Cable ferrule conductivity question

user1842

I know that putting ferrules on cable ends is considered good practice but can someone answer the below question.

Nearly all ferrules are tin platted copper. Tin is thus the primary contact surface in the connector. Tin is 15% the conductivity of copper. Does using ferrules not significantly reduced the conductivity of the connection?

Am I missing something here.

2011

I think you need a bit of perspective.
Imagines a copper cable that is 100m long with a ferrule on each end.
What percentage of the entire conductor would you estimate is made up of tin?
Less than 0.001% perhaps?

Lenar3556

2011 wrote: »

I think you need a bit of perspective.
Imagines a copper cable that is 100m long with a ferrule on each end.
What percentage of the entire conductor would you estimate is made up of tin?
Less than 0.001% perhaps?

And is it fair to compare it like that? The tin isn’t evenly divided out across the length of the cable, but rather is concentrated at one point, and may comprise the full contact surface for the connection. Could it contribute to a hot spot at this point?

Obviously not proving to be much of an issue in practice anyway!

user1842

2011 wrote: »

I think you need a bit of perspective.
Imagines a copper cable that is 100m long with a ferrule on each end.
What percentage of the entire conductor would you estimate is made up of tin?
Less than 0.001% perhaps?

This I understand but the connection made is via tin. Therefore do the electrons not have to flow through the tin first which only would have a conductance of 15% of the copper cable.

I'm struggling to understand this and im clearly missing something.

As Lenar3556 pointed out, in reality, it is not an issue. But why is it not?

2011

user1842 wrote: »

This I understand but the connection made is via tin. Therefore do the electrons not have to flow through the tin first which only would have a conductance of 15% of the copper cable.

I'm struggling to understand this and im clearly missing something.

As Lenar3556 pointed out, in reality, it is not an issue. But why is it not?

Take the above cable, overall the resistance is very low. Therefore no issue.

On the other hand if the cable was tin and the ferrules were copper the resistance would be significantly higher so there may well be a volt drop issue.

hesker

Interesting question. I wonder is the reduced conductivity good enough anyway.

I don’t think the explanation offered is correct. If you replaced the ferrule with a perfect insulator you wouldn’t think the insulating properties would be spread over the length of the cable.

2011

As the ferrule is so thin the resistance is so low the conductivity makes no odds.

Look at the formula for resistance

Resistance = (resistivity x length) / cross sectional area = a very small number

Lenar3556

2011 wrote: »

As the ferrule is so thin the resistance is so low the conductivity makes no odds.

Look at the formula for resistance

Resistance = (resistivity x length) / cross sectional area = a very small number

Ok I think I get it. We are perhaps confusing resistivity as equating to a fixed resistance which is a multiple of copper - that is not the case.

Rather resistivity is merely a factor in determining the effective resistance introduced.

When the other factors (such as length) are low it becomes of little material importance, unless it was much higher.

2011

Lenar3556 wrote: »

When the other factors (such as length) are low it becomes of little material importance, unless it was much higher.

Exactly

DublinDilbert

The ferrule is probably electro-plated (or similar process) with tin, so the coating will be very thin, probably measured in microns. This is more than offset by the fact you are effectively increasing the cross section of the cable at that point by inserting it into the ferrule.

user1842

2011 wrote: »

Exactly

Thank you and this was along the same lines as I was thinking. But I was hoping to fine a paper on this where it was tested with a full physical explanation. No matter how much I googled I cannot find such an analysis. I find this very strange as there is an effect on conductivity (albeit miniscule and even could be offset with the increased contact the ferrule makes).

Surely such an analysis was done.

user1842

DublinDilbert wrote: »

The ferrule is probably electro-plated (or similar process) with tin, so the coating will be very thin, probably measured in microns. This is more than offset by the fact you are effectively increasing the cross section of the cable at that point by inserting it into the ferrule.

Indeed and I wonder does a micron coating of another metal effect copper conductivity at all.

