Multimeter for testing engine coolant

[Oro_O]

I found a few references online for using your mulitmeter in DC V mode to test the health of your coolant. One was a six year old thread from BITOG.

thread

I have one vehicle that sees little use, and when driven, is driven moderately (14k miles in the 3.5 years since I last changed any coolant, when I did the WP and timing belt as well as both radiator hoses). Since we live in a very mild climate (it rarely sees temps over 80F or under 40F), I speculated the coolant was not being stressed and didn't need a change despite the time. It has has no make-up coolant in that time and the level in the overflow tank is also unchanged. It is regular green in the system. I ran the multimeter test and got ~70mV (.07V as in the guidelines).

So, my instinct is to just monitor this value before the summer starts each year and then drain/fill the radiator, barring any visible changes, when this value starts to climb towards the threshold 0.4V. I was considering getting test strips and then read and learned about this method, which was news to me.

It's not that I am maintenance-shy; I do all my own and enjoy it. But I have less free time for that and higher priority little tasks on the cars I'd like to do first. Plus throwing out perfectly good fluid doesn't seem bright if a simple test is available.

Q: Does anyone else use this method or have experience with it?

 

[Chris142]

Almost daily. Its been my experiance that anything over .3v will eat radiators

 

[Ducked]

Some online references (there are also youtube videos but I didn’t bother with them, a word being worth a thousand redneck-illiterate pictures)

http://www.familyhandyman.com/automotive/car-maintenance/coolant-testing-with-a-multimeter/view-all

This one wants you to test it running (2000 rpm) at normal operating temp, (impossible? It'll boil)

< 0.4 Volts OK > 0.4 Volts BAD (which is plainly an arbitary cutoff)

http://www.sancarlosradiator.com/VoltageDrop/testing_for_electrolysis.htm

- lead to ground + in the coolant (rad)
- - >0.3V BAD
- > 0.3 V with battery disconnected = “Type A”(self sustaining) electrolysis.
-
- FLUSH and change. A range of flushing techniques/chemicals is discussed/pitched
-
- 0 volts with battery disconnected = “Type B” (earth leakage) electrolysis due to a bad ground/current leakage to earth – locate and fix.

It seems that the engine is OFF for these tests, though it isn’t explicitly stated.

http://www.ve-labs.net/electrolysis-101/how-to-test

Similar numbers, and they want the engine at 2000 rpm testing for current leakage, but don’t say it has to be at temperature.

They suggest using a bit of copper tube on the + electrode in the coolant.

They also suggest testing the alternator for overcharge (>14.7 volts), and using pH strips, specific test strips, a hydrometer or a refractometer, and pitch a range of chemicals and gadgets.

I had/have a refractometer in the UK, though it was designed for sucrose solutions so I dunno if it’d work with coolant.

http://www.experts-exchange.com/Other/Automotive/Q_27776180.html

Conflicting numbers

https://macsworldwide.wordpress.com/2011/10/03/acdelcos-guidelines-for-replacing-engine-coolant/

ACDelco say
http://www.audizine.com/forum/showthread.php/263790-Antifreeze-Experiment-amp-Testing

Forced electrolysis experiment, essentially testing the performance of coolant mixtures as electrolyte – not sure how relevant this is to real-world in-car situation.

Testing on the Skywing, voltage was initially about 0.14 volts but fell quite quickly to about 0.09V, and continued to fall more slowly thereafter. Subsequent re-tests started about 0.09V and fell (will have to look again to establish a rate). Ignition/fan/lights on didn’t seem to make any difference.

I didn’t try running the engine.

This didn’t give particular ground for concern, (though the instability of the voltage reading isn’t mentioned in any of the refs.) but I wasn't very convinced and changed it anyway since it was about 5 years old at the time and I didn't know what was in it.

It was quite rusty, as it was when I changed it again just recently.

 

[Y_K]

Thank you for a bunch of useful links. Just reinforced my stay-with-the-manual attitude in this department. In my 'fleet' I have a G-05, Toyota SLLC (pink) and an old school green vehicles. All those already have and keep putting a lot of miles/hours. I simply drain and fill radiators every April and October regardless.

