There REALLY are reasons...

[SailingHarmonie]

There really are good reasons that the standards are written they way they are. One of the requirements that generates no end of argument here and other places, is the general prohibition against soldering electrical connections (with a few exceptions). Here is one example showing WHY the standards are written this way. It also shows how little we can trust people who should know better to get things right.

This is on a customer's boat. The engine is a 4 Cylinder Yanmar with the isolated ground option installed. This consists of a two relays, and some wiring that disconnect the battery negative from the engine block except when it is needed. This option was, as far as I know, installed at the factory or as part of the initial engine installation at the boat builder.

The boat has recently struggled with intermittent difficulty of engine starting, and stopping, indicating that there were issues in the ground system. Much time and effort (and money) was put into diagnosing this before the boat was brought to me. The primary ground relay had been replaced, which seemed to fix the issues with starting, but now the engine would not stop using the electrical system at all, and had to be manually shut down at the injection pump.

Another hour of troubleshooting, and I found the problem hidden under some heat shrink tubing. The wires that carry +12 Volts to the coil of the ground relay had been crimped and then also soldered into the ring terminal. One wire had completely fractured at the end of the hard solder joint, and the other had only one or two strands of the wire left hanging on. Not surprising at all that there had been hard to diagnose, intermittent problems.

Now, there is a bunch not to like about this assembly as supplied by Yanmar. The wire is NOT proper boat wire, lacking the higher strand count needed for good flexibility. It might pass the specification, but just barely. Putting multiple wires in one terminal is not great, but probably isn't terrible. The terminal does look to be properly sized. The crimp itself is crap. It looks like it was just mashed together with flat jaw pliers instead of a proper crimping die. But, even though they aren't good, none of these things failed.

All of those things are, at best, sub-optimal, but the single proximate cause for this failure is the solder joint, and resulting lack of strain relief. This forced all of the engine vibration related movement of this wire to focus on this one point, resulting in metal fatigue and failure. Both wires failed RIGHT at the end of the solder joint. The solder joint itself appears to be well done. It is not a cold joint or otherwise deficient, as far as I can see. The amount of solder looks appropriate, and there is not excess wicking up the wire. It was under thin-walled heat shrink, so had at least a little strain relief, but clearly not enough.

Now, I get that this is an extreme case with wires attached to a part bolted to a running engine. This does have more movement stress than most wires will ever see. But this is not the first time I have tracked down electrical problems on engines to fractured solder joints. It shows that this failure mode DOES happen, exactly as the standard writers would have us avoid. I am sure an argument could be made that solder is fine for joints that are not subject to "too much" movement, but how much, exactly, is "too much"?

I'll point out that all this mess was with factory original parts. There is no excuse for it. Such stupid and avoidable quality problems have long since been wrung out of the automobile industry (mostly). It's about time that marine engine suppliers learn the same lessons.

The second photo is how the part was reinstalled. Separate terminals, quality crimps, made with the right tool, and quality, heavy-wall heat shrink.

[noelex 77]

My understanding is soldering is permitted when this is not the only means of securing the wire so I don’t think Yanmar have done anything against regulations by soldering in this example. However, a lack of strain relief is not permitted.

Was there any additional heat shrink that was removed before the second photo was taken?

[skipperpete]

In my personal experience Yanmar does not supply a factory wired insulated earth engine, this one was probably wired as a special request either by a dealership or by the installer.
I replaced a series of Perkins Prima‘s and Volvo MD22’s in French built Alloy yachts (which were all wired above ground) and had to convert the new Yanmars to fully insulated. Yanmar did send me a wiring diagram but refused to supply the engines already modified. I had previously had a boat with an insulated earth Lister diesel so there was no mystery about how and why the wiring needed to be modified and I did em all myself but strayed from the Yanmar sketch to make fault tracing easier, mostly by fitting an earth bus that every single ground wire terminated on.
I’m not interested in the “solder/not solder” discussion other than with battery cables, long ago all battery cables were potted in solder and every now and then during hard or prolonged startups the solder would melt and the cable would fall out of the socket with potentially disastrous results if it found an earth. I guess “Anderson “plugs are off the menu too, the terminals on them are “ potted solder” only

[Statistical]

I have no idea why people still insist on soldering connections on a boat. It is beyond terrible. Technically ABYC doesn't prohibit it as long as it has other method of physical connection but it is still terrible.

I think people do it out of mistaken belief that if crimping is good then crimping plus solder must be super good. No it is ****. You are doing extra work to make things worse.

Solder is never used on connecitons subject to movement. A circuit board even subject to vibration is going to be equally subject to vibration. A soldered component won't be subject to elongation and compression stress. If a circuit board needs to be connected to another circuit board crimp connectors will be used. $2B stealth bombers use crimped connectors on wires.

Right tool for the job and for wiring subject to movement that is crimping not solder. The one exception for me is VHF connectors simply because I have never found one that is solder free which is decent but it also isn't a high amperage connector. For most VHF radios it is pushing less than two amps.

[SailingHarmonie]

Originally Posted by skipperpete

In my personal experience Yanmar does not supply a factory wired insulated earth engine, this one was probably wired as a special request either by a dealership or by the installer.
I replaced a series of Perkins Prima‘s and Volvo MD22’s in French built Alloy yachts (which were all wired above ground) and had to convert the new Yanmars to fully insulated. Yanmar did send me a wiring diagram but refused to supply the engines already modified. I had previously had a boat with an insulated earth Lister diesel so there was no mystery about how and why the wiring needed to be modified and I did em all myself but strayed from the Yanmar sketch to make fault tracing easier, mostly by fitting an earth bus that every single ground wire terminated on.
I’m not interested in the “solder/not solder” discussion other than with battery cables, long ago all battery cables were potted in solder and every now and then during hard or prolonged startups the solder would melt and the cable would fall out of the socket with potentially disastrous results if it found an earth. I guess “Anderson “plugs are off the menu too, the terminals on them are “ potted solder” only

Actually, Yanmar USA does refuse to supply the insulated ground assembly. I have never quite understood why, other than some vague and undefined concern about "liability." The insulated ground kit is a standard Yanmar part outside the USA, and away from its lawyers.