As noted above it would be nice to see an analysis of this.

hesker

While it might not contain a detailed analysis the following paper had this to say about tin plating

Tin is used because of the very slow growth of its oxide layer and due to the fact that its oxide film remains very thin under many technically relevant conditions and can be mechanically broken by a low contact force. Good electrical conduction can therefore easily take place.

https://res.mdpi.com/d_attachment/metals/metals-02-00450/article_deploy/metals-02-00450.pdf

whizbang

Its probably not tested, because the advantage of using a ferrule significantly outweighs not using one.

2011

I think what is being forgotten is that although tin is not as conducive as copper it is still highly conducive.

user1842

2011 wrote: »

I think what is being forgotten is that although tin is not as conducive as copper it is still highly conducive.

Indeed, tin is still very conductive and thus it is an non-issue.

I think my overall view was that at some point somebody coated copper ferrules with tin to prevent oxidation of the copper. Surely they would have done all the necessary tests to make sure the conductivity difference between tin and copper did not affect the connection.

The PDF linked by hesker previously is very good but it focuses on gold coating and corrosion rather than conductivity. However the below figures from the PDF note that, indeed, tin coating is an non-issue (silver as the comparison has 105% the conductivity of copper):

Coating thickness in micron: Sn bright (tin) 6, Ag (silver) 6.1

Contact resistance in milliohm: Sn bright (tin) 2.29, Ag (silver) 2.64

hesker

2011 wrote: »

I think what is being forgotten is that although tin is not as conducive as copper it is still highly conducive.

I think that point was already made more or less earlier in the thread when I mentioned “good enough”.

But it’s still an interesting question.


Here's a paper that might have an answer somewhere within it or more likely in the references. I waded through a good bit of it quickly but didn't find an answer but it seems to be about as relevant as anything else I've found.

https://rc.library.uta.edu/uta-ir/bitstream/handle/10106/24069/Scully_uta_2502D_12395.pdf?sequence=1&isAllowed=y

Chapter 3 is fun.

God bless Maxwell!

hesker

Spending way too much time thinking about this but as a final couple of comments

There is an assumption maybe that we are comparing pure tin to pure copper. However the materials will actually be oxidised to some extent. So the conductivity of tin plated copper should be compared to oxidised copper.

Also the ferrule is a hollow cylinder so an unplated ferrule might be more vulnerable to the effect of oxidation than solid or stranded copper cable. This is pure speculation on my part.

2011

hesker wrote: »

I think that point was already made more or less earlier in the thread when I mentioned “good enough”.

I beg to differ, what you actually said was:

I wonder is the reduced conductivity good enough anyway.

^^^ I read this as a question not a statement.
Anyway it seems we agree so its makes no odds.


Tin is highly conductive and in this case is only a few microns thick with a large contact area. Therefore it presents an extremely low resistance path.
I can't see what is supposed to be so confusing

user1842

2011 wrote: »

Tin is highly conductive and in this case is only a few microns thick with a large contact area. Therefore it presents an extremely low resistance path.
I can't see what is supposed to be so confusing

I think the confusion on my side is that it is difficult to find evidence to prove the point. Why is there not a nice research paper measuring resistivity/conductivity of tin ferrule based connections under various loading situations.

Most papers point to the corrosion effects not resistivity/conductivity.

Im sure such a paper exists and there are bits of answers in the papers hesker pointed too.

Of course in reality, as you pointed out, the conductivity of copper and tin is so high that the relative conductivity difference between them becomes irrelevant (even if this difference looks big on paper).

I would like to see the numbers though

2011

user1842 wrote: »

I think the confusion on my side is that it is difficult to find evidence to prove the point.

We know what the resistivity of tin is and we have the formula for the resisatnce of a conductor:

R = ρl/A

Where:
ρ = resistivity
l = lenght
A = area

Not only that but apart from the theoritical aspect we have truck loads of proof of how well this works in reality.
What evidence is missing?

