Fundaments Friday – Crimp Vs Solder Connections

[funwithbikes]

Solder vs Crimp

If like many new e-bike builders, you use the forums to find information about how to build an e-bike you will soon come across a solder vs. crimp discussion. Many builders are passionate one way or another.
Let’s sidestep the passion. Like other aspects of building an e-bike, choosing how to make electrical connections is an engineering decision. Every engineering decision has strengths and weaknesses. Those strengths and weakness can be understood and weighed against each other. In some situations, there is a clear winner. In some cases, the best solution is less clear. I thought I could take a few minutes to look a few of the issues engineers use when choosing how to make electrical connections.

• Quality Assurance
• Installer Skill
• Mechanical Connection
• Tooling and Consumables
• Electrical Connection

If this article falls into the tl;dr category. The takeaway is that if done correctly, soldering and crimping provide equally good electrical connections. Other considerations are more important.

Quality assurance - In a situation requiring high reliability, the final decision often comes down to quality assurance. It is very easy to see if a crimped connection has been correctly made. It is more difficult to inspect solder joints. A solder joint may look good on the outside while note making good mechanical or electrical contact.

Installer skill and speed - With the correct tools, it is easier and faster to make good crimp joints than it is to make good solder joints. Soldering is a skill that takes practice and research to do well. Crimping is just a matter of using the proper tool to compress a connector into place.

Network, phone, and cable installers use crimp connectors. It is quick and easy to make high quality connections in the field.

Mechanical connection - The most often cited argument against solder connections is that when subjected to vibration they can be less reliable. To understand this issue we need to look at the physics of soldering. The soldering process involves liquid solder wicking between the components being connected and then hardening into a solid mechanical and electrical connection. In a stranded wire, the wicking action causes solder to creep between the strands of a wire.

Stranded wire is more flexible than solid wire of the same size because the strands slide past each other as the wire bends. If solder has wicked between the strands near the solder joint, the strands cannot slide past each other as easily as in unsoldered wire. The wire becomes stiffer.

This can cause failure if the joint is subject to vibration. The effect is particularly pronounced in automotive and marine applications. An engine vibrating at a few thousand hertz for 10,000s hours can wreak havoc with soldered connectors on thick wires.

Tools and consumables – The most significant advantage of solder joints is that nearly all small to moderate sized joints can be soldered with a $20 soldering iron and a spool of solders.

Crimp joints require properly sized connectors and tools for the specific application. Over time builders ends up acquiring a small inventory of connectors and surprisingly expensive tools

If properly done, the electrical conductivity of crimp and solder joints is so similar that designers consider differences in conductivity insignificant.

In most e-bike situations, the quality of the joint is more important than the specific type of joint. You will get more bang for your buck spending time in the shop honing your skills at make connections than participating in forum discussion about them.

[Alan B]

I agree with most of the OP's posting above, however I will present a somewhat different set of conclusions.

Soldering works well for small wiring connections - electronics. It's use for larger connections is problematic. Solder has higher resistance than copper by a large factor, and it essentially guarantees joint heating and subsequent failure if the wire ever gets hot enough to melt the solder. As an engineer, I cannot recommend solder for the high current connections we find on ebikes, except as a last resort. The highest current connections are between the controller and motor, where the motor current can be multipied many times over the input battery current.

Ebikes present significant vibrations to their wiring, especially those without suspensions, as their jarring shaking is very harsh. Soldering removes flexibility in large conductors and makes them susceptible to failure from this vibration.

Soldering is not used commercially for high current situations. Crimping is the rule there. Look at how the power company connects your power lines - with a portable crimper. Never solder. Many ebikes run more current to a hubmotor than the power company feeds to older houses. When I spoke with the West Coast Sales Manager for Anderson, makers of the PowerPole connectors we often use, they never recommend soldering large connectors and are appalled at the soldering people use on them.

Get the proper tools and crimp those connections, unless they don't support that mode of connection or you just cannot afford it. These days hydraulic or third party crimp tools are not very expensive, and make very good connections. If you don't have the right tools to crimp properly soldering can be used as a less reliable solution. We frequently see failed soldered joints in ebike use.

