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 - VDO Gauge Wiring Problem. 11

Hi

Follow up for the Thread VDO Gauge Wiring Problem.

http://www2.mgcars.org.uk/cgi-bin/gen5?runprog=mgbbs&mode=archiveth&archiveyear=71_2007.dat&access=&subject=71&subjectar=71&source=T&thread=2007050620412729669

Since my last posting I have replaced the alternator due to a bad regulator. I have had the starter rebuilt due to an internal bolt which backed off and into the armature shorting out the starter. I have replaced All connectors and with a dremel, cleaned all bullet connectors as well as cleaned the multi pin connectors under the North American steering wheel.

Nothing has changed. When the system is loaded and I apply the brakes the speedometer gauge dips (or I turn the fan on). I have 100/80 headlamp bulbs with 2 additional 50W auxiliary lights and Halogen brake lights.

I took some readings and they are as follows.
I have checked the battery so many times and it always corresponds with the starter reading I have started my readings at the starter.

Starter to ground 14.02
Alternator + grounded to a bolt in the engine bay 14.34
Alternator + grounded to alternator case 14.41
Brown wire cluster (this is not the brown wires welded together) but a brown wire comes from the welded wires to what I call a brown wire cluster and feeds my 4 relays. At this point I have 13.30. A bit of a drop. What I dont understand is why? This is a direct wire from the starter to the welded brown wires and to the brown wire cluster.
Ignition Brown wire which I used the big ground under the dash 13.27
Brown fuse to the big ground under the dash 13.44

The white circuit
White wire at ignition to the big ground under the dash 12.26
From the white fuse to the big ground under the dash 12.35
From the white fuel pump wire to the big ground under the dash 12.25
Reading from the white fuse side 12.43 to the same fuse green side 12.40

Green circuit
Green wire which goes to the gauges to the big ground under the dash 12.25
Green wire which goes to the brakes to the big ground under the dash 12.31

The drop is obvious. The question is why? I think if I was able to find the reason for the drop in the brown circuit that this might solve the problems in the white and green circuits.

Quite honestly guys I am out of ideas. Could it be one of the components in the system. A bad gauge, a bad fan. A bad ignition switch or hazard switch.?
As the car warms up the problem is not as blatant.

Thanks for any suggestions

Bruce



Bruce Mills

Bruce-
Your readings are mostly excellent, with two exceptions:
1V drop in the brown feeds, almost as bad is the further 1V drop by the time you get to the W after the IGN switch. Both will cause the instrument problem, since the gauges are fed after the switch, as are the brake lights.

I forget what year the car is. Looked it up!

1V drop in the brown feeds:
You are pulling well over 25 amps, and more like 30 once you get the system right, just to run your excessive lights - which are NOT necessary or advisable IF you have GOOD lights.

If you are trying to feed all those lights from any OE wire, it won't fly. Run a minimum 10ga wire from the solenoid to the fusebox feeding the relays. I have found that even with an added 10ga wire, there seems to be a limit on the carrying capacity of the wire stack at the solenoid - on a 77 with two leads stock + added 10ga, I still had .25 to .30V drop with standard 60W headlamps and electric fans. The total there is about 30A, same as just your lights.
The best way to fix this is to run the added feed direct from the Alt+ to the relay fusebox, on the basis that you only are using the lights with the engine running. Some Alts have an extra + connector that is not normally used, and the required connector is in the plug from the factory. Other Alts have a - connector, also not normally used - if you have that, use it to run a ground direct from A- to the main body ground below the fusebox.
Actually, it's immaterial, since the stock Alt won't keep up with demand anyway, so upgrade it to a Bosch or Delco, and run the relay feeds direct from A+. And by all means, run the direct ground too.

If you have an electric fan, it also should have a relay off the same box.
**********

Further 1V drop by the time you get to the W after the IGN switch:
If you still have this after you fix the preceding, then you have big loss either in the browN from the feedpoint, or in the IGN switch.
The cure is to install an IGN relay, as on the later cars. Feed this off the same fusebox as the other relays. N(feed) = 30, B(grd) = 85, W(control; use the W that feeds the fusebox) = 86, WN (to OE fusebox, where you took the W from)= 87
This will remove all the things fed by the OE W-G fuse from the IGN circuit, greatly reduce the load on the switch, and eliminate a bunch of heavily loaded wire and connections from the circuit feeding your instruments.

Given that you have a problem with the brake lights causing the gauge behavior, install Dave DuBois' brake switch diode circuit to kill the spike caused by the brake lights - it will also save the switch (mandatory on replacement switches!).

Yer gettin' there!

FRM


FR Millmore

Thanks FRM

I will be doing some body modifications this winter which included the installation of a fiberglass bumper. I wasn't sure what I was going to do with the extra lights. It was looking like it was going to be a lot of work to incorporated them into the fiberglass bumper. So I will unhook the extra lights and take some readings.
Since cleaning up the electrical system and the installation of a new 100A Delco the lights are noticeably brighter.
When I move my wiring to under the dash I installed a new fuse block which still has places left for additional circuits. I like you idea of running another brown wire feed to the fuse box and power my relays off of that.
I appreciate all the help and it is starting to make a lot more sense now.
I will post my results.

Thanks


Bruce
Bruce Mills

Bruce-
Your relays should be on the RH inner fender or the rad support. The power feeds should be as short as possible. This means the fusebox feeding the relays should also be on the RH inner. Use a Buss 6 fuse block with a single stud common input = $8 from any parts store. Run Power from the Alt+ direct to the fusebox. That gives you about 1 ft wire from Alt+ to fusebox, a few inches to relays, and another few inches to the light pigtails. It makes no sense whatever to run heavy power into the cockpit without fuse protection, then run it all the way to the front again. Defeats the whole point of the relay system, and gives you lots of unprotected and other trouble points, wiring mess, ... and power loss.