Why is there not a nice research paper measuring resistivity/conductivity of tin ferrule based connections under various loading situations.

I would have thought it pointless for the above reasons.

Most papers point to the corrosion effects not resistivity/conductivity.

That makes sense because resistance is a function of corrision. When the resistance increases it can lead to over heating.

Ah yer all missing the most important factors gents.

Put down the schoolbooks, calculators and reference charts.
I will allow thermal imagers.

In the real world where terminals aren't perfect and we're pushing grub screws into finely stranded flex cable strands get pushed aside, sheared, torn or missed altogether.

When you ferrule finely stranded cable it affords you to use a higher clamping force on the entire conductor that's been compressed to mitigate moisture ingress and wicking as opposed to clamping a portion of twisted strands marmaladed into whatever sized terminal you were offered.
Some mechanisms will not clamp small cables that spread low because the design of them resists it..like putting a 1.5mm² cable in a 63A RCD.

Furthermore, in the legacy of the device, the time and expense taken to ferrule a cable is negligible until that device is rewired or modified, whereupon it starts paying back that invested time exponentially.

hesker

Sir Liamalot wrote: »

Ah yer all missing the most important factors gents.

Put down the schoolbooks, calculators and reference charts.
I will allow thermal imagers.

In the real world where terminals aren't perfect and we're pushing grub screws into finely stranded flex cable strands get pushed aside, sheared, torn or missed altogether.

When you ferrule finely stranded cable it affords you to use a higher clamping force on the entire conductor that's been compressed to mitigate moisture ingress and wicking as opposed to clamping a portion of twisted strands marmaladed into whatever sized terminal you were offered.
Some mechanisms will not clamp small cables that spread low because the design of them resists it..like putting a 1.5mm² cable in a 63A RCD.

Furthermore, in the legacy of the device, the time and expense taken to ferrule a cable is negligible until that device is rewired or modified, whereupon it starts paying back that invested time exponentially.

That’s all great stuff but as your teacher in school used to say: it doesn’t answer the question that was asked.

Does adding mass and clamping torque to a conductor reduce the conductivity?

Academia's got a lot to answer for.

hesker

Sir Liamalot wrote: »

Does adding mass and clamping torque to a conductor reduce the conductivity?

Academia's got a lot to answer for.

So simply that the contact area is increased?

The CSA is increased, the clamping area and pressure is increased, the contact resistance has reduced, the junction temperature has reduced therefore heat-induced resistance, the heat sinking has increased and therefore reduced heat-induced resistance, atmospheric atrophy has decreased, the second time you use it's as good as the first.

Another thing is if I have a choice between a sparks who gives an eff enough to use ferrules and one who doesn't or claims "you don't need to crimp them" for the same price. I'll go with the former and I guarantee you it'll be a better more reliable installation as a product of that attitude.

The books would have me believe that all RECs uphold the same standards.

2011

Sir Liamalot wrote: »

Ah yer all missing the most important factors gents.

Put down the schoolbooks, calculators and reference charts.
I will allow thermal imagers.

In the real world where terminals aren't perfect and we're pushing grub screws into finely stranded flex cable strands get pushed aside, sheared, torn or missed altogether.

When you ferrule finely stranded cable it affords you to use a higher clamping force on the entire conductor that's been compressed to mitigate moisture ingress and wicking as opposed to clamping a portion of twisted strands marmaladed into whatever sized terminal you were offered.
Some mechanisms will not clamp small cables that spread low because the design of them resists it..like putting a 1.5mm² cable in a 63A RCD.

Furthermore, in the legacy of the device, the time and expense taken to ferrule a cable is negligible until that device is rewired or modified, whereupon it starts paying back that invested time exponentially.

I agree with all of the above but as hesker said this does not answer that was asked in the OP.