The $20 soldering pencil is fine for soldering small wires, but to solder the heavier high current wires takes a lot of heat and the small tools will fail to produce a quality solder joint. Much larger soldering irons, guns, or torches are required. These often cause secondary problems with too much heat and damage to neighboring items.

Solder if you have no better choice.

[craigsj]

"Solder has higher resistance than copper by a large factor, and it essentially guarantees joint heating and subsequent failure if the wire ever gets hot enough to melt the solder."

The melting temp of solder is high enough that you would have far greater problems than joint failure if it comes to that. "As an engineer" you should know this and I'm sure you do. Otherwise the "joint heating" doesn't mean anything. Yes, solder has higher resistance than copper "by a large factor", but so does nickel yet that stops no one from advocating its use in battery packs. In fact, solder has only twice the resistivity of nickel yet nickel is five times as resistive as copper, so it would roughly correct to say there's little difference between solder and the stuff you build your packs out of. Somehow that's OK when it's battery current but not for joining wires. ;)

As an engineer, I would say that bikes don't have "high current" connections all things considered.

"Ebikes present significant vibrations to their wiring, especially those without suspensions, as their jarring shaking is very harsh. Soldering removes flexibility in large conductors and makes them susceptible to failure from this vibration."

Can you quantify "significant vibrations" and "very harsh"? In comparison to what? Soldering removes flexibility in large conductors but so does crimping, so apparently both are susceptible to failure from "this vibration". Why the double standard?

"Soldering is not used commercially for high current situations. Crimping is the rule there."

In many cases this is due to reasons other than the quality of the result. Crimping has a non-recurring cost that Is meaningless commercially but may matter a lot to DIY. At sufficiently high currents soldering is a problem but that's not the case here.

"If you don't have the right tools to crimp properly soldering can be used as a less reliable solution. "

Prove that soldering yields a "less reliable solution". All you're offering are vague references to things that don't apply. Let's see proof that a soldered connection is inferior to a crimp for a wire designed to carry 100 A.

"Solder if you have no better choice."

Crimp if you have no better choice.

Soldering is not a problem and you don't have to buy a bunch of expensive, one-purpose tools.

[Eric Luna]

excellent well written post... thanks for this funwithbikes :)

[spinningmagnets]

Thanks for starting this useful and detailed thread. Whether a builder uses solder or crimp, I firmly believe the best possible joint is one where the strands of both butt-ends are intermeshed before they are either soldered or crimped.

There are plusses and minuses to either method. I can cut, intermesh, crimp, and then heat up the heat-shrink insulation with a lighter. I can do this on a mountain side, while being miles away from an electrical outlet for running a 100W soldering iron.

high heat applications can melt solder, but if you are getting an electrical joint that hot, your wires are too skinny for the peak amps you are running. That being said, if your joints are hot, use a crimp.

I solder most joints as a method of protecting the copper strands from oxidation. That is not an issue near the dry desert regions, but ebikers near the ocean have seen the green cancer from humid salty air. I use marine-grade glue-lined 3:1 heat shrink insulation over all joints, even if crimped (although, I have some "liquid tape" that I may use some day...)

Concerning a soldered joint, if you lay two copper stranded butt-ends side by side, the current will flow from one to the other through the solder, which is actually a poor conductor. Even if you twist the two butt-ends, it's the same. Intermesh the strands first, then spiral-wrap the joint with a couple thin strands from a scrap of copper wire, and then solder.

I know solder makes a joint stiff, but I never place them so that they are under strain. I only use enough solder to seal the intermeshed copper strands in the joint.

[funwithbikes]

Alan,
Would you mind posting a howto on your crimping technique? I agree that a well done crimp provides an excellent connection.

Any soldering gurus willing to share their techniques?

