrofit. Not a big deal at all. No need for great big wires at these lengths, I use 14ga US, or 28 strand by old Lucas, voltage drop in the wire is essentially not measurable. I use two modern blade fuse (15A) holders, which come with leads that can go directly to A1, then to relays then to front harness lamp connections. Best to use sockets and the double 87 terminal relays, then separate wires to each headlamp. You will want to also replace the headlamp pigtails, as detailed further down. Ground these directly at wing mount bolts as soon as they come through the fender, and it doesn't hurt to also connect the ground points to the harness, helps with other things.

PLEASE measure voltages at generator and lamps with the system hot and loaded before you do anything, and record it all, Then repeat after each improvement stage. That way you know what and where things are happening, and you might help to get some other people to pay attention.

Send an email asking for "Electrobabble" and I will send it, much more detail.

I greatly appreciate your knowledge and extraordinary website, which I recommend to all and sundry, both for mechanical stuff and as an outstanding bit of web design. However, I must take exception to pretty nearly everything you said in this discussion. Much of it is just plain wrong.

The existing wires are more than adequate for the job, excepting the pigtails, which might be OK on an MGA. The advantage of the relays is in eliminating the circuitous route and a lot of bad or potentially bad contact points, including the light and dip switches. It is minimally invasive to the harness, easy to fit, and IT WORKS.
If you think your lights are good, or even adequate, then I submit you have never had good lights. I'm sure your car is better than most, since you are careful and go over it pretty thoroughly, but I'll bet there is still plenty of improvement possible.
$10 halogen bulbs and condemnation of relays on the basis of a cheap aftermarket part is just as silly as condemning our cars for some of the garbage parts we now get.

There is nothing wrong with real Lucar connectors, and the term should not be applied to crappy aftermarket stuff (including later Lucas ones). If you don't overload or bend them they last forever. I too like screw terminals, but used to replace a lot of expensive components due to heat damaged and galled threads when they were common - like anything else, hamhanded yahoos can destroy those too.

You too should try measuring hot loaded voltage at the bulbs and compare it to Alt V.

"the relay has as many joints as the original switch"
NOT true. A good relay has two contacts in the socket - both better than most wire connectors other than new original Lucas, and a single contact pair made of special silver/silver oxide material. All other internal connections are welded. Most switches, and it's worse on newer ones, have two wire connects, two double junctions of connector/rivet/internal contact, and two for the internal bronze contact shuttle. Heat degrades all these, sometimes severely, from bad wire connectors, shorts in the unfused wiring, and from arcing that inevitably happens due to the slow switching speed. That's six bad internal switch contacts compared to the single high grade relay one. Double all that for the dipswitch.

Even if your assertion that "The contact pressure on the inside of any switch or toggle switch is the same as or more than any relay!" were true for a new switch, which I dispute for all but the most heavy duty switches (these ain't them!), it is certainly not true on switches that have gotten hot, in which case contact pressures can and do go down to zero.

"The resistance of a wire is extremely minimal"
Generally true for Lucas cars, but the headlamp pigtails on MGB era onward cars are a notable exception, being reliably responsible for a drop of 0.33V for the standard pigtail (cumulative for both legs)- that's 8% of light output. MGA may be better than later cars, I recall the pigtail wires as heavier, but one has not wandered by since I started measuring voltages.

"A volt drop of 2 or 3 volts over and above a similar relay circuit suggested by FRM would mean someone had a seriously bad connection somewhere."
It is generally not a matter of "...a seriously bad connection somewhere", but of many somewhat bad connections all over the place.

Following are actual measured voltages from several cars, before and after some remedial work. I have measured dozens of cars and these numbers are typical.
Alt V, Initial headlamp V across bulb contacts/final headlamp V.
74 MGM 13.30, 10.25/12.89 Relays, connectors, pigtails (needs Alt!) Also fitted H4+30 Hellas, called to complain of number of deer I had put along the road!
73 MGB 14.0, 10.2/13.5 same fix, + H4 Hellas
77 MGB 14.0, Headlamps 10.5/13.8, Taillights 10.9/12.8, Ignition 12.5/14.0. Same fix + new ign relay, + fan relay + revised fusing. + H4+30 Cibie - like driving in the daytime.
77 MGB NA, 10.2&9.6/11.5 some sleeve connectors only
72 TR6 14.23, 11.26/NA, no work done.
So, 2-3V is the norm, that's 40 to 60% of light output.
Compare these V to the table of lamp output above.

" fit 10 watt bulbs instead of 5 watt!!"
Mostly that is just going to increase the losses in the system, leading to more heat at bad points, with subsequent failure. At best it will just get the light output somewhere near where it would be if the system were fixed correctly. And,it doesn't do doodley for the invisible panel lights!