And with those lights, you MUST upgrade the pigtails - the OE (less than 18ga)or replacement ones (some are 22Ga!) are not even adequate for 60/55W lights (0.30-0.35V drop from harness to ground = 8% of light output). 14Ga wire and Hella HD plugs from Daniel Stern.

FRM
FR Millmore

FR

I never even though of the pigtail for the headlights. I will see about getting some new ones.

I disconnected the auxiliary lights and went for a drive. Just the main headlights on and when turning on the fan or applying the brakes the speedometer drops.

Turned off the headlights and when turning on the fan or applying the brakes the speedometer drops.

Guess the lights aren't the problem?

I guess where I am is now trying to isolate the problem by by-passing components. For example by-pass the heater switch and see if the problem still persist? Maybe there is an internal problem with the switch?

Bruce
Bruce Mills

Bruce-
By "speedometer drops", do you mean long term or momentary?
If momentary, the arc suppression diode circuit (Dave D) should fix it. I believe it should work on any switched load. Goes across the switch contacts to kill the spikes and also preserves the switches.
Or, in Electrobabble, I speak of varistors, which will do the same.

If it's more than momentary, it would seem that you still have a bad feed to the speedo, before the heater/brake light/instrument split, or a bad ground causing a ground drop or loop.
Seems to me that the speedo should not be so sensitive; such devices normally have spike suppression filters built in - perhaps the speedo is defective. However, spikes are always best killed at the source, either switch contacts or inductive loads like solenoids and motors.
In any event, now that you've got the system basically good, the lights are not the problem causing the speedo thing.

FRM
FR Millmore

FR Millmore

The speedometer drops are momentary. When the fan is turned on, or the brakes applied the speedometer needle will drop from 45MPH to 25MPH and then start to climb back up.

The warmer the vehicle gets, the shorter time duration of the drop, it might drop 1-5MPH, immediatley climbing back to 45MPH.

Right from the start I have always suspected the speedometer gauge. I have keep in touch with VDO through-out this problem and I am fast coming to a point where I provide them with the link to the thread on this supject and ask them at what point do we say the speedometer is defective?

With your help I have almost exahusted any other source of the problem.

I have touched basis with David Dubois and I will try by passing the switch's this weekend end. After that I will contact VDO.


Thanks for all your help


Bruce
Bruce Mills

Much the same as FRM. You have two big volt-drops - one from the brown 'cluster', ignition switch and fuse, and another from the ignition switch to the white.

If the four relays are feeding your high power items then the brown feed to these must come from the solenoid stud as FRM has said.

However with another big drop through the ignition switch, either that is defective and exhibiting a higher than normal resistance, or something on the white is pulling a big current as well. In the latter case that also needs a relay that should be fed off the solenoid.

But even so, the brake lights (assuming they are standard wattage and not something exotic) isn't that much of an additional load and it seems to me the spedometer simply shouldn't be that sensitive, unless there is a partial short on the brake light circuit which is causing it to pull more current that it should, but still within the rating of the fuse.

Bypassing the ignition switch (temporarily) may reduce its volt-drop, but you still wouldn't know whether it was caused by internal resistance of the switch (or connections to it) or excessive current. What you could do is use a clip-on (i.e. non-intrusive) ammeter on the brown at the ignition switch. Operating the brakes will cause an increase in that load. If using an intrusive ammeter i.e. connected to the brown and white at the ignition switch and turning the switch off, you will need a pretty hefty one, ideally 50 amps. A typical 10 amp hobbyists meter may simply go to FSD, or pop the current limiter or fuse. Also bear in mind that intrusive ammeters are not zero insertion resistance and so will cause their own volt-drop, which may be more or less than the ignition switch.
Paul Hunt 2

Well this problem continues to totally amaze me.
Today with no load on the system and the car at 2000RPM, I hooked the + probe up to the speedometer + and the - probe up to the big ground under the dash 13.60
Applying just the brakes 12.82
Applying just the heater fan 12.86
Applying fan and brake 12.19
Applying fan and wiper 12.34
I unhooked the brake lights and the readings were similar.
I then also unhooked the back up lights and the readings were similar.
I then unhooked the wipers as well and the readings were similar.
I won't bore you with the readings.
This was all done with the heater switch removed and by passed
I installed a different heater fan and nothing changed.
I unhooked the headlights and no change.
With all the above still unhooked and I then unhooked the temperature gauge, the speedometer, the tach and the gas gauge in sequence. No real change.
The only thing which made a difference was unhooking the big red plastic connector under the dash.
So as far as I can figure out I have eliminated the green circuit and it's components as the source of the problem. Windshield wipers/tach/speedometer/fuel gauge/temp gauge/fan/heater switch/brake lights/brake switch etc.
The only stuff on the white circuit is the fuel pump and the coil, the OD no longer is applicable.
The relays I have are 1 for the high beam, 1 for the low beam, 1 for the cooling fan and 1 which was used to bypass the ballast resister when starting with my mini hi-torque starter, which is now no longer used as I have a coil with an internal ballast resister. (The wires have been taped back).
The voltage drops very noticeable when the car is cold. When it warms up the drop is sporadic, at times with the heater fan on and applying the brakes, nothing, at other times, just turning on the fan and the gauge drops immediately bouncing back.
We previously talked about installing a new circuit with a brown wire from the solenoid and I bought the stuff to do so but since the lights are not the problem and the cooling fan is the only other draw on the circuit (3000CFM 16") I wasn't sure it was warranted.
Could the cooling fan be drawing enough to be causing all these problems? Which were not noticeable before the addition of the VDO gauges? Thus the sporadic nature of the gauges?
When the cooling fan is on (I have no way of knowing) is it drawing enough power that with the addition of the heater fan and/or brake lights the gauge dips?
I guess there is only one way to find out.
Install the new circuit to run the relay off it?
Thoughts welcome

Bruce

Bruce Mills

Bruce-
All the things you measured/disconnected are IGN switched items.
They all point to a common high resistance problem, and the only place that can be is in the browN feed to the IGN switch, or the switch itself, or the W from the switch to wherever the individual feeds split.

