Bongo Fury

Must have a chat sometime....you are obviously clued up on this subject and I still have a lot to learn...
The low coolant alarm will not do any harm to the inside of your engines folks......the voltage/current involved is negligible.
The only case I have heard of where this problem has affected a BOngo was a guy who connected a set of powerful horns behind the Radiator and and tried to earth it to the radiator, as the radiator was insulated from the chassie it was earthing via the coolant.....It quickly moved his rad onto the inside of his engine.....(he had also added a second "thin" wire to the chassie so the horns did "sort of" work.
A little knowledge can be a dangerous thing.

This is a good discussion -- but I wasn't concerned about the LCA initially -- and I'm not now either. What I wanted to know was whether it's a good idea to check the coolant for voltage, and whether it is likely that bongo cooling system failures may sometimes be related to electrolysis. Bongos vary in how much onboard electrical gadgetry they have, but does it only take a bad ground or faulty electrical component left unrepaired to wreck the cooling system? Or is this such an unlikely scenario that checking the voltage of the coolant is pointless?

g8dhe wrote:Electrolysis can be confusing because it can arise from two sources.
The corrosion occurs because a current is flowing between two differing conductors it doesn't matter why or the cause of the current.
The first source is due to the use of dissimilar metals, in there own right they will generate a voltage difference, which will cause a current, as you can see in my experiment above, no external source of power is required. In the second case applying a voltage across different parts of the system will cause a current to flow, and hence corrosion, this is the case with the LCA and why Haydn went to such lengths to reduce the current flow down into the region of μA. Normally both situations apply, you have different materials present (first source) and car electrics use the chassis as the earth return for electrical circuits (second source). To do serious damage takes either a lot of current or a lot of time, or the combination of both time and current!

The other thing that causes scratched heads is that the voltage and current don't always obey Ohms Law, because there are chemical reactions going on between the metals and the liquid (electrolyte) these vary with time and temperature and cause all sorts of unexpected results just to catch the unwary!

Thanks Geoff. Sounds like my dusty old recollections on electrolysis are about right. I find myself then trying to make the following hypothesis:

I'd imagined that the potential difference we are interested in in this case, was the one between the the two screws that are screwed into the header tank (at least in my Mk.1 system there are two screws - but did I read that the current design only has one screw?) as this is where the coolant is required in order to complete the circuit. Those two screws appear to be of identical material (stainless steel did Haydn say?), so influence of dissimilar materials can be ruled out. The issue then is what effect putting an amount of power (voltage x current) across an electrolytic solution might have. Both current and (especially) voltage may have an important role to play in promoting chemical breakdown within the coolant, but I imagine the promotion of a chemical reaction within it is the likely effect; one where the dielectric constant of the molecule(s) in the coolant is exceeded, polarisation takes place and chemical reaction occurs, possibly with the creation of a salt which may possibly contribute to coating on the surface of one or other of the s/s screws. I can't remember but did I read Haydn has had reported to him one or two incidences of this?. I suppose if this happens then a failsafe should occur as the circuit will not close properly and the alarm will start to sound. In any case, it is very hard to see how the low currents in the LCA could produce any catastrophic chemical breakdown. But it would be good to know that data on units in service that given problems, is being analysed by someone with an understanding of electrolytic behaviour so that owners can undertake any necessary maintenance etc.

I'd be interested to get your views on the above, as its quite an interesting (if somewhat esoteric - apologies to bored readers!) subject. I'd be interested too to do some bench testing, if I was retired and rich - ha ha!

PS - just read nfns post as I submitted this. Well I reckon that an LCA with 2 screws in the header tank ought not to be promoting electrolysis anywhere except across itself and any coolant in between. However, if the LCA is using the vehicle as part of completing the circuit, then theoretically it is. HOWEVER, the amount of power involved is so low (assuming the design protects against a breakdown where raw 12v power could end up across the system) that it seems v. unlikely it could contribute significantly. Thats my guess (and I stress its a guess) anyway.

The original low coolant alarms have been fitted now for more than 8 years...The test vehicle in Oz has 3 fitted all working together.