This issue came up for me as I was building some wiring harnesses the other day. I was trying to figure out how to connect some XT90 connectors for my battery cables. It took four attempts before I felt comfortable with the joint.
1. The first one was cold and ugly. It looked alot like a tiny dog took a dump :(
2. The second one looked good... so I used a cutting wheel on a dremel tool to cut the joint into slices to verify it was solid all the way through.
3. I screwed up the third one by reversing the + and - connectors. arggg.
4. Fourth time's a charm.

For full disclosure, I ended up using two soldering irons to rapidly heat the joint before applying the solder.

fwb

[Alan B]

The XT connectors require soldering of course, since they are not designed for crimping. For them I use a small flame butane torch, hold the connector in a small vise (it helps to mate it with another connector to reduce the chance of melting and pins going out of alignment), pre-tin the cups and wire with solder; position the heatshrink or cover on the wire, and then reheat as the tinned wires are slid into the cup. The last step is shrinking the tubing or sliding the wiring cover into place.

[funwithbikes]

Excellent post. https://www.electricbike.com/matchin...to-controller/

One of the reasons I started documenting this step by step cyclone 3kw build was to curate the e-bike information I found searching around the internet while doing my first build for other new builders. Sadly, your mid drive build articles did not show up using any of the terms when I searched google. Probably because I hadn't figured out to search for "mid drive electric bike diy build" :(

[craigsj]

Quality assurance - In a situation requiring high reliability, the final decision often comes down to quality assurance. It is very easy to see if a crimped connection has been correctly made. It is more difficult to inspect solder joints. A solder joint may look good on the outside while note making good mechanical or electrical contact.

I don't agree that a crimped connection is easier to verify visually than a soldered one nor do I think it's relevant to DIY. You don't have separate inspectors with DIY.

One advantage a production line has is the ability to assume that the right crimp tool is used. This makes crimp inspection easier ifthey do it at all. With DIY you can't assume the right tool was used. Without adequate crimp force, a crimp connection is suspect just like a solder joint could be. I think the issue is moot.

Installer skill and speed - With the correct tools, it is easier and faster to make good crimp joints than it is to make good solder joints. Soldering is a skill that takes practice and research to do well. Crimping is just a matter of using the proper tool to compress a connector into place.

And this issue is important in production but not really in DIY. Stocking parts and tools is more important. Soldering wins in that way.

Network, phone, and cable installers use crimp connectors. It is quick and easy to make high quality connections in the field.

But these connections are not for power so it's irrelevant.

Stranded wire is more flexible than solid wire of the same size because the strands slide past each other as the wire bends. If solder has wicked between the strands near the solder joint, the strands cannot slide past each other as easily as in unsoldered wire. The wire becomes stiffer.

True also for a crimped connection.

This can cause failure if the joint is subject to vibration. The effect is particularly pronounced in automotive and marine applications. An engine vibrating at a few thousand hertz for 10,000s hours can wreak havoc with soldered connectors on thick wires.

Surely there is some reference that shows that soldered joints are more susceptible to this failure than crimped? If not, you shouldn't suggest otherwise.

[Alan B]

There are plenty of references as well as manufacturer's data and electrical standards that show crimping is preferred and/or required over soldering for power wiring. Check with Andersen, manufacturers of the PowerPole and SB connectors that are commonly used in electric vehicles such as forklifts, for example.

Electrical utilities do not solder - crimping, welding or bolting is required.

Electrical building codes require crimping and don't allow soldering.

Vehicle standards don't use soldering for power wiring.

Soldering is restricted to electronics and components or wires that don't move. Soldering creates a very high stress point where the wire enters the rigidly soldered volume.

This is not to say that soldering doesn't work, it produces a serviceable result if done properly. But solder wicking deep into the cable creates a more failure prone situation that crimping avoids. Also the large mass of copper often wicks heat away so quickly that the solder joint is not properly made deeper in the cable where it is not visible.

So solder if you like, but use strain relief and don't be surprised when maintenance is required. Using techniques such as glue-filled heatshrink can rigidify the soldered volume and keep the flexing portion of the wire away from the solder joint itself and avoid the strain failure. For example quality soldered connectors generally have built in cable clamps to avoid the flexing that causes the soldered joint to fail.

