Bad chassis earth?

[BlackDogZulu]

I've had the truck since May but it's only recently I have been driving with headlights a lot. Hence I didn't notice I had a problem until this week. 1997 90, 300Tdi, unmodified. Symptoms:

• Temp gauge goes from below half (normal) to the red when headlights turned on (only when engine is warm - doesn't seem to do it when cold)
• Headlights very dim, like a brown glow, both dip and main beams
• When accelerating up the road in the dark (no street lighting) I can see the lights getting slightly brighter and dimmer - seems to change with engine revs. but in 'steps'. (My old Mondeo used to do this as well, and I just put it down to 'one of those things'.)

Someone has suggested that I need to run a new earth from battery to chassis and engine/gearbox as I have a bad earth connection. This seems reasonable to me, and I am going to have a look at this before I start looking at the wiring generally. Does this sound like a likely solution to the problem - at least as a first port of call?

If so, I have been recommended to get 2 x 18" earth straps - one from battery to chassis, one from chassis to engine/box. Would that be long enough? Just give me an idea - it's raining outside and the Landy is standing in a huge puddle of water. Not the day for crawling underneath!

Thanks all.

 

[neilly]

Hi There,

I had similar issues when I got my 90.

One way to test it , is to run a jumper cable from the chassis to battery, chassis to engine , chassis to gearbox, fit one of them take it for a run and see if that helps.

Plus remove what earths you have, clean the ends and where they bolt to , retighten them.

Not sure on lengths.

But, most "car" places have a lot of different lengths.


Cheers

 

[Oldseadog]

Check the output from the alternator at the battery, which can be done from the nice, dry inside.
It should be between 13.8V and 14.4V, and whilst you are at it, check to see if there is any 'leakage between the battery and the frame earth.
Start simple and work outwards friom there.

 

[BlackDogZulu]

Hmm. Checked the alternator when I got the truck and it was fine, but haven't looked at it since. How do I check for 'leakage between the battery and frame earth'? Not sure what you mean there. Off outside with the multi-tester. Thanks.

 

[lynall]

You simply cannot have to many earth straps.
My 200 has one direct from battery to the chassis, another direct from battery to the engine and one engine to chassis.

 

[Disco1BFG]

OP:- any electrical weirdness of this nature is often a bad / failed / failing return path. Earth straps are cheap, and I'd make sure you have one from gearbox to chassis, one from engine to chassis, and one from battery to chassis. @lynall is not wrong - too many earth straps will not cause problems.

IF you have a handbrake cable, rather than a linkage, then be careful that the cable doesn't become the only "earth strap" - as the heat will weaken it, possibly causing it to fail, with results that you do not want....

You may have other problems, but making sure everything is earthed properly is a good place to start.... Then the "boomslang" loom addition to your headlight circuit is worth doing, as this will save the crappy switch, and increase the brightness of your lights too.

 

[BlackDogZulu]

Alternator is putting out 13.9V at idle, falling to 13.7V with headlights and HRW on. Curious thing is if I increase the revs to a fast idle (2000-ish) the alt output falls slightly to 13.8V no load and 13.6V with lights and HRW on. When I have checked vehicles in the past, the voltage usually increases slightly at fast idle. Could this be a voltage regulator problem? Possibly related: when I leave home I drive up a long unlit lane. With the headlights on main beam, they appear to brighten and dim slightly as the revs rise and fall, but in 'steps', not smoothly. My last car did this, and I assumed it was the way the voltage regulator worked, but maybe not.

 

[BlackDogZulu]

You simply cannot have to many earth straps.
My 200 has one direct from battery to the chassis, another direct from battery to the engine and one engine to chassis.

I have heard this before. Makes perfect sense. As soon as the driveway isn't under water, I'm on it.

Then the "boomslang" loom addition to your headlight circuit is worth doing, as this will save the crappy switch, and increase the brightness of your lights too.

This is a new one for me. I'll have a look into that too. The headlights are pretty ineffective, much worse than I remember on my n/a 90.

Thanks both.

 

[Kwakerman]

Simple checks first, start the engine and with a multimeter measure the voltage from the engine block to the battery negative terminal, should be less that 20millivolts or so, now turn the lights on and measure again. You will probably find it has increased to 300+millivolts, it is this voltage difference between the 'ground' on the temp sender (the engine block) and the chassis ground which causes the temp gauge to rise when the lights are on. If it does rise then that is indicative of a bad main earth cable. Fixing that will help the lights a bit but reality is that the lighting wiring isn't that good to start with and a boomslang loom, orsam night breaker bulbs and new clear 'crystal' light/lenses will make a major difference.