Finally, I cannot tell you how many people just don't drive at night because they can't see. Many times they don't even know why until you question them intently. I had one customer who had been to the eye doctor to find out why he couldn't see at night; they couldn't find anything so he scheduled more tests. After I measured his voltages and fixed his MGB and diagnosed his Geo as having the same problem, he canceled the appointments, but it took a couple of hours discussion re lights and a trip down the road before he figured it out. He called me when he got home and complained that now he had to use the dipswitch! They don't drive at night, they don't go to events where they might have to come home in the dark, they don't take trips. Bad lights, slow wipers, and poor heat/demist are the results of badly maintained electrics. All this makes them generally uncomfortable in their nice British cars, and the cars stay in the barn - that's what I'm trying to fix. (I need folk to use their cars so I have some work, too.) And, I don't want to hear about any of my friends crashing in the dark!

FR Millmore

you create a good case FRM and to a great extent it is pretty true. However the gain from relays is minimal unless the system that is being replaced is fundementally faulty. Like when I hear the performance upgrades from a K and N filter as an example. The figures stated will probably be 10% more efficient! yeh compared to what? well compared to a clogged up paper type, why do they not simply compare it to a fully functioning new paper element? because if they did that they could only maybe claim a 1% increase.

However an increase is an increase and irrespective of measuring voltages we need to wait for Lindsey to give his impressions of enhanced night driving.

I/we look forward to your opinions on the improvements that you see Lindsey! I do agree however that for the head lamps it is a very easy modification, not to good however for those keen to keep their cars standard and up to concours condition. That in itself almosts gives me the enthusiasm to do it! :-)

Robert (Bob) Midget Turbo

I beg your pardon Lindsay! (Poor spelling on my part!)
Robert (Bob) Midget Turbo

Hi FRM, yes I was planning on taking "before and after" voltage measurements at the A1 terminal and at the headlamps, I have even ordered a new digital multimeter for the job as my old analogue one, after years of use, is just about clapped out! It wasn't until I sat down and thought about what you had said and figured out how to do it, that I could see how simple a task it was.
I have ordered the 28 strand wire that you mentioned in your last post and even got it in the right colour for "fuse to headlamp relay",namely Blue/Brown! I have also got relay sockets so that I can change them out if I need to. It will be interesting to see what the headlight mod does for the tail lights, they do dim slightly when the headlights are turned on but that is not a fair test because I can't run the engine at the moment so it could be a voltage drop on the battery rather than the tail light circuit.
Lindsay Sampford

Just a warning about your digital multimeter. The cheap one that I have is badly affected by the ignition when the engine is running and produces very dodgy results.
Malcolm Asquith

Not affected by the ignition Malcolm, but the voltage regulator is continually chopping the output voltage to try and maintain 13/14 volts and the digital meter simply can not update quick enough so appears to be reading all over the place. I use an analogue meter (AVO 8) when I set up my voltage regulator charging voltage.
Robert (Bob) Midget Turbo

Man, a lot happens here in one day.

Yes, FRM, I would like a copy of "Electrobabble". Click email link above.

Lindsay, -- It is not necessary for anyone to travel long distance to get "in the ring". I wouldn't intend for a civil discussion to deteriorate into a fisticuff or a religious war.

I don't think I have said anywhere that FRM is wrong in anything he has said (yet). Surprise? Some of us may take exception to his inference the everyone from Cecil Kimber to 90% of all current MGA owners might be "less than sane" simply because we do not find it necessary to add relays. This is a general philosophical difference, comparable to installing disc brakes on a 1500 car or adding a brake booster or a third brake light, and deciding for yourself if such things are advantageous or appropriate for your car. The only problem comes when someone insists that their non-original modification is the only sane way to allow a vintage car on the road.

The flip side is the common axiom, "If you want a modern sports car buy a Miata". In between there is a lot of opportunity for anyone to make whatever mods they deem appropriate for their own car. One of my initial comments was, "I get on quite well at night". That means I treat my vintage car like a vintage car, I don't overdrive the headlights, and during tens of thousands of miles of night driving I have avoided hitting any of the multitude of deer and bears I have encountered on the night roads. So far I haven't been rear ended either (perhaps due to care on my part for watching the rear view mirror and not slamming on the brakes unnecessarily in front of someone). I fully understand people who have some extra time and money to spend on their car and might like to make some modifications. I've done some of that myself, even upgrading the headlight bulbs, but just haven't come to grips with the cost/benefit ratio of the substantial mods and increased complexity required with relays for the small additional increment for brighter lights. I see it as a matter of personal preference, not a matter of life or death.