In the opening post user1842 seemed to think that as tin has a lower condictivity than copper that this would somehow be an issue.

user1842

2011 wrote: »

I agree with all of the above but as hesker said this does not answer that was asked in the OP.

In the opening post user1842 seemed to think that as tin has a lower condictivity than copper that this would somehow be an issue.

Indeed that was my initial thought but in reality it is an non-issue. I should have been more specific in my initial post and asked what is the measured conductivity/resistivity difference of using a tin plated copper ferrule compared to just a non-plated copper ferrule.

Based on the good discussion in this thread I would think that any very minor difference in conductivity/resistivity would be more than offset by the non-oxidation properties of the tin plating as opposed to bare copper.

2011

user1842 wrote: »

Indeed that was my initial thought but in reality it is an non-issue. I should have been more specific in my initial post and asked what is the measured conductivity/resistivity difference of using a tin plated copper ferrule compared to just a non-plated copper ferrule.

Based on the good discussion in this thread I would think that any very minor difference in conductivity/resistivity would be more than offset by the non-oxidation properties of the tin plating as opposed to bare copper.

I think that the difference in resistance of a coper ferrule and a tin plated copper ferrule is so small that it is amost immeasurable.

2011

The question makes sense but the answers don't entirely to me and are overlooking the significance of practical applications. Let me put it another way;

Do turbos save fuel? Theoretically yes or at least maybe.

In the real world; no, you use more fuel because you get better torque, hp, acceleration and top speed.

2011 wrote: »

I think that the difference in resistance of a coper ferrule and a tin plated copper ferrule is so small that it is amost immeasurable.

It's very measurable on a boat after 5 years.

2011

Sir Liamalot wrote: »

It's very measurable on a boat after 5 years.

I think your boat connections has more significant issues than whether the ferrules are tinned or not

Yurp, the owner sank it for a week.
I ran for the hills, one of those cheaper buy a new boat jobs.

hesker

Was reading up about another aspect of ferrules when I found this pdf below. Thank God the lockdown restrictions are easing as I clearly need to get out of the house.

I think it supports what I said earlier about oxidation.

https://tinyurl.com/hud38sy3

In summary a properly applied ferrule breaks through the oxidised copper and provides a connection with superior conductivity vs bare copper cable.

As soon as you drive a screw terminal into the ferrule it'll etch away the tin on the contact area more than likely while preserving the rest that's exposed to the environment.

Don't get me wrong I support nerd behaviour...I have on occasion ferruled my crimps..ahem..for CSA mechanical robustness reasons...

I generally think you learn more on the tools and verifying empirical results than theorising...

When you're dealing with big cables and lugs a crimper is effectively compressively welding the copper together..it's a no-brainer that it's better.

hesker

Sir Liamalot wrote: »

I generally think you learn more on the tools and verifying empirical results than theorising...

They both have their place in learning but it’s in the practical application of something that you usually really understand the basis of theory.

Agreed although I usually find it as a reference for how much I was lied to by marketing and advertising and ascertain the percentage derating I have to apply to the device I was building.

The difference between pdf sell sheet performance and actual real-world performance in the off grid industry is a shockin' disgrace.

Just for context after 2 years testing I have yet to see li-ion beat lead-acid at anything but energy density.
In fact Li-ion is performing way worse when you factor the efficiency losses associated with thermal management and the cost per kWh.

10-10-20

Tin is the sacrificial coating, it's more reactive than the underlying metal, but also much weaker as it's a soft metal.

Tin makes its own problems too. It has a very nasty habit of creating hair-like growths called "tin whiskers" when used in power equipment and for that reason is no longer used in a lot of IT equipment (such as rack power distribution units).
https://nepp.nasa.gov/whisker/background/index.htm
https://nepp.nasa.gov/whisker/anecdote/af114-transistor/index.html

The growths are often attracted to the opposite pole and will grow until they reach the flashover distance, short and cause an arc/plasma path and completely vaporise. You'll be a while root-causing those ghosts if a breaker drops!