 

[pete1987RRC]

Your alternator output is a little low but start with the Earth straps as all have said above.
They have a horrible knack of coroding under the pvc outer, they are cheap so replace the lot.
Vehicle wiring products will make them up if you can’t do your own.

Then give the battery terminals a good clean and gell them up.


If your still at 13.8v check the fan belt is tight, if still low the voltage reg is an issue.

 

[BlackDogZulu]

Simple checks first, start the engine and with a multimeter measure the voltage from the engine block to the battery negative terminal, should be less that 20millivolts or so, now turn the lights on and measure again. You will probably find it has increased to 300+millivolts, it is this voltage difference between the 'ground' on the temp sender (the engine block) and the chassis ground which causes the temp gauge to rise when the lights are on. If it does rise then that is indicative of a bad main earth cable. Fixing that will help the lights a bit but reality is that the lighting wiring isn't that good to start with and a boomslang loom, orsam night breaker bulbs and new clear 'crystal' light/lenses will make a major difference.

That's a very helpful reply, thank you. I'm OK with using a voltmeter for continuity checks and reading raw voltages, but reading voltage differences is new to me, However, what you say makes sense and I will give it a go today. The headlight units are Wipac with clear lenses and a rippled reflector, and look fairly new. I am guessing that the bulbs are normal halogens. I'm a big fan of Nightbreakers (used them a lot before) but to be honest a 50% increase on the current light output would be hardly worth doing at the price. There is obviously something wrong that I need to fix first. I haven't checked the voltages at the bulbs, but I am guessing it will be less than 13.8. Looking into the new loom.

Your alternator output is a little low but start with the Earth straps as all have said above.
They have a horrible knack of coroding under the pvc outer, they are cheap so replace the lot.
Vehicle wiring products will make them up if you can’t do your own.

Then give the battery terminals a good clean and gell them up.


If your still at 13.8v check the fan belt is tight, if still low the voltage reg is an issue.

Plan is to replace all the earth straps and add if necessary - battery/chassis, engine/chassis, battery/engine. I'm a fan of belt and braces. I'll get on to VWP as soon as I have measurements.

Fan belt is tight, but perhaps the voltage reg is faulty, then? I have never seen the voltage go *down* with increased revs before. I have an autoelectrics place nearby, so I can get it tested there if need be.

Guys, I am back on shift tomorrow but will get some free time later next week when I will be sorting this issue out. Thanks for all the advice - much appreciated.

 

[Kwakerman]

I'm OK with using a voltmeter for continuity checks and reading raw voltages, but reading voltage differences is new to me

Its exactly the same thing, negative probe on the battery -ve terminal and the +ve probe on the engine block and measure the voltage, you may need some additional wire to make the probes reach.

 

[BlackDogZulu]

Its exactly the same thing, negative probe on the battery -ve terminal and the +ve probe on the engine block and measure the voltage, you may need some additional wire to make the probes reach.

Thanks for this. So - if there is perfect continuity between battery/chassis/engine block, a voltmeter between them would read zero. Any voltage (measured in millivolts) would indicate a bad connection. Have I got that right?

 

[Discokids]

The voltmeter will read zero regardless of how good/bad a connection you have because nowhere you mention are there volts present to begin with. It's resistance you need to be measuring, between all the places mentioned. I think what kwakerman is saying is measure your battery voltage, then measure your voltage again with pos lead on pos terminal, but by putting your neg lead to block, g/box, Chassis, body etc. And see if there's a difference. Also try keeping neg lead on neg terminal and try pos lead on alternator output, starter flag and see again if Theres a difference.

 

[Kwakerman]

The voltmeter will read zero regardless of how good/bad a connection you have because nowhere you mention are there volts present to begin with. It's resistance you need to be measuring, between all the places mentioned. I think what kwakerman is saying is measure your battery voltage, then measure your voltage again with pos lead on pos terminal, but by putting your neg lead to block, g/box, Chassis, body etc. And see if there's a difference. Also try keeping neg lead on neg terminal and try pos lead on alternator output, starter flag and see again if Theres a difference.