[craigsj]

Certainly connectors designed to be crimped recommend crimping. No need to ask Anderson about that.

Standards and codes don't inform us in any way about the reasons or whether they are applicable to the task at hand. I hope you aren't suggesting that high voltage building wiring is relatable to electric bikes.

"So solder if you like, but use strain relief and don't be surprised when maintenance is required. Using techniques such as glue-filled heatshrink can rigidify the soldered volume and keep the flexing portion of the wire away from the solder joint itself and avoid the strain failure. For example quality soldered connectors generally have built in cable clamps to avoid the flexing that causes the soldered joint to fail."

I've been soldering for 40 years and would absolutely be surprised "when maintenance is required". Soldering is not advised under certain circumstances but none of that is mentioned here nor is it applicable. "Rigidify"-ing a "soldered volume" would be the opposite of what is needed if the problem is that solder "rigidifies" a joint as is claimed. Crimping can cause stress points as well.

I will, of course, "solder if I like" because I've been around the block and don't need advice on the issue. The general point seems to be that both techniques will work if done properly and that is true. The opposite is also true...that both can produce poor resuits if not done properly, yet only half of that point is made. Instead there seems to be an effort to demonize soldering.

I've certainly seen my share of bad solder joints but I haven't seen solder joints that, while properly done, "require maintenance". On the other hand, I've seen plenty of poor crimp jobs that have to be repaired. I suspect this is more a religious issue than anything else. Oddly enough, there are also solutions that both crimp AND solder and there are applications where soldering is specifically REQUIRED. Wonder why those aren't mentioned?

[Alan B]

There is plenty of industrial information showing that crimping is superior to soldering for cabling.

Soldering can be good enough. Don't breathe the smoke, and the lead is bad for your health.

Make your own choice.

[Alan B]

I've been soldering for considerably longer than 40 years, but that doesn't make it any more well suited for high current and high vibration applications. Talk to the manufacturers of the connectors, find out why they make them with crimps and not solder. Lots of people solder Andersons. I have had that conversation with one manufacturer, and they know why crimping is better, and what the problems are with solder. Only in the RC industry do we see small high current soldered connectors, and the life cycle of RC toys is not significant. They are more concerned about cheap, small and light than highly reliable. And we do use some of those connectors, even though they have problems, some of which I describe below.

Making a good solder joint takes both equipment and skill, making good crimped joints requires equipment but not much skill, and the consistency and success ratio is higher. Soldering tends to damage the insulation and possibly nearby items, it is one thing on a circuit board but in the wiring harness of an ebike is another. Would you suggest encouraging folks who have no experience and are just learning to solder to be doing that? As a percentage I've seen many more failed solder joints in ebike cables than failed crimps. If you had an ebike shop, what would you want your employees doing?

Even on a circuit board the leaded components must be physically restrained, (glued or potted) or their leads will fracture off at the solder junction due to the high stress created there. It is really basic physics. Anyone who has ridden a (non suspension) bicycle knows how jarring the bike is on rough roads. No bicycle experience here, perhaps?

In ebikes we often push the envelope. Have you had your connections come unsoldered due to high current? This happens from time to time in ebikes, especially in battery or motor high current leads when folks push the envelope. When pushing lots of current the solder fails long before the wire. And the wire is often marginal for the peak currents. One of my ebikes routinely pulls 80 amps for the duration of a steep hill, and that's conservative, the controller will handle more and a simple setting change would increase that current and heat up the connectors and wiring. Yes, there is high temperature solder, and copper welding, but neither of these is within the scope of practicality for ebike users. At work we often check high current wiring with an infrared camera, if those connections were soldered they would melt and fail, whereas crimped and bolted connections just get hot and have a greater safety margin. There are plenty of examples of melted solder in ebike forums.