You can figure this out by measuring V to grd at each side of the IGN switch. Or, you can measure Vdrop between the two ends of any segment. As, N from solenoid to IGN switch; Across IGN switch; or IGN to split. Measuring V drop lets you know if and where drops are - it is a waste of time bypassing switches and such when all you need to do is measure drop.

None of these has anything to do with lights AFTER the point where the N feeds separate. If you have not divorced the light load from the IGN feed then the lights will cause a drop. If the cooling fan is on the IGN circuit, it will kill you - cooling fan requires its own feed and relay - only the control power for this may come off the IGN side. (very easy to tell when the fan is operating - put an indicator light after the control, in parallel with the fan motor on the wire from the relay) (the OE fans pair pull 18A cold, 15 after running a few minutes)

All of this is why I said to put a relay in the IGN feed, instead of running that circuit through the switch and associated wires. Heater blower, instruments, wipers and IGN itself are a substantial draw. Even the factory figured this out and fit the relay on late cars.

What big red plastic connector?

The differing behavior cold vs hot is characteristic of bad connections. Constant V drops can be either bad connections or undersized wires.
If the drop is in a ground connection, it can create a ground loop which will affect everything on that ground section: for instance, a bad ground between dashboard and main grd will screw up everything grounded to the dash, AND any measurements taken to the poorly grounded section. That's why you find reference points for both hot & grd, and stick to them.

Remember what I wrote in Electrobabble:
V from any hot to any other hot = 0
V from any grd to any other grd = 0
V from any hot to any grd = system V, as across Batt or A+ to A-.
You will have inevitable deviations from this, but so long as you KNOW WHERE they are and WHY they are, you'll be OK.

The IGN voltage drop is annoying, may indicate a potential trouble point, and slows down the things fed by it, but:
I still think the speedo is goofy - most electronic automotive devices are normally regulated to work OK with system V from 10 to 16, and to withstand "normal" spikes without error. If not, the documentation should say that spike or voltage control is critical AND specify limits.

FRM
FR Millmore

Like FRM says what big red plastic connector and also what effect does it have i.e. what were the voltages then i.e. with fan and brakes on as well as off? With this connector unhooked (disconnected?) what circuits no longer work i.e. what is it feeding?

From your earlier readings:
"Ignition Brown wire which I used the big ground under the dash 13.27"
and
"White wire at ignition to the big ground under the dash 12.26"

You are losing more than a volt through the ignition switch and its connections, in addition to the more than one volt lost between the alternator/starter and the ignition switch brown. These are large losses and should be investigated irrespective of the speedo action.

Whilst in theory the system voltage should probably always be above 12.8v (or the battery is likely to be discharging) I still say that the speedo is more sensitive to supply voltage than it should be. OTOH things like brake lights take a relative large current when first powered because the filaments are cold and low resistance, but only for a very brief instant. This could be pulling the system voltage down well below the 12. something a meter will show, which could be below the minimum system voltage for the gauges (do you know what this is?) and is something you could only see easily with a recording/storage oscilloscope.
Paul Hunt 2

FRM and Paul

The big red connector.
If you look at the Haynes Manual page 196 Fig 10.31, on the left had side of the page about 1/2 way up between #116 (*Heated rear window)and #24 Flasher unit, is junction of wires. On my car this a red connector. The wires going through this are the hi bean and low beam, heater motor and wiper switch, heater motor switch, fuel gauge signals tach negative and voltage stabilizer which has been removed.
I can't remember where I had the probes but I do remember when I separated this connector the reading was 13.60
Next chance I get I will run the new feed and install the ignition relay.
I have suspected the speedometer for some time and what I would like to do is to explore every option and if the problem still persists, then contact VDO re replacing the speedometer. In the interim period my 30+ electical system is getting a good going over and I am learning things.
Thanks both of you for all your patient help.

Bruce
Bruce Mills

Right, that is one of the multi-way connectors that connects the dash wiring to the rest of the harness.

However in my edition of Haynes from left to right the wires in that connector are:

Red/green out to the parking light fuses
Green/yellow and green/pink out to the heater fan
Blue/white in to the main-beam tell-tale
Green/blue out to the coolant temp sensor
Green in to the tachometer
White/black in to the tach sensor
Green/black out to the tank sender
Green/white in to the right-hand turn-signal tell-tale

Blue from the light switch to the dip switch goes through a different multi-way dash connector, and the dip and main wires through one of the column multi-way connectors. The wiper wiring goes through the other column multi-way connector, And the voltage stabiliser goes through another connector that includes two red/white (dash lighting) wires, albeit unconnected electrically.

But there are other confusions, you say that applying the brakes the voltage went down to either 12.82 or 12.19, but unhooking the brake lights the results were similar. Where did you unhook the brake lights? The implication is that if you disconencted the lights, say, at the back but pushing the pedal still caused a volt-drop, then there is some other load on the wiring connected to the switch. If so, this could be caused by a pinched green/purple wire, partial short, taking more current through the brake light switch than it should, but not enough to blow the fuse.

But the two volt-drops I mentioned earlier are the crucial ones, in my opinion. Fixing those may well solve the problem, or indeed if you do have a trapped wire may allow sufficient current to flow to blow the fuse!
Paul Hunt 2

My apologies for not being more precise. You are right of course with regards to the wire colours of the big red connector.

With the no load on the system and the car at 2000RPM, I hooked the + probe up to the speedometer + and the - probe up to the big ground under the dash.
With the brakes unhooked at a green wire 4 way connector
Fan only 12.82
Signals only 13.42
Wiper only 12.99
Fan & signals 12.70
fan, signal, wipers 12.01

"Fixing those may well solve the problem"
I agree. As per FRM's article I want my readings to be close as possible to "V from any hot to any grd = system V"



Bruce


Bruce Mills

That was the period when they went mad changing connectors - seemed every car that came by was different, can't keep them all straight!

