Stray Current ARGHHH!!!

[love ke70]

Hey Guys,

So the pootrol ate a all aluminium radiator in three years.
diagnosis was stray current.
trying to find the source, I had all sorts of strange things going on with it last week and advice was get a analogue mulitmeter, and flush the system so you have water only in there.

OK, thats done.

So I went to it this morning, bearing in mind I filled and drove and drained the cooling system about 10 times at least last week with tap water to get it as clean as I could, and it is clean water in there now and I havnt driven it since then.
went to it and with everything off including ignition, it was reading 0.5 V
FUCK I thought to myself...
so I started unplugging things from the battery 1 by 1 until there was nothing left.
still 0.5V (depending how far you dip the probe into the coolant)

So I disconnected the USI-160 battery controller, no change, and started disconnecting things from both back batteries until they were isolated completely, IE no earth or power connected to them. Still no change.

Then I thought, ahhhh of course, the laminova intercooler has coolant mix in it, that must be storing a current blah blah, checked that and it had -ve voltage of about 0.2V
FUCK I thought, so I have no drained that and flushed it with water and then blown the lines dry so there is nothing in it, to eliminate variables.

Feeling chuffed that I had got it sorted and there would be no stray current now as all 3 batteries do not have their +ve terminal connected and only the cranking battery has its earth on, I check again.
0.5V STILL!

So I am stumped.

I ran some water into a bucket and tested that, 0.2V with both probes in the water. :shock:


No I am a little confused so I have supplied photos...
but basically the digital multimeter set on 2V (cause on 200mV it maxes out) reads 0.5V
the analogue one read higher earlier (5) but as in the photo is reading around 2.4 and as per the chart it says multiply by .01 (2.4*.01 = 0.024) photo doesnt show it, but yes it is in 0.1V setting...

SOOOO, is the digital or the analogue gauge right?

photos for reference:


The test piece:





the reading on the digital meter as shown in the arrangement above




and the analogue reading (pulled probs from digital and plugged to analogue)



reference table







Reading in the bucket (I know the analogue one is a bit hard to read cause its off centre, but its ~0.2) so 0.2 * 0.01 = .002




Why is the digital and analogue so different?!




And more to the point, why does my water have a voltage?!

the tap water and the water in my coolant....

Just for reference, on 20v setting on the digital gauge it reads 12.01V on the battery and 0.5 something in the coolant, so that scale seems to be correct...

blew the ass out of the analogue one by sticking it on the battery in the wrong setting...oops, might be taking that back for a swap over

any and all help appreciated guys!



Cheers, Andy

[dumbass]

its the bloody water
there is so much shit in it that not earthing the radiator or earthing the radiator does jack shit
goto your local woollies and buy some demineralised water which is mainly used for irons
so they don't rust

ive been thru this many times with many customers and this water normally fixes the problem
and using a digital multimeter should be fine aswell
most shops like pwr don't know how to cure these types of problems
many years ago they used to say don't earth radiator now they say earth it
they wont give u a straight answer because they just want to sell more radiators and not have to give warranty on them

[trdee]

Galvanic corrosion. Run demin water. Also make sure your engine grounds are all in good condition

[love ke70]

I wont be earthing the Rad
I do run demin water with the coolant, but I was told to run tap water for the testing? Can swap it out if I have to, but its actually really hard to get all the water out of the block

apparently the coolant will hold the charge so its impossible to problem solve with coolant in there, hence the tap water. I was going to clean all earths etc, but got stumped by these strange findings.

Its just really baffling me that there is so much voltage in the water just sitting in the engine.

Do I treat the half volt as a datum and work to it, as its hard to aim for 0 when 0 cant be achieved completely isolated?!

Cheers, Andy

[foamy]

It's pretty hard to hit zero (our newish work Corolla reads 0.01 of a volt. Anything above 0.03 can be classed as damaging depending on who you talk to). Mixing different coolants will even give you a reading, might be a silly question but when you flushed everything did you do the heater core as well?

Out of interest, who knocked you back on warranty? Was it a tube to header leak, or further along in one of the tubes? Did they give you any photos or reasons as to why the thought it was electrolysis? (If they're lazy and it's a tube to header leak, they might not of bothered cutting it open and just knocked you back because it's easier).

Another potentially silly question, with your positive lead, is it just touching the coolant only, or touching the radiator as well? It only needs to touch coolant, if it's on the radiator then it's going to mess with the reading...

[trdee]

do you have any copper/brass/bronze items anywhere in your cooling system?

http://en.wikipedia.org/wiki/Galvanic_corrosion <- go down to the anodic index

if you have copper and aluminium in an electrically conductive environment, you can see voltage differences of up to 0.5V, basically creating a battery... and as aluminium is more electronegative, it will be the metal that corrodes. aluminium and iron dont have massive differences in potential so they shouldnt corrode much.

[Indelible]

just an idea and might be completely off but maybe look into something like this?

http://www.flex-a-lite.com/accessori...tor-anode.html

[foamy]

Surely that couldn't cause an issue Vito? 80% of the cars on the roads have a radiator with an aluminium core and a brass oil cooler?

[oldcorollas]

the water doesn't have a voltage. the coolant doesn't store current

what are the tips of your probes like? if they are plated, but one is a bti scratched with copper plating exposed or whatever, that can affect measured voltage

get some copper wire, wrap around probe ends and have a few cm sticking out.. scrape a bit of the copper wire clean, then stick ONLY the copper into the water (NOT the probe metal).
having same metal only on both probes should have same reactions at the metal surface, and should have no voltage between them.

stray current and stray voltage are not the same thing. you can have a large voltage difference, but no corrosion if there is no current (flow of electrons)..


what sort of alloy radiator was it? different aluminium alloys (ie, that used for the main cooling bits, and that used for the caps or tubes etc) have different impurities and other stuff that can change their electrochemistry, and when they are welded up into a part like a radiator, it can be destined to fail..

still need ions in the coolant to make it bad.. and stuff like chlorides can be far far worse for corrosion than other things like sulfates.. pH also can have a big effect, pushing an alloy from self passivating, to actively corroding...

for the multimeters, got a 1.5V battery? connect both multimeters up in parallel to the battery, see what they read?

[oldcorollas]

in your second picture, what's the white stuff adjacent to the weld?
to me that looks like it could be corrosion of the flat bits, galvanically driven by the weld metal

where did the radiator fail? got pics of the failure points?


edit:
if radiator is dead, you could cut bits of the different aluminium bits, and use them as electrodes in a water mixed with a bit of salt (or tap water, or your coolant mix).. so you can see the voltage generated between different bits.
it might take a couple of minutes, or half an hour or more to reach a stable voltage, or longer...

[trdee]

Surely that couldn't cause an issue Vito? 80% of the cars on the roads have a radiator with an aluminium core and a brass oil cooler?

It most certainly could. Especially if the water is conductive. Batteries work by the same method. Sacrificial anodes on boats work by the same principle too

edit:

just an idea and might be completely off but maybe look into something like this?

http://www.flex-a-lite.com/accessori...tor-anode.html

that is a sacrificial anode