You are welcome to do your own research and have your own opinion, or religion as the case may be. There are plenty of folks who like soldering connectors, and that is your and their choice, and it has already been presented in this thread as a viable option, with some caveats. I have repaired failed soldered connections on ebikes that were soldered just fine (the wire breaks right next to the soldering), but have not had failures in properly crimped connections. If a crimped connection fails it is generally due to an improper crimp, or the wire was physically yanked out and would have failed regardless of the connection type. The bicycle is a far more violent and jarring environment than you generally experience.

I've used plenty of soldered RC connectors on ebikes and another thing I don't care for is the production rate for soldering them. It takes a lot more time and care to make good soldered connections with #12 and heavier wire, and they take considerable time to cool. We are forced to do it or face changing the connectors on every Lipo brick, but that doesn't make it a great connector choice. These connectors are not easy to solder well as the connector body is not made from a heat resistant plastic and they make zero mechanical connection which is rule #1 in soldering (or it was before surface mount came along), but they are popular and generally reliable enough. The connectors often have some important hazards such as not physically protecting against cross polarity, which allows two bricks to plug into each other with reversed polarity. Not great connectors from either a safety standpoint or a production rate perspective. So use them with extra respect, merely plugging them together wrong can instantly melt the connector and wiring and flash molten copper onto your fingers or worse. Wear leather gloves and eye protection when working with live battery connections (and batteries are always energized).

It still seems to me that "crimp if you can, solder if you must" is the best approach for reliable ebike wiring. This goes along with providing good wiring support, protection and insulation such as heatshrink and cable ties, as well as using safe connectors and standard wire color codes to help insure correct wiring both initially and later as the wiring is maintained and modified.

[Alan B]

Ebikes are not a typical situation when it comes to wiring. The vibration and jarring when hitting a pothole with small tires operating at high pressure (and no suspension) generates large G forces. It can be enough to make the riders hands numb if gripping the handlebars tightly. The rider learns to grip softly and many wear gloves, but the bike and wiring is subject to these forces. Vehicles don't see these forces beyond their suspension components. Ebike components and wiring have a rough life. Look inside an ebike controller, the large components are glued to the PCB, otherwise the vibration will fracture their leads near the solder joint.

Soldering a wire reduces its current carrying capability compared to crimping, and it introduces a failure point. It is easy to test this yourself, here is one example:

http://www.esatinc.ca/News_Letters/Crimp_or_solder.pdf

Take the wire and add soldered and crimped connectors to it, then subject it to high current until it fails. Observe what fails first. Tin and lead have a significantly higher resistance than copper, so if the current passes through the solder the increased resistance will cause a hotspot. The RC connectors typically don't have a good mechanical connection, thus forcing the current to flow through the solder, and mechanically failing as soon as the solder melts. The common bullet or XT connectors have a puddle of solder with the wire end sitting in it, which makes no attempt to minimize the current path through the solder and provides zero mechanical support for the wire once the solder is melted. RC connectors are typically operated with extreme airflow and short runtimes so their ratings are extremely optimistic and they should be considerably derated for ebike use. The RC ratings should be considered "peak" values.

We have seen solder melting type joint failures occur in ebikes (where the wiring gauges used are commonly below industry ratings considering the peak currents). Failures are generally in the conductors between the controller and the motor where current multiplication causes greater currents than on the battery side of the controller. The most common that I've observed is inside the motor where the temperatures compound, less often we see this type of failure in a connector or splice in the motor lead. A direct drive hubmotor can see hundreds of amps at low speed where back EMF is near zero, and the wiring is likely to be only 12 gauge or smaller wire size. The wiring is sized for the "average" current so these current peaks raise the temperature and stress the connections. If the bike is operated on a steep gradient at low speeds this stress condition can last long enough to cause failures, especially in cases where the user has increased the current settings for additional torque.

Since the wiring in ebikes is often operated briefly at higher currents than wire tables recommend, we are in a different situation than "normal wiring". Thus we are interested in the performance of these connections at these extreme performance levels. Depending on your ebike, these conditions may not occur, a 250 watt ebike is not pushing the high currents, but they often use much smaller wire and may end up in a similar situation. A 750 watt hubmotor can easily see double or triple that power level when accelerating from a stop or climbing a hill, and this is where the quality of connections gets tested. If the hill continues long enough it can cause failures of the connections (or the connectors if they don't have adequate margins).