It's still a matter of all the IGN switched things being low, hence still in that feed line before all those split. As OE, that would be the N-IGN switch-W-fuse-G circuit. But, you've moved the fuses and presumably changed all that, so you're on your own as to where they cease to be common. As stated, IGN relay fed from the A+ and/or B+ direct is the answer.

FRM
FR Millmore

When I moved the fuse box I keep the fuse and wire configuration as it was. Simplicity. I did fuse the headlights though and I do have a couple of spots left.

I will be adding the new relay and feed this weekend.

And I will report back. Hopefully with good news.

Thanks

Bruce
Bruce Mills

I suppose if adding bits cures the problem then any port in a storm :-/ But unless you power the speedometer directly from the alternator/battery and manage to switch it along with the ignition, or power everything else bar the speedo that way, i.e. the only common point is at the alternator battery, the bad connections will still be there, still causing a volt-drop, and still be affecting the speedo. And if you do the former then the low voltage will still be affecting everything else adversely. You should have a lot more than 12v left on the green with fan, signals and wipers on.
Paul Hunt 2

New circuit installed and all relays feed off it.
No change

I ran out of time and will try the relay this weekend.

After all the checking/cleaning etc I have done, and all the things I have tried with your help. If the relay doesn't correct the problem I think even VDO will have to admit there is only thing left that can be causing the problem.

The speedometer.

I will report back after installing the relay.

Bruce

Bruce Mills

You can't blame the voltage drop on the speedo. It could well be over-sensitive to the voltage drops, but yours are excessive. You must investigate why you are getting a volt or more dropped in the brown circuit, and another volt plus in the ignition switch, white and green circuits. I'm sure all VDO will do is demonstrate that at a given system voltage their speedo works fine.
Paul Hunt 2

FRM
I just got back from the shop where I went to install the relay
"Feed this off the same fusebox as the other relays. N(feed) = 30, B(grd) = 85, W(control; use the W that feeds the fusebox) = 86, WN (to OE fusebox, where you took the W from)= 87"

I understand the N (feed) and the B (grd) but I am a little confused with the white which feeds the fuse box.

I have 2 white wires joined together which attach to the male connector on the fuse box, the white from the ignition switch and a white which goes to the ignition coil as per diagram 17 http://www.advanceautowire.com/mgb.pdf

Do you mean for me to separate these white wires and run the white from the ignition switch to #86 on the relay and leave the one to the coil hooked up to the fuse box?

Also I am confused over the WN wire.

I have a WN wire which goes to the starter relay.

Did you mean for me to undo the WN wire from the starter rely and put that onto the fuse block where the ignition white came from and then run another WN wire to #87 on the relay?

Thanks

Bruce
Bruce Mills

Confusingly, on late model cars there is a white/brown from the ignition relay to the fusebox, as well as one from the starter relay to the solenoid. It must be the former FRM is referring to, but if you have white wires on the fusebox it sounds like you don't have the ignition relay, which would be pre-77, and I don't see the point of adding a second one.

If you want to *add* an ignition relay to power the coil you would have to split the two white wires at the fusebox and find out which one comes from the ignition switch and which goes to the coil, the one from the switch goes to 86, and from 87 would be a white/brown to the coil and fusebox.

However that is likely to cause the ignition to stay on as the relay will be held operated by the ignition warning lamp, a problem that afflicts 77 and on cars when there are problems with the emissions system. Better to *not* split the whites, put *both* of them on 86, and run a new one from 87 to the fusebox.

But all that is likely to do is bypass one of the bad connections that you obviously have, it will still be there and affecting the coil. The other one seems to be in the fusebox or green circuit and will still be affecting the speedo. Unless you use the new relay to power the speedo alone!
Paul Hunt 2

Thanks for answering Paul
I have added the new feed as FRM suggested.
The readings with fan, wipers, turn signals on at 2000RPM are
starter 14.05
the new brown feed to the relays 13.71
The welded brown connection 13.83
Brown wire at plastic connector to ignition 13.57

So these readings are within @.5 volts.
Is this close enough?
Assuming it is, then the next reading at the plastic connector which contains the wires which go to the ignition switch, the white is 12.82

As I look at the wiring diagram I realize that the brown wire feeds the ignition and in turn supply's power to the white wire.
If the white wire is this low at the plastic connector, could the ignition (steering lock) switch be the problem?

A quick look at the Moss book shows me it's going to cost 100-150cdn to replace the switch. Is there a way I can prove or disprove my theory.

Thanks


Bruce
Bruce Mills

Bruce-
Been away for a week.

"Do you mean for me to separate these white wires..."
You can, but it's not necessary. The W from switch is the control power for the relay; the W to coil can be left attached to it and the lot plugged onto the relay 86. The load from the coil is not an issue here, as it is small. What you are trying to accomplish is removing the rest of the W & G circuit loads from the switch and it's wiring.

"Also I am confused over the WN wire.
I have a WN wire which goes to the starter relay.
Did you mean for me to undo the WN wire from the starter rely and put that onto the fuse block where the ignition white came from and then run another WN wire to #87 on the relay?"
Yes. I mentioned use of WN here (for the relay output) solely to maintain the OE Lucas colors as used in the 77-on cars, and keep it "standard" and distinct from the W coming from the switch. WN is relay output, but it could be any color that is distinctive - not my fault that BL used WN twice and confused all of us!

Paul's comment on the "keeps running" trouble are noted - I forgot about that. My solution for it is to disconnect the W from the warning lamp, and use a WN from the relay output terminal 87 = fusebox (where the switched W used to be) to feed the lamp. That puts the lamp connection on the relay output rather than on the relay control.