I'm glad to see FRM's follow up notes with more details of his relay installation. Now I get to nit pick it a little bit in public. His 14 gauge wire actually is a "great big wire" compared to the 16 gauge headlamp circuit wires in the original harness (and I agree on the size choice). The blade fuses are a good choice over tubular fuse clips. If you have an alternator with internal regulator you don't need to connect to the A1 terminal of the control box (which may have been removed), but can connect direct to the main battery cable. Also notable that he grounds the lamps directly to the body bolts to bypass the original return ground wire and snap connectors, nothing to do with relays but a good idea regardless. These are things you can do to improve the original lighting before adding relays (none of which I did).

I don't suppose the wire gauge of the headlight pigtails is very significant, as it amounts to two feet of wire carrying only 5 amps of current (for a 60 watt bulb). Even if it was only 18 gauge originally that comes to a voltage drop of 0.065 volts (anyone interested can look it up and do the calculation). You might recover half of that loss with larger wire (likely not worth the effort). To have anything like the 0.33 volts loss stated you would have significantly dirty end connectors (part of the original problem).

I agree that modern sealed automotive relays are very reliable, even the plug-in socket contacts are good. The issue is how you connect the harness wire to the relay socket. Modern cars have the female contact in the relay socket crimped onto the wire with a gas tight crimp joint, so the socket connector is effectively part of the harness. If you use a relay socket with Lucar connectors on the sides to connect the harness, you take a giant step backward in reliability by incorporating many additional wire connectors.

Now there was a little personal challenge I couldn't pass up, because I didn't believe a small detail in FRM's first message above, that he was measuring voltage at headlamps of only 10.5 volts before adding relays. So last night I did the same test on my car. My alternator puts out 14.5 volts at all times regardless of engine speed. The (prior) generator and control box were adjusted to have slightly higher output voltage at higher engine speed, about 15.5 volts at the "D" terminal on the generator with headlights on. Voltage at the A1 terminal on the control box would be lower, more like 14.5 volts (similar to alternator output at higher speeds). The main difference then is that with the generator at low idle speed the system voltage could be as much as 3 volts lower, which does indeed make for dim lights at idle. But that's a difference of alternator vs. generator, nothing to do with relays.

When reassembling my car a year ago in July after body restoration I put the wiring harness and lighting system back together in original configuration using the 1986 vintage standard type wiring harness and mostly new snap connectors. Nothing special, just ensure everything works, and I haven't touched it since. Now I connected my rather expensive Fluke digital volt meter to the snap connectors at the left front headlight pigtails (farthest from the power source), and I measure a nice steady 11.9 volts with engine idling. This includes the original lighting switch, dipper switch, long runs of original size wire, and all of the snap connectors (including the black ground return wires). Just to be sure I switched the lights on and off and kicked the dipper switch several times while checking voltage with each step, and no change. This is a significant difference in both voltage and light output from FRM's noted starting point (10.5 volts). This may serve to show that there is much to be gained by putting the original components in proper working order first. It also takes a bit of the edge off of the ratio of light increase due to relays, and the severity or dimness at lowest voltage.

A agree with FRM's overall assessment that that a change of voltage from 10.5 volts to 13.8 volts at the headlamp will make for substantially more light. I do not agree that it all comes from installing relays. If I start with my measured 11.9 volts at the final snap connectors, how much improvement might I expect by grounding the lamps on the body and soldering wires together to eliminate a multitude of snap connectors? That would be the starting point before consideration of how much improvement you get from the relays.

The point is, if my headlights always have a minimum of 11.9 volts in standard configuration, maybe that's good enough for normal night driving. (Apparently I think it is). The personal choice then is, at what point do you call it good enough to keep as original, or decide it is not good enough for your personal taste and feel the need for modification? That's pretty much all my side of the discussion was about.

As to FRM's offer, I would be happy to post information about his headlight relay installation(s) on my web site. I am delighted to place the facts and data in public access. He can even write the article himself. However, I will not post 10.5 volts starting point without some qualifying statement (when I measure 11.9 volts on my standard system), and I will not repeat his religious style insistence that relays are the only sane way to configure an MGA. There are some very good reasons for using relays if that's what you have in mind, and also some good reasons for not modifying the car.
Barney Gaylord

This is great fun!

I'm with Barney on this one. Keep it simple and original--This old stuff really works good if properly installed and connected. With that in mind it is most important to pay partucular attention to good tight CLEAN connections. The only thing I add to help this is some dialectric grease on the bullet connectors.

A parting comment--the factory only used two relays on the MGA and those were for options. One for the optional horn/flasher switch and the other for dual fog lamps. Both relays were deemed necessary to take the load from the switches and increase their longevity.