Production rate may not matter to some, but when putting together a battery pack with 24 Lipo bricks requires soldering 48 wires to the connectors, and another 48 at the other ends of those wires to make up the series/parallel harness it adds up, and having several ebikes the time saved can be substantial. Crimping is quicker. Crimp tools are becoming more available at reasonable prices. Soldering requires more than one soldering tool for the range of connections on an ebike, most folks don't have proper soldering gear for the heavier wiring and the result is poor connections, damage to insulation and connectors, etc. Crimping requires a few tools also, I use about 4 crimping tools on my ebikes, but most can get by with fewer.

Crimp when you can, solder when you have to.

[commuter ebikes]

A reminder to all members to please keep it friendly and uplifting. Argue ideas until the cows come home, and if you happen to encourage, help or befriend another forum member along the way, all the better. Your ideas are welcome, but please do not post personal attacks or insults here.

Ebikes are a super fun hobby. You get all the mechanical and electrical fun as well as the riding excitement.

[commuter ebikes]

The following is a heavily edited text of craigsj's responses to Alan B.'s posts which I edited in an effort to keep things friendly and constructive. It is restored here for the purpose of free speech and data integrity. I apologize to the reader if my editing has deleted any useful data or distorted the original text.

Alan B.: "Talk to the manufacturers of the connectors, find out why they make them with crimps and not solder."

craigsj: "What are "the connectors"? The ones typically used in e-bikes? Not one could make a case that it's because solder is inferior. Furthermore, plenty of such connectors are solder only."

Alan B.: "Only in the RC industry do we see high current soldered connectors, and the life cycle of RC toys is not significant. They are more concerned about cheap, small and light than highly reliable."

craigsj: "Cycling is concerned with "cheap, small, and light", however, so it seems like a good fit. The question is suitability and solder works here. Crimping is great for volume and repeatability, it is not a requirement at these power levels."

Alan B.: "Even on a circuit board the leaded components must be physically restrained, (glued or potted) or their leads will fracture off at the solder junction due to the high stress created there. It is really basic physics. Anyone who has ridden a (non suspension) bicycle knows how jarring the bike is on rough roads."

craigsj: "Potting boards is not simply about "physically restraining" components, in fact, it's not primarily about that. Lots of PCBs don't get potted and don't have their parts "glued" despite being in higher stress environments than e-bikes. I don't even accept that e-bikes are high stress environments. Are e-bikes are high stress environments requiring the gluing or potting of components and the avoidance of solder joints. Ever looked inside a commercial battery pack?"

Alan B.: "Have you had your connections come unsoldered due to high current?"

craigsj: "No."

Alan B.: "This happens from time to time in ebikes, especially in battery or motor high current leads when folks push the envelope"

craigsj: "Properly soldered? At what current? At what temp?"

Alan B.: "Yes, there is high temperature solder, and copper welding, but neither of these is within the scope of practicality for ebike users."

craigsj: "You aren't melting any solder putting 80A through a properly constructed joint. Few, if any, motors can withstand anywhere near the melting point of lead-based solder. Likewise, all the heat shrink you use to reinforce those "weak" solder joints will have long since melted as will the hot melt glue you see all over commercial packs. Melting solder is irrelevant. Solder may be unsuitable for 1000 amp connections but is fine for 100 amp ones. If you can use spot welded 0.1mm nickel you can use solder."

Alan B.: "The connectors often have some important hazards such as not physically protecting against cross polarity, which allows two bricks to plug into each other with reversed polarity."

craigsj: "Crimp doesn't protect against this. The choice of connector doesn't argue the superiority of crimp over solder."