In your latest figures, the 0.5 drop on the new relay N feed is more than I would expect - should be very near zero, about 0.02max. As I said somewhere, I have found that there seems to be an inherent limit to the amount of power that can be fed off the solenoid stack without a power drop of this order. That's why I recommended feeding the relays direct off The A+ alternator terminal.
The drops of 0.2 at the welded,and 0.5 at the IGN N feed both are a bit excessive, given that you have removed so much load from them. All pointing to something still wrong in that N circuit; it can be lived with functionally, but indicates a potential failure point.
The 1.2 drop across the switch and its immediate wires is not acceptable under any circumstance, now that you have removed all but the IGN coil load from that circuit. If all the wires and connections are correct, then it is the switch, and at that level means that the switch is sufficiently bad as to cause "dirty" power, which can give spikes everywhere in the car - including your electronic instruments.

In as much as you have never given a direct voltage drop measurement in all this discussion, I suspect you missed my points on that, given in Electrobabble and earlier in the rest of this. Measuring V drop across sections of circuit or single components is the direct way to find bad things, at any desired level of isolation. If drops are zero, then the other basic rules will be true; if not then you will know precisely where the problem is, without guesses or wasted work or substitution of components.

It is possible to dismantle the IGN switch, clean up the contacts, lubricate with OxGard, and put it back together.

If you do have to fix or replace the switch, the new IGN relay will mean that you should never have this problem again - the switch failures are caused by the inherent borderline overload in the non relay cars.

FRM
FR Millmore

Bruce - I have to say that I've been saying that your figures indicate a big volt-drop theough the ignition switch amd another through the white to green at the fusebox, all along, and recommending you test further in these areas. With the load on i.e. volt-drop at the speedo present measure the voltage on *both* sides of the brown connector in the multi-way plug by the steering column, on the brown and white terminals of the switch, and both sides of the white conenctor in the same multi-way plug. That will show whether any bad connections are in the plug/socket or the switch itself. Of course you could have more than one.

Then with the same load I'd test the spade connector on the white wire(s) at the fusebox, the fusebox spade they connect to, that end of the fuse *holder*, that end of the fuse, the other end of the fuse, its holder, that spade on the fusebox and the spade connectors on the wires (could be more than one) that are connected to it. the reason for testing both the fuse holder and the spade is that there is a rivetted connection between these two on the back of the fuseblock and these can corrode.

If you measured the 14.05v at the starter on the stud itself, then you are getting a reasonable voltage to that, but could still be getting a drop to the individual browns going up to the rest of the car's electrics, I've had this problem in the past. I'd use a probe like a needle to measure the voltage on each spade bolted to the stud and ensure they are all the same.
Paul Hunt 2

Another day in the shop chasing down this problem.
First off I ran a #10 brown wire from the Alternator + to the fuse box and to the relays, replacing the wire which I had run previously run from the solenoid.
The readings are as follows. 2000RPM with lights, heater fan, wipers and turn signals on.
Starter (End of stud) 14.12
Alt + to ground 14.25
Brown wire from alternator at fuse box 14.07. Brown wire on the other side of the fuse 13.99
Relay Brown wire cluster 13.97
This is the end of the new wire I just ran. I would have thought this would be closer to the alternator reading?

Ignition Brown wire multi connector 13.67 and 13.68 on the ignition side of the connector

I just do not understand why there is a .57 drop from the alternator to the brown wires to the multi connectors. These wires are continuous from the welded brown wire connector to the multi connectors. Where is the voltage going?? I have cleaned all connections?

Ignition White wire multi connector 13.03 and 13.07 on the ignition side of the connector
Brown wire right at the ignition switch 13.47
White Wire right at the ignition switch 13.28
I also put a probe on the white and the brown at the ignition switch and got a reading of .21. It is suppose to be zero isn't it?
White wire at fuse box 12.76

So the switch (as Paul has told me) is faulty. I took the switch apart and I can't see anything wrong with it. Only that the piece of metal inside is actually 2 parts. Is it suppose to be one?

I also tried to install the relay today. When I went to turn the key, the car started before it got to the ignition starting point and the starter would not dis-engage.
I have moved the 2 whites from the fuse box over to #86 on the new relay.
I put a brown wire on #30 and a black ground on #85.
I took the brown/white wire from the Lucas starter relay (leaving the red/white and brown wire on the Lucas starter relay) and placed it in the spot on the fuse box where the 2 white wires had been.
From the Brown/white wire on the fuse box I ran 2 more brown/white wires, one over to the new relay #87 and the other was going to be for the warning light. But as I turned the key on to activate the red light to help me locate the correct white wire the car started and wouldn't release the starter.
I have put the wires back to where they were originally to get the car started and home

Bruce

Bruce Mills

Bruce-
First off I ran a #10 brown wire from the Alternator + to the fuse box and to the relays, replacing the wire which I had run previously run from the solenoid.

>> There is no reason to not keep the N from the solenoid to fusebox in addition to the N from A+. It will function as an additional feed from A+ to B+, and reduce voltage drop, especially in Off-charge situations if such ever happen.<<

The readings are as follows. 2000RPM with lights, heater fan, wipers and turn signals on.
Starter (End of stud) 14.12
Alt + to ground 14.25
>>Do you have a ground from alt case to chassis? - the Jag mentioned in Electrobabble had over 1V drop from case to engine, all clean and tight. Fixing this by adding a grd caused the A+ to ground (chassis)voltage to drop, while bring the A+ to grd very close to B+ to grd. This is a result of the internal regulator seeing a different grd V than the rest of the car.<<

Brown wire from alternator at fuse box 14.07. Brown wire on the other side of the fuse 13.99
Relay Brown wire cluster 13.97
This is the end of the new wire I just ran. I would have thought this would be closer to the alternator reading?