This is odd, because both Beta and Volvo have offered isolated grounds in the USA as a option for some time.

This boat was built in France and the isolated ground kit was a catalog Yanmar part.

[SailingHarmonie]

Originally Posted by noelex 77

My understanding is soldering is permitted when this is not the only means of securing the wire so I don’t think Yanmar have done anything against regulations by soldering in this example. However, a lack of strain relief is not permitted.

Was there any additional heat shrink that was removed before the second photo was taken?

There was no strain relief other than the thin walled heat shrink which was obviously gone by the time the photo was taken. Unless you count a cable tie. Which I do not, because it did not secure the wire to the actual point of attachment of the terminal, so did nothing to reduce mechanical load on the terminal.

My point is solder added nothing of value to the assembly, and only caused a point of failure.

[alctel]

It's also worth noting a crimp with a proper tool is a ton easier to do a good one for a novice than soldering

[Lake-Effect]

Originally Posted by noelex 77

My understanding is soldering is permitted when this is not the only means of securing the wire

It's really more of a case that soldering a crimped connection isn't expressly prohibited... but as this example shows, it didn't do anything beneficial.

[Ballsnall]

Originally Posted by SailingHarmonie

My point is solder added nothing of value to the assembly, and only caused a point of failure.

Interesting that the solder causes a hard point but a crimp does not. I find that logic extraordinary, Have you never seen an non soldered crimp fail at the crimp?

Seems to me the likely root cause was the wire strand count and the lower quality heat shrink used, possibly even mechanical damage.

[skipperpete]

Originally Posted by SailingHarmonie

Actually, Yanmar USA does refuse to supply the insulated ground assembly. I have never quite understood why, other than some vague and undefined concern about "liability." The insulated ground kit is a standard Yanmar part outside the USA, and away from its lawyers.

This is odd, because both Beta and Volvo have offered isolated grounds in the USA as a option for some time.

This boat was built in France and the isolated ground kit was a catalog Yanmar part.

Can you post a link to the standard Yanmar part, I’d really like to get that harness rather than endure the hassles of insulating the out of the box engine. What model engine is that in your pics?

[JPA Cate]

Originally Posted by Ballsnall

Interesting that the solder causes a hard point but a crimp does not. I find that logic extraordinary, Have you never seen an non soldered crimp fail at the crimp?

Seems to me the likely root cause was the wire strand count and the lower quality heat shrink used, possibly even mechanical damage.

To me, the above is a very interesting statement. I don't know about the general membership of CF, but I'm certainly aware how copper work hardens. I don't know about any other metal than s/s, but those two work harden like crazy!


Ann

[s/v Jedi]

Soldering connections is fine when done right. Your electronics is full of it.

The problem is when DIY’ers or “installers” are doing it wrong.

[Matt Johnson]

Originally Posted by Ballsnall

Interesting that the solder causes a hard point but a crimp does not. I find that logic extraordinary, Have you never seen an non soldered crimp fail at the crimp?



Seems to me the likely root cause was the wire strand count and the lower quality heat shrink used, possibly even mechanical damage.

Here's my take:

If you were to crimp right on the end of the terminal, pinching it closed on the wire, then it would be a hard spot like a solder. But you are supposed to crimp further up the barrel, and the remaining part supports the wire helping to eliminate the hard spot.

[Spot]

I am wondering if the copper lug shown in the first post was another interim repair and not factory. I would expect the factory crimps to be tinned. Nice job by SH in troubleshooting and documenting the situation.

One would be surprised how many unsoldered crimped and eyelet connections are used in the automotive wire harness industry. The worked fasteners usually have a surface finish or are made from something other than plain copper.

I just had a inexpensive RV fan with untinned wires fail at the motor to wire solder joint, where the untinned portion between the insulation and solder joint corroded. The one that failed was also not mechanically connected to the motor but its neighbor on a properly formed solder lug with a hole had the same failure mode at the stripped wire.

I was resistant to getting a ratcheting crimper at first...but am so glad I did, between that and better crimps with the heat-shrink ends and additional heat-shrink tubing, one can make really nice working and looking connections.

[wyb2]

Originally Posted by Ballsnall

Interesting that the solder causes a hard point but a crimp does not. I find that logic extraordinary, Have you never seen an non soldered crimp fail at the crimp?

Seems to me the likely root cause was the wire strand count and the lower quality heat shrink used, possibly even mechanical damage.

A good example of conventional wisdom not really standing up to scrutiny for one reason or another.

Seems that lots of people with lots of experience have observed that solder joints fail at a higher-than-acceptable rate. That should be enough to accept that it is a problem. But figuring out why is a slightly different animal.

Encouraging fatigue right where the solder ends seems like a logical mechanism, except as you point out, it’s not clear why a crimp wouldn’t have the exact same problem. I don’t know the answer, but I’m wondering if it’s actually the solder that fatigues and begins to crack first, and the stress concentration/ crack propagation carries on through the copper strands.

Looking at the picture, it seems there is solder on both sides of the break.