So to all you "wire nuts" out there, I say "rela-shmeelay", do it right and keep it simple.
James Johanski


Back on my digital meter (Small yellow job bought on special offer from Maplin).

I have electronic ignition and the meter was giving a variable reading when held near the electronic bits without connecting the meter wires to anything. I wouldn't have thought the voltage regulator cutting in and out happened often enough to cause a problem.

When I changed to an analogue meter it was quite easy to notice the regulator operate and adjust its operate point which was the object of the exercise. It is now set at the upper end of the voltage range and there was a small but noticeable change in headlamp brightness.
Malcolm Asquith

No problem Malcolm, we are all entitled to our own opinion and we must accept the consequencies of our beliefs. :-)

Just for your benefit I too adjust the voltage output of my regulator when the engine is running. However I do NOT have electronic ignition I simply use points.

If I use my FLUKE digital multimeter to measure the voltage it is not stable enough to make a good adjustment (due to I believe the refresh rate of the meter and the opening and closing of the regulator points) Therefore I use my old MK 8 AVO moving coil analogue meter to do the job easily.
Robert (Bob) Midget Turbo

I have a cheap RadioShack pocket size DVM for my traveling tool kit. That meter has short test leads. When I connect it to the back end of the generator it goes wacky from being too close to the ignition coil. When I extend the test leads to keep the meter farther away from the coil it works fine. My old Fluke DVM works well under any circumstances, even if I lay it right against the ignition coil. It is a mid 1980's vintage lab quality instrument. Maybe they don't make them like they used to, or maybe you just have to buy the right one.

Even though the regulator relay will be chattering incessantly for square wave modulation of field voltage, output from the generator will be pretty smooth, because the magnetic field does not change that fast. Output voltage is dithering slightly, but only within a very narrow range, so any DVM should work okay (of you keep it away from the ignition parts). Advantage of the analog meter is when you can see the needle drop back slightly when the regulator relay begins to function at the intended control voltage. Digital meter does the same thing, but you only get a reading a couple of times per second, so it's trickier to see the drop back point.
Barney Gaylord

Barney the discussion was not about measuring the consistant voltage at the back of the generator it was about measuring it effectively across the battery for the output of the voltage regulator. I would admit that using a DVM on the back of a generator would have no problems, as the voltage is pretty stable at that point.
Robert (Bob) Midget Turbo

The A and A1 terminals of the control box fall in the circuit between the generator output and battery cable. This should be slightly more damped than the output directly at back of generator, so should be no problem for a DVM.

The "F" terminal on the control box is the one dithering constantly between 0 and 12 volts so it cannot be measured with most cheap digital meters. Some more expensive digital meters may have a special input averaging function that could damp out the pulses and give a good reading of average voltage (similar to using an analog meter).
Barney Gaylord

It would avoid the problems of ignition and regulator effects if you just checked voltage drop on the battery alone - kill the engine. The results would be the same for lost voltages and the readings more stable.
Art Pearse

Gosh what fun reading this!!

As a somewhat side note, I have used many models of Fluke Digital Multi Meters professionally and find that some models are suseptible to noise interference more than others. By this I mean that locating the meter toward a noisy source, while keeping the wires on the same device, will affect the reading. The MGA has many noise sources, Generator brushes, Ignition components, heater motors, and Gen Control box's are just some.

I have personally own a very nice DMM and some $4 DMM's. BUT when I go to set the regulator voltage on my MGA, I use my tried, trusty, antiquated Simpson 260 Analog VOM. Why? because I never have to guess what the actual measurement is. Any iron-vane meter is by its nature impervious to noise injection and yields a true RMS reading, even on DC circuits. Now if I wanted to see the voltage drop across a Lucar connector in a circuit drawing say, 4 or 5 amps, then I would switch to a digital meter. Each type of meter has its own application.

By the way, I have the factory designed headlight circuit in my car, a new(9 years ago) wiring harness,a generator, and have Wagner Britelite (Xenon) headlamps. I can see better in the MG than I can in my DD's.

Now on with the banter.


Chuck Schaefer

Voltage drops are not exactly the same with engine off, but it will be proportional to system voltage. Voltage drop is a function of circuit resistance and current.

With engine running the system voltage is higher, so the voltage drops will also be proportionately higher.

Suppose with engine off you find 12V supply and 10V at headlights (2.0V or 16.67% loss). Then with engine running the alternator bumps supply voltage up to 14.5V with proportionately higher current, and you still have 16.67% voltage loss in the same circuit, so now it drops from 14.5v to 12.1v, then 2.4 volts drop. In any case you must have the engine running to measure final operating voltage at the headlamp.
Barney Gaylord

This thread was discussed between 29/08/2009 and 03/09/2009

MG MGA index

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