>> You are pulling a fair amount of power, and there will always be a V drop, per Mr Ohm. I have the impression that the fusebox/relays is under dash, giving maybe 6-8 feet of wire. Not where I would put the box/relays by any means; I put the new box just behind the old one and the relays on the Rad support, giving minimal wire length - the A+ to box is maybe 16". These readings are OK under the circumstances.<<

Ignition Brown wire multi connector 13.67 and 13.68 on the ignition side of the connector

I just do not understand why there is a .57 drop from the alternator to the brown wires to the multi connectors. These wires are continuous from the welded brown wire connector to the multi connectors. Where is the voltage going?? I have cleaned all connections?

>>Agreed, there should be less loss. I cannot fathom why or how you moved the fusebox underdash and kept the 'same" configuration and wire colors. It would appear that there must be some splicing and wire extensions, but I don't know where or how. EVERY splice or other connection, and EVERY inch of wire gives a V drop. Are all splices soldered, and any added wire at least as big as the ones being extended? Remember that reducing the current draw on a circuit will also reduce the V drop on that circuit. Once you fit the IGN relay, the total draw on the IGN switch line will only be about 4A if it is still powering the coil, and much less if it's only operating the relay, and V drop should be negligible. Further, at this point you have removed nearly all the major loads from the OE (as modified) N wires, so the drop on those should be low.<<

Ignition White wire multi connector 13.03 and 13.07 on the ignition side of the connector

>>This is another .6V drop across the IGN switch and its immediate wires; while it indicates a problem of bad or undersized wires and possible bad switch, again the IGN relay will make it immaterial.<<

Brown wire right at the ignition switch 13.47
White Wire right at the ignition switch 13.28
I also put a probe on the white and the brown at the ignition switch and got a reading of .21. It is suppose to be zero isn't it?

>> In theory yes, but it never will be. .02 would be nice, and above comments on relay apply.<<

White wire at fuse box 12.76
>> Another excessive drop, .5V from switch. Same on relay.<<

So the switch (as Paul has told me) is faulty. I took the switch apart and I can't see anything wrong with it. Only that the piece of metal inside is actually 2 parts. Is it suppose to be one?
>>Haven't had one apart in years, some are like that but I don't think this one is supposed to be, should be obvious on examination with magnification.<<

I also tried to install the relay today. When I went to turn the key, the car started before it got to the ignition starting point and the starter would not dis-engage.
I have moved the 2 whites from the fuse box over to #86 on the new relay.
>>OK<<

I put a brown wire on #30 and a black ground on #85.
>>OK<<

I took the brown/white wire from the Lucas starter relay (leaving the red/white and brown wire on the Lucas starter relay) and placed it in the spot on the fuse box where the 2 white wires had been.
>>BIG mistake! This is why the starter engaged when the IGN switch was turned ON - you were powering the starter direct from your new IGN relay. This is the confusing NW Paul spoke of, leave it on the START relay; it is NOT the same as the other NW I spoke of, as used on later cars. The WR is the control wire from the IGN switch START position to the START relay; remove it from the relay and connect a test light from it to ground for testing. IF the START function is triggered, the lamp will light, without the stater engaging. Don't forget to replace it when you get this sorted!<<

From the Brown/white wire on the fuse box I ran 2 more brown/white wires, one over to the new relay #87 and the other was going to be for the warning light. But as I turned the key on to activate the red light to help me locate the correct white wire the car started and wouldn't release the starter.
>>OK here, BUT: It is important to describe things correctly- wires are main color first, tracer second - the start control wire is WR NOT RW (dash lights), and the relay to fusebox is WN NOT NW (used in generator circuits on older cars). Also describe circuits according to current flow - the WN wire goes FROM the relay TO the fusebox. Else we gets confuzzled.
The easiest way to sort the warning light W wire is to cut it right near the light and splice your new WN from the relay 87-fusebox to the lamp W side near the lamp. This wire splits off the rest of the W wires under the dash in various places depending on year, not accessible under the bonnet. Paul has a different method of preventing the no-shutoff issue, which I don't recall at the moment, but it's on his site.<<

>>Not to further confuse you, but there are a couple of things that might help your original problem, and possibly help in troubleshooting. There is no reason why you can't use more than one IGN relay - maybe a completely separate one to feed half the circuits. All IGN relays would be wired the same, with only the outputs different. In particular, you might want the electronic instruments on their own dedicated circuit, with a diode equipped relay or a varistor across both the control and load legs. Folk more electronic than I can tell you how to construct filter circuits to reduce noise on that system.<<

I have put the wires back to where they were originally to get the car started and home.
>> Have a brew - you're earning it!

FRM
FR Millmore

Bruce-
>> There is no reason to not keep the N from the solenoid to fusebox in addition to the N from A+. It will function as an additional feed from A+ to B+, and reduce voltage drop, especially in Off-charge situations if such ever happen.<<
I kept the wire from the solenoid. It is currently attached to the fuse box with out a fuse for future use.

>>Do you have a ground from alt case to chassis?
No I do not have a wire ground from the Alt to the chassis. I do not think I need one? The alternator is bolted directly to the block at two locations. The reading I have from a couple of weeks ago are Starter 14.02, Alt to Chassis 14.34, and Alt + to Alt case 14.41. .07 difference

The fuse box is under the dash. I am sorry I thought I had mentioned that. I have a Rover 3.5 and a few years ago I moved the fuse box and added relays under the dash http://www.britishv8.org/MG/BruceMills.htm
As you mentioned, perhaps not the best choice but I am sure you will understand why when you see the engine compartment. I don’t remember it being a hard task. I had a sacrificial MGB harness and used the same colour coding as the existing wires, soldered and heat shrunk all the splices. New bullets were soldered on. I added a relay for the low beam and another for the high beam and one more for the cooling fan. I drilled a hole up through the side transmission tunnel on the passengers side and ran the wires through that to the new relays and fuse box behind the glove box. .
The car is a 74.5 so I use 3 different wiring diagrams to make it all work. The Haynes 73/74 N. American models schematic and #17 73/74 and #18 75/76 from http://www.advanceautowire.com/mgb.pdf to make it all work. The WLG/W wire to the ignition coil was removed and a W wire installed with an internal ballasted resister coil.
The biggest problem I had when I did all this electrical work the car wouldn’t start. I had removed the sequential seat belt system and once I ran a new WR wire from the Lucas Starter relay to the ignition switch all was ok.