Between them they inject 12 times the current of the bongo units.
The car still has it's original 1993 radiator fitted and is in good condition internally.
2 volts times 25 micro equals: W=V x A ?????? watts

Its quite a possibility (but quite difficult to prove) that electrolysis has/will cause problems, given the number of rubber / silicone hoses and an insulated radiator (didn't know that!) in the Bongo. I always run Earth wires wherever possible for new electrical equipment myself, but its not always easy or possible, especially if the item is going to be switched on for any length of time. My main reason for doing this is that I never trust the chassis connections, I might know that my bolt/washer has a good connection but I have no idea really about all the other connections between the bit of chassis I have connected to and the battery! Plus rust and corrosion over time changes things

The big problem with trying to measure the currents involved in electrolysis is that the current flows through the liquid ( I won't call it water-would you drink it ?) and its very difficult to break the liquid in to two parts to measure the current flowing - basically impossible! Hence you tend to use the voltage as a proxy for the current - but this is fraught as well, as shown above, you can get platinum based probes which are more stable (and very expensive) , which can be inserted like and endoscope to measure the voltage at various points along pipes and tubes, as you would expect the voltage varies as you pass along the tube, especially when you get to a connection between different metals, so you measure the voltage ( taking into account the Platinum offset voltage +1.188v) at two different points and assuming you know or have measured the resistance (or conductivity) you can work out the current flowing between those two points.

Just reading the messages come in whilst writing the above - the LCA is NOT a problem! First the currents that Haydn uses are really, really, small, secondly they are within the plastic box, connected by rubber tubing to the nearest metal item, which is where any current would flow to, which means a long and relatively high resistive path and the first metal the currents meet are thick and not exactly in critical points!

The worst points for corrosion are where two differing metals meet, i.e. steel pipe to copper pipe or radiator say, that's where the currents will be highest for simple electrolytic corrosion. Where external currents are involved then you are normally talking of currents above a 100mA into the Amps range (0.1 to 10Amps say) (they LCA uses 0.0001 Amps I believe) then it will normally be worse where the current first leaves the liquid and flows into the metal at the junction between say a rubber pipe and a metal part.

g8dhe wrote:Just reading the messages come in whilst writing the above - the LCA is NOT a problem! First the currents that Haydn uses are really, really, small, secondly they are within the plastic box, connected by rubber tubing to the nearest metal item, which is where any current would flow to, which means a long and relatively high resistive path and the first metal the currents meet are thick and not exactly in critical points!

The worst points for corrosion are where two differing metals meet, i.e. steel pipe to copper pipe or radiator say, that's where the currents will be highest for simple electrolytic corrosion. Where external currents are involved then you are normally talking of currents above a 100mA into the Amps range (0.1 to 10Amps say) (they LCA uses 0.0001 Amps I believe) then it will normally be worse where the current first leaves the liquid and flows into the metal at the junction between say a rubber pipe and a metal part.

That all makes sense to me Geoff. The whole thing is probably a Storm in a Header Tank

Haydn - does the newer LCA have only one screw inserted in the header tank, or is the design still the same as when originally developed for the Bongo (i.e. two screws)?

We now only use one screw as the electronics and calibration have been updated....All it means is a slightly longer "test" beep/flash...If anyone finds it is excessively long then a 2nd screw will shorten it.
It is very much down to the condition/strength of the coolant....we have it calibrated in fresh 40% coolant.
Unless your coolant is in very poor condition or less then 20% I doubt you would be aware of the difference.

haydn callow wrote:We now only use one screw as the electronics and calibration have been updated....All it means is a slightly longer "test" beep/flash...If anyone finds it is excessively long then a 2nd screw will shorten it.
It is very much down to the condition/strength of the coolant....we have it calibrated in fresh 40% coolant.
Unless your coolant is in very poor condition or less then 20% I doubt you would be aware of the difference.

Useful to know that Haydn and thanks. Presumably the longer beep is due to the greater capacitance effect from having a greater distance between the two dependant ends of the circuit (what used to be the two screws on the ealier models). Out of interest, what advantages does the single screw design offer that made the changeover attractive, as I certainly find the older design perfectly robust and reliable in operation.

only one hole to drill...

haydn callow wrote:only one hole to drill...

Thought that might be the case. Customer logic I suppose though as you've said many times, the holes are harmless and easy to do and I think the thinking behind the 2-hole design is very good.

less for us to put in the kit = more "lovely" profit = Haydn a happy bunny

haydn callow wrote:less for us to put in the kit = more "lovely" profit = Haydn a happy bunny

thought that might be the case (again)