Alan B.: "Ebikes are not a typical situation when it comes to wiring. The vibration and jarring when hitting a pothole with small tires operating at high pressure (and no suspension) generates large G forces. It can be enough to make the riders hands numb if gripping the handlebars tightly. The rider learns to grip softly and many wear gloves, but the bike and wiring is subject to these forces. Vehicles don't see these forces beyond their suspension components. Ebike components and wiring have a rough life. Look inside an ebike controller, the large components are glued to the PCB, otherwise the vibration will fracture their leads near the solder joint."

craigsj: "Not all e-bikes have no suspension, not all have small tires and high pressures, not all are "hitting potholes", and all this is not what makes hands "go numb". Furthermore, "vehicles" are not immune from these kinds of stresses simply because they have suspension. If that were true, then e-bikes with suspension would have no such stresses. Meanwhile, many "vehicles" with suspensions have vibration sources far greater than road variations that e-bikes encounter. What matters are the actual stresses that actual components must endure. There are far more stressful environments than e-bikes; car or motorcycle are common examples."

Alan B.: "Soldering a wire reduces its current carrying capability compared to crimping, and it introduces a failure point."

craigsj: Both soldering and crimping introduce a failure point, the same failure point and for the same reason. Don't assume any crimp is part of an engineered solution that solves this problem. The failure point in a crimp is arguably worse than in a solder joint. Soldering, when properly done, can result in a joint with lower resistivity than the wire it joins.

Alan B.: "The RC connectors typically don't have a good mechanical connection, thus forcing the current to flow through the solder, and mechanically failing as soon as the solder melts."

craigsj: Regarding solder vs. crimp, just because some solder joints can fail does not mean that soldering is inferior. Furthermore, just because current is forced through solder does not mean the solder will melt. Solder has conductivity about half as good as nickel, it's poor compared to copper but it is not a poor conductor and its melting temp is quite high when you consider it is connected to the best possible heat sink material. A joint fails because it carries far too much current (in e-bikes), not because it is soldered. "Wiring tables" don't "decide" what's normal vs. "extreme" nor do they say anything about "operated briefly" nor do they consider the relatively short lengths of wire typically used in e-bikes.

Alan B.: Production rate may not matter to some, but when putting together a battery pack with 24 Lipo bricks requires soldering 48 wires to the connectors, and another 48 at the other ends of those wires to make up the series/parallel harness it adds up, and having several ebikes the time saved can be substantial. Crimping is quicker. Crimp tools are becoming more available at reasonable prices.

craigsj: Most won't be putting together 24 LiPo bricks for a bike. However, those that have a lot of joints can factor in both the time and cost and make a decision on whether a proper crimp tool is worthwhile. For large projects, the cost of tools is immaterial and a strong argument can be made for crimp provided suitable solutions can be found. This isn't proof that solder is inferior for other tasks.

Alan B.: "Soldering requires more than one soldering tool for the range of connections on an ebike, most folks don't have proper soldering gear for the heavier wiring and the result is poor connections, damage to insulation and connectors, etc. Crimping requires a few tools also, I use about 4 crimping tools on my ebikes, but most can get by with fewer."

craigsj: This is, for the most part, not true. You can do fine electrical work AND solder wires larger than 10 gauge with the same tool...if you have the right tool. Furthermore, the right tool costs less than many single crimp tools. But yes, its true that a bad solder joint is worse than a good crimp. A good solder joint is better than a bad crimp and, yes, bad crimps exist. If you want good joints, you need proper tools and proper skills. When you have these things, either will work.

[commuter ebikes]

My ebikes have both precrimped JST connectors (which I have found to be very reliable) as well as soldered joints (particularly on the larger wires). I have not had any failures yet, but I will certainly post them here if and when they occur. I put a lot of miles on bikes with no suspension, so we'll see how that goes.

[Chris Duke]

Why not both?

I've been rewiring motorcycles for decades. When I'm not being lazy, I'll do both on most connections.
The advantages as I see them, you can use only a small amount of solder vs just doing a pure solder joint, and your mechanical connection will never become loose enough for a wire to pull out.

It works for me, and I've never had a failed connection due to vibration.