Sorry Discokids, got to disagree there, the best way to check for a poor connection (or less than perfect ground connection) with a 'standard multimeter' on large cables is to measure the voltage drop across the cable / connection. Obviously you need to have a current flowing through that connection for a voltage drop to be developed (ohms law V = I x R). The reason this method is better (albeit more fiddly / time consuming) is that a good quality generic DMM may be able to measure resistance acuratley down to say 0.1Ω but will also have a resolution of 0.1Ω so could actually read 0, 0.1 or 0.2Ω If you use the voltage method then the DMM should be good for something like 0.01 or 0.001 volts accuracy & resolution. If we use the 0.1Ω example above and say the DMM reads somewhere between 0 and 0.2 then that could be considered to be a good connection, however, if it were the headlamps earth connection then it could be passing about 10 Amps, that would equate to roughly a 1 volt drop over the connection when the lights are on which is well within the DMMs voltage tolerance but not good for the lighting. If on the other hand the voltage drop was 0.5 volts @ 10amps then we would know the connection resistance is 0.05Ω alternatively 0.25 volts would equate to 0.025Ω both of which are way beyond the meters resolution on resistance but well within its voltage capability. Hope that makes sense?

In the case of the temp gauge miss-reading high with the lights on this is normally because the engine block (and therefore the temp senders ground connection) is at a slightly higher voltage than the temp gauges earth reference. The engine is normally a bit higher as the charging current from the alternator (which is grounded via the engine block) increases when the lights are on, as this current also flows through the battery straps any increase in current will also induce an increased voltage drop over any poor connection which makes the engine 'ground' float higher than the chassis ground.
From experience I've noticed that normally (engine running, lights off) I get about a 30mv (IIRC) difference between the engine block and battery -ve terminal. However I had a similar issue to the Op where my VDO temp gauge would rise 10 degrees with the lights on and drop with them off, in that case the engine block ground voltage went up to 300mv with respect to the battery -ve. As the engine ground voltage increases it also affects the temperature senders output and therefore the gauge readings are affected. In my case it was a poor connection between the battery strap and chassis rail, a good clean up sorted it out in the end.

Phew!

 

[Kwakerman]

Thanks for this. So - if there is perfect continuity between battery/chassis/engine block, a voltmeter between them would read zero. Any voltage (measured in millivolts) would indicate a bad connection. Have I got that right?

Set you DMM to voltage, connect the black lead to the battery -ve terminal and the red lead to a good clean point on the engine block. Start the engine (with lights off) and measure the voltage (should be in the 0.03v region), now turn the lights on, is there a fairly large increase in voltage (0.25 - 0.5)? If there is then you have a poor earth strap or earth strap connection somewhere. You can use this technique to measure the voltage across each connection point and also over the length of the earth cable (to check its condition). No (or minimal) change between light on & off means good, large change = bad connection or cable.

 

[Kwakerman]

The voltmeter will read zero regardless of how good/bad a connection you have because nowhere you mention are there volts present to begin with. It's resistance you need to be measuring, between all the places mentioned. I think what kwakerman is saying is measure your battery voltage, then measure your voltage again with pos lead on pos terminal, but by putting your neg lead to block, g/box, Chassis, body etc. And see if there's a difference. Also try keeping neg lead on neg terminal and try pos lead on alternator output, starter flag and see again if Theres a difference.

Discokids, having re-read my reply to you reply I've realised it sounds a bit argumentative (the 'got to disagree with you' bit), obviously you are just trying to help the Op too, sorry if any offence was taken! I'll get back to painting my skirting boards now (after a bacon sarnie)

 

[Discokids]

Discokids, having re-read my reply to you reply I've realised it sounds a bit argumentative (the 'got to disagree with you' bit), obviously you are just trying to help the Op too, sorry if any offence was taken! I'll get back to painting my skirting boards now (after a bacon sarnie)

No problem kwakerman, no offence taken We're all only trying to learn and help as we go. I just read the post prior to my own reply and couldn't understand where said voltage was coming from, because no battery was mentioned. I didn't get why bad connections would present +voltage, when there was no voltage present to begin with.
Basically, I read the ops post as though he were looking for voltage on something that would never naturally contain it without something supplying it first. I.e., a piece of cable that isn't connected to a battery. It was too early in the day for me to get involved, and your post more than has it covered.
I'll step back, and hope the op disregards my ramblings, as while I understand this now in my own mind, I can see I've more than likely just added to the confusion

 

[Discokids]

Discokids, having re-read my reply to you reply I've realised it sounds a bit argumentative (the 'got to disagree with you' bit), obviously you are just trying to help the Op too, sorry if any offence was taken! I'll get back to painting my skirting boards now (after a bacon sarnie)

And a fried egg with brown sauce would go very nicely with that bacon

 

[Kwakerman]

And a fried egg with brown sauce would go very nicely with that bacon

Egg yes (but too late now!), brown sauce.......no way!