 

[Ducked]

Originally Posted By: Y_K
Thank you for a bunch of useful links. Just reinforced my stay-with-the-manual attitude in this department. In my 'fleet' I have a G-05, Toyota SLLC (pink) and an old school green vehicles. All those already have and keep putting a lot of miles/hours. I simply drain and fill radiators every April and October regardless.


Some of the links are probably dead. It's a while since I looked into it.

I think I wouldn't rely on this method alone. In combo with pH, hydrometer and/or refractometer it might be a useful check but I probably still wouldn't again over stretch the change interval beyond a couple of years with the "standard" green stuff.

Well, not deliberately anyway.

 

[Silk]

I've been checking coolant this way for years, and you can also use the method for brake fluid. I use 600mv for coolant and 300mv for brake fluid.

 

[i_hate_autofraud]

Sorry, but the voltmeter test for coolant is false.

It turns out the coolant mix ratio can make the meter swing the most, that is the mix ratio is the dominant factor,
not if the coolant is worn out, hot/ cold or anything else. (use test coolant with litmus test strips - pH)

Since the coolant is in contact with different metals in the engine and the radiator and connected
to the battery thru the ground and wires to sensors that also make contact, a small voltage is generated.

You can do it in a coffee cup; add 2 part water / add 1 part coolant / mix / add steel / add aluminum strip / measure
voltage, then add more 1 more part coolant - watch voltage change!

you can use different metal strips and see voltage change.

If you use a meter with a needle movement, the voltage will be lower, (low impedance of meter uses tiny current to move needle)
if you use a digital meter with a 1 MegOhm input, voltage higher, if 10 MegOhm input, higher still, etc.

Pat Goss' Garage on Motorweek also got fooled by meter test too:

See PDF with Pics and see:

BOGUS COOLANT TEST – Pat Goss Blows It!
https://app.box.com/s/mc3kaetdmj38ebzomyscxe9wt8rvotdg

 

[i_hate_autofraud]

Forgot to mention, actual corrosion due to a 'voltage' in the coolant, etc, requires
significant current. While taking voltage readings I also switched to current mode and
saw currents in the micro-amp range.

In the garage on the daily driver, coolant voltage was .4 V and current was 12 micro amps and dropped over several minutes. Considering the size of the engine block, radiator and total immersed metal surface area, 12uA is nothing.

On the other hand, old coolant with acidic pH or high base pH does corrode the cooling system parts, but it does not show up in voltmeter wires dipped in coolant.

Funny thing, if you do the test with the engine off, you got one reading, start and run the engine and it goes up a bit, 12V power for engine systems creates a normal voltage drop in the wire harness if you ground your meter to the neg term on the battery, less so if you ground to the block.

Looks like the only proper test for coolant is pH test strips...

 

[Oro_O]

My problem with the detractors linked above is that there IS a positive correlation between acidity and conductivity (it's pretty axiomatic...). But no one ever does the simple thing of actually lining up samples and taking both readings for comparison.

The "busted" web page is full of a lot of soft-headed double talk that doesn't seem to grasp the real issue of acidity/alkalinity or build a real test. They just make a battery with different electrolytic materials and then effectively say "our batteries made with different materials have different conductivity and thus "voila" - it's not valid because we say so." Big waste of time.

However, that is not me defending the method. Only pointing out they didn't debunk it - if they used real systems instead of table-top batteries, that would mean something.

But the fact remains that conductivity is correlated to voltage, and other places (like civil engineering) do use voltage as a measure of pH. I just think the material presented isn't well documented.

I notice in the AC Delco reference, all that they say the .3V threshold is useful for is finding electrolysis from faults - NOT as a sign of worn out coolant. On the two vehciles I tested, there was no significant drop between running/not running. If that does happen, I think that is more indicative of a poor ground somewhere than it is of totally invalidating the whole approach.