>>Haven't had one apart in years, some are like that but I don't think this one is supposed to be, should be obvious on examination with magnification.<<
I took a look at it through a magnifying glass but the supposed broken edges (right through he middle) look fairly smooth. I suspect it has been broken a long time and worked against itself. It makes no sence for it to be in 2 piece’s and there is no way it will go back together broken the way it is. I took a look at my electrical parts today and I have the switch from my old ignition switch/key assemble which was replaced about 15 years ago with the now faulty one. I will hopefully get that installed Monday evening and have some new readings (up to par hopefully) to post

I have moved the 2 whites from the fuse box over to #86 on the new relay.
>>OK<<
I put a brown wire on #30 and a black ground on #85.
>>OK<<

I found some information of relays
A General rule
Power is applied to #30 (in my case the brown wire)
Ground = #85 (black)
Receives power to operate relay #85 (white)
Output power #87 (this wire, colour arbitrary, goes back to the fuse box where the white use to be)

This, along with your instructions makes sence to me. The out put power (whatever color) goes to the fuse where the white use to be. From this same point the wire which goes to the alternator light should be attached? Correct?

>>Not to further confuse you, but there are a couple of things that might help your original problem, and possibly help in troubleshooting. There is no reason why you can't use more than one IGN relay - maybe a completely separate one to feed half the circuits. All IGN relays would be wired the same, with only the outputs different.
As I learn more about the system and in which direction the voltage flows and how relays operate I am sure I will put this new knowledge to use. Like daytime running lights. But that is a whole different/new issue. One problem at a time

>> Have a brew - you're earning it!
:) Thanks

Bruce
Bruce Mills

Bruce-
"...currently attached to the fuse box with out a fuse for future use. "
>>So it's not presently helping carry the load. While the 10ga feed is rated at 35A, this depends on length, and will give a drop if long. In addition, bullet connectors and 1/4" Lucar (flat tab)connectors are rated for 17.5A continuous/35 max short term - because of drop and consequent heating. This applies in any circuit. Main feeds are preferably run through bolted connections, or sometimes 3/8" or 1/2" Lucars. Certainly 100A alternators and the loads that require them call for bolted main connections.
>> If you are feeding all the power from N leads through a fuse going into the box, similar rules apply, around 30A max through standard glass or plug-in fuse holders. My Mazda has main feed fuses - BIG plug-ins with big blades.

"...do not have a wire ground from the Alt to the chassis"
>> I detail the story of the Jag alt in Electrobabble. I've found this a common situation, though not to that level, 0.5V is more common.

Yes, you mentioned moving the fusebox, but I missed the bit about the V8! The MGB wires are mostly adequate (except headlamp pigtails!), at the factory sizes and lengths. Extending them at the same size may not be sufficient, and increased loads make them almost certainly undersized to the point of explaining your persistent V losses. Also, for common wire colors (W,G,N), there are several sizes in the harness; using a section of undersized wire WILL give major drops on a heavy circuit. Checking Vdrop over sections of wire will show you where the problems are.

"...Ground = #85 (black) >> Right
Receives power to operate relay #85 (white) >>ERROR! should be 86.
This error may be in Electrobabble also, depending on where I was in the revision process. It does not matter if 85 & 86 are reversed, UNLESS the relay is diode equipped. In that case, reversing the control leads causes the diode to fail short, giving a 70A drain to ground (when I did it recently!). Best to keep these consistent.

" From this same point the wire which goes to the alternator light should be attached? Correct?"
>>Yes.

A point that is easy to loose: All this searching for Voltage losses can consume you. Automotive systems are designed to allow for "reasonable" loss. Lights for example, are typically rated for illumination levels at 12.8V, and for lifetime at 14V; same idea for everything else. Lights have a +/- 15% output though, with large variations for small V changes; and marginal items like MGB heater blowers need all the help possible. Thus, it is desirable for at least these things to have delivered V as close to system max as possible. The most important thing is that you know where and why you have a drop; then you can decide if you need or want to live with it. Your evidently busted switch was letting the car operate for some time, but it was a possible "leave you stranded" event, which you can be sure would have happened in the most inopportune situation!

DRL can get really confusing. Daniel Stern has a neat module that he says makes it easy, uses the turn signal lamps as DRLs. I haven't used it yet.

Have another brew for me - I'm needing it !

FRM
FR Millmore

A quick report.
I installed the new ignition switch and took new readings. Nothing really changed enough to bother posting here. A potential road side problem has been eliminated.
I also got the relay installed. This solved the speedo problem. HOORAY. The gauges no longer dip when the brakes, wipers, turn signals and heater fan are applied.
A huge THANK YOU to Paul and FRM. It is time for a beer.
I will straighten out the wiring mess I have under the dash and take some reading tonight and report back.
You guys have no idea how good I feel about getting to this point.

Thanks very much again

I did blow a 20A fuse This is the fuse that the relays run off. Hi beam, low bean, cooling fan and now all the accessories. I will put a 30A in there and see how I get on

Bruce
Bruce Mills

Bruce-
Great!
A couple of points.
"...fuse that the relays run off. Hi beam, low bean, cooling fan and now all the accessories."
>> Not good practice at all! Each relay should have its own power feed fuse. A pair of 100W headlamps will pull about 15.6A at rated 12.8V, 17A at 14V - but they won't be getting 14V because the fuseholder is overloaded once you add the other loads you are feeding. Low beams around 75W, typical for 100W highs, will pull 11.7/12.8A. You can have all beams on if you are on low and you flash the highs = 30A draw. Then add the other loads.

You also do not want everything going out at once because of a fault or miscalculation on your part. When the high beams die, you will hopefully have the presence of mind to switch to low before you become an off-roader - would be nice if the lows work!

Go back to what I said about the six circuit box with single stud feed and feed your relays separately! You want minimum 20A fuses in the high and low beam feeds, probably 10 or 15A in the other feeds. Remember that inrush currents on cold parts can be much higher than steady state currents. My rule of thumb is to use fuses rated at 1.5 or 2x the measured max draw on the protected circuit.

Comparing voltage readings taken at different times can seriously mislead you. Both the system V and local section V changes greatly with load and (component) temperature conditions, more so if there are bad connections. It's another reason to measure local drops rather than hot to ground measurements when troubleshooting. Do your final checks with the system hot and loaded for 15 or 20 minutes, and be sure the battery is fully charged. I will send you direct a copy of an extensive MGB rework I just did, which will show you what happens pretty well.

FRM.
FR Millmore

FRM

For the moment I can separate the relays to some degree and use the new unused N wire to the solenoid to run the accessories.
But I like the idea of "single stud feed and feed your relays separately!".
Since I have spare wires in the harness that have been eliminated for one reason or another and I have new wires to tape into the harness.
I will be looking into upgrading my electrical system over the winter.
I will seperate my relays and take some new readings and post the results.

Bruce
Bruce Mills

Hi

I separated the relays yesterday and put the hi & low beam headlight (just the 100/75W not the auxiliaries) relay and the cooling fan relay on the new feed from the alternator. So that equates to 17A for the hi beam and 15A for fan. Just slightly over 30A. I think it would be better to separate them but will get to that later.

The new feed from the solenoid is carrying the accessories. The fuel pump draws 4.5A and the wiper motor 2.7-3.4, I am not sure about the turn signals nor the heater fan, stereo and amplifier, dash lights etc

I understand not to put a lot of importance on the comparison of different set of readings but on the individual set itself. However you can see from below that things have improved substantially. Not as drastically in the N circuit as in the W and G circuits.

All figures are under load 2000RPM with wipers/lights/turn signals and heater fan on.
Aug 25 was prior to installing the relay.
Aug 27 was after installing the relay and
Sep 01 was after separating the relays to different feeds

Aug 25 Aug 27 Sep 01
Starter 14.12 14.16 14.23
Alt to Ground 14.25 14.27 14.40
Alt to Case 14.27 14.46
*Relay N Cluster 13.97 14.00 14.14
Welded N Cluster 13.99 14.05 13.84
N at Ignition Switch 13.47 13.53 13.72
W at Ignition Switch 13.28 13.26 13.63
Across Ignition NW wires .21 .18 .12
White at Fuse 12.76 12.96 13.82
Green at Fuse 12.70 12.82 13.80
New Feed From Alt at Fuse 14.10 14.01

*On Sep 01/07 this is now the hi & low beam relays as well as the cooling fan relay and is feed directly from the alternator instead of the welded N cluster.

These figures are, hopefully, acceptable. I can see that I am loosing some voltage at the ignition switch from the N and sequentially to the W & G circuits. I do not understand why (after all it should be 1 solid wire from the welded N to the ignition) and this winter will take a much closer look at it.

As I accept these figure for the moment and will now try and get a lot of driving in the last month of the driving season, my mind is turning to something else you mentioned. “six circuit box with single stud feed and feed your relays separately”

I have been unable to located a picture of a six circuit box with single stud feed on the internet for clarification.
My understanding is that I would feed the six circuit box from the alternator (1 wire) and thus to the relays? If this is the case I could use the new feed I have already run from the alternator to feed the six circuit box?


With the alternator putting out @14.4 volts feeding a six circuit box which feeds my existing relays (and future relays), would this be enough voltage? Or would I have to run additional wires from the alternator to the individual relays to ensure they are receiving max voltage.

For anyone who has read this far through both lengthy threads, I have tried to be as detailed as possible (from my layman’s point of view) in the hopes that it will assist someone in the future in dealing with a perplexing electrical problem. My advice is to buy a multi tester and practice using it. Don’t be afraid to post questions, That is how we learn. And believe me, I have learned a lot from this experience

FRM, Paul and Kimberly have all been extremely helpful, patient and more than generous in sharing their knowledge and time

Thanks all.


Bruce

Bruce Mills

Bruce-
Your present readings are just dandy. Anything more takes really big wires and space shuttle work - totally unnecessary.

The six circuit fusebox is Buss # BP/15600-06-20, takes six ATC fuses, from any standard parts store - mine from Advanced Auto Parts = $8. Has a single 10-32 stud feeding all six fuses, which have standard 1/4" push-on tabs for output.

Run your feed from A+ to the stud; feed each relay off a fuse output. Your 10 ga feed may be a little undersized for all those loads, IF you have all that stuff ON at once, and IF the feed is long because you have the relays under the dash where they should not be. Wires from relays to heavy loads should be as short as possible, and amply sized - 12 ga for cooling fan and your overwattage headlamps.

You can attach all N wires to the stud as well, to allow them to share loads in parallel from A+ to solenoid.

Once you get the relays fed correctly, You are about ready to measure final V across light filaments and fan terminals!

Note that one factor in doing this is that as you improve connections and reduce resistance, current draw goes up with delivered voltage, making remaining marginal items show greater V drops.

FRM
FR Millmore

Much better, but what does the speedo do?
Paul Hunt 2

Paul-
See Bruce post 28 Aug above!

FRM
FR Millmore

Fairy nuff.
Paul Hunt 2

This thread was discussed between 07/08/2007 and 04/09/2007

MG MGB Technical index

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