Crimped vs Soldered and Crimped Wire Terminals

Technical
1

Tonight 95 tested crimped vs crimped and soldered wire terminals.

Preparation:
We cut 14awg APP and ring terminal leads off of old CIM motors from ~2015-2019 that we haven’t recycled yet. We divided the population into crimp only (black) and terminals we would solder (red). This ensures each sample has a mate in the other population that was crimped by the same person and the same tool. The spread of motors means that we are looking at terminations made by numerous students with different training and experience levels with different tools at different states of wear.

image
image454×629 49.5 KB

The soldered leads were randomly soldered by 4 students with different skill levels and one coach with a high skill level. A cordless and corded soldering iron were used with two different rolls of solder. I believe this randomization gives a better representation of what an average team might get from this practice.

Measurement

Clip lead to bar
Clip student to terminal
Look at scale
Record minimum weight
Subtract minimum weight from student static weight

image
image414×855 33.8 KB

image
image547×728 55.3 KB

Results

Here are the results, divided by terminal and finishing technique.

image
image842×510 7.21 KB

The solder joints that failed real low generally appeared to be okay from visual inspection, but had not flowed into the entire joint. Generally they did not bond well to the un-split side of the terminal.

Both populations had failures where the wire slid out of the terminal or the terminal itself broke. The very strongest soldered joints failed in the wire itself.

Conclusions
-It is totally possible for an FRC team to execute high-quality crimps over the course of years (every crimp exceeded UL strength threshold, a vast majority passed NASA and MIL strength)
-Soldering does improve the average strength of the joint
-Soldering stress-relieves the terminal and wire, so if there is a defect/incomplete solder flow the joint is dramatically weakened

This prep and testing was conducted by myself, my assistant coach, and 5 rookie Grasshoppers… mere crickets, really, but they show much promise.

30 Likes
2

The tests I never knew I needed to whitness, but am happy I now have.
Awesome work as always from everyone at 95!

2 Likes
3

Doing the lord’s* work @JamesCH95.

*Clarke. Duh.

3 Likes
4

An important part of every good experiment

8 Likes
5

Is this true? Can you explain more about what you mean?

1 Like
6

Sure thing.

Copper stress relieves at 150-275C
Copper anneals at 200-400C
Soldering temps start at over 300C and get up to around 380C

So we have lots of overlap between soldering and temperatures that soften or weaken copper and some copper alloys. In my experience soldering definitely has a metallurgical effect on cooper materials.

Terminals and wires we (FRC) use are generally pure, or nearly pure, copper (sometimes with plating or tinning). Copper can only be work-hardened, so we can’t solder them heat treat it later to regain strength. Crimping on the other hand is a wonderful work hardening process, and that is part of the reason it works as well as it does. That and copper-copper interfaces have a coefficient of friction of 1.4 or so.

5 Likes
7

Awesome work! Would it be possible to get a spreadsheet of the raw data? I’d love to play with it. It may even prove useful (and relevant) for teaching some stats concepts to our team.

8

Wild that the soldered connectors have such a broad spread from “weakens” to “strengthens” - good job collecting a variety of skill levels into your experiment.

2 Likes
9

That’s probably why Anderson themselves say to only solder the closed barrel terminals.

Good stuff, I’ve looked for a thread with data like this to point to in the past, now I have it!

4 Likes
10

To me this data says that crimped connections are better than soldered simply because there is less room for error. “Squeeze crimper” is a lot easier than “solder properly”, at least for APPs. And, well, I guess this just validates my understanding of the crimp vs solder debate. Soldering can be stronger, but it also can be way weaker, and crimping is more consistent and almost always “good enough”.

15 Likes
11

The tin-lead solder commonly used for electronic applications melts around 183 C (361 F). The various “lead-free” solders melt around 217 to 220 C (423 to 428 F). I have confirmed this when testing circuitry in an oven set to 200 C and parts fell off because tin-lead solder was used instead of a higher temperature solder.

12

I assume these melting temperatures are correct, and exceeding the melting temperature will certainly cause joints to fail. However, solder is not applied at its melting temperature.

(I don’t mean to sound condescending to you here, but I want to thoroughly explain my thoughts for the benefit of those reading this who are much less familiar)

  1. If the soldering iron/heat source were to be at or just above the solder melting point only a small, near-zero, area of the joint would be soldered because the heat sinking of the terminal and/or wire would rapidly cool the area to below the melting point of the solder. Using an oven like you did avoids those pesky temperature gradients, but I doubt any FRC teams are oven soldering their power leads.
  2. Solders’ surface tension is reduced with elevated temperature, meaning it will flow into complex joints like a crimped wire terminal much more effectively when it is hotter. I would expect to see most soldering irons set to 650-750F for this kind of operation (360-400C). You can draw a mental map of the temperature starting at, say, 350C where the iron is and decreasing to around the melting temperature wherever the solder stops flowing down the wire. Much of the terminal barrel will be well above the liquidus temperature.
  3. People are impatient. They will crank up their iron’s temperature or use a torch to flow the solder joint with nearly zero regard to temperature control.

I cannot find a source I had for the stress relief and annealing temperatures of pure copper, which is frustrating. I recall pure copper requiring 160-170C, ish, to be held for 1hr/in of material thickness to stress-relieve, which would make most soldering processes completely stress relieve the relatively thin terminal. But I can’t reconnect to that source to ensure my memory is accurate.

4 Likes
13

One lesson I learned from running a big winch on an off highway vehicle is that 1/0 cable painstakingly soldered inside a big copper lug, will, when on a long heavy pull, get hot enough to melt the solder inside the lug, allowing it to run out of the lug and cause the connection to become very “high impedance”.

Hopefully this sort of thing wouldn’t be relevant in our scenarios. But I now crimp all of my automotive battery connections.

2 Likes
14

If you can find the winch on my truck, you will also find all of the terminals crimped as well!

15

I used to work with solder dipping robots at a previous job, 63/37 Sn/Pb solder has the lowest melting point (the 183C you noted) but we would typically maintain much higher temps (I think 250C) in our solder pots. Different applications but your soldering iron is typically going to be much higher than 183C and much less controlled.

For reference 60/40 has a melting temp around 190C

16

What crimpers was used for these tests?

17

Yes.

APPs were crimped with one of two Tri-Crimp tools we purchased in 2016. Ring terminals were crimped with one of two or three insulated terminal crimpers we purchased in… 2010? 2000?

No real magic or expensive tools. Just adequate training and occasional checking and adjusting of the crimpers’ performance.

For those who do not know, this is a typical crimper adjuster wheel.

image
image728×410 46.3 KB

4 Likes
18

PDF Report attached, with data.
Crimped vs Soldered and Crimped Terminals.pdf (282.9 KB)

2 Likes
19

I’ve always annealed Copper by running up to red heat and quenching.

Doing some digging, 371 to 649C seems like the right range; you need to hit dull or dark red. Certainly visible in a slightly darkened room.

Longer time at lower temperature will also produce grain growth and strain relief… But I’m pretty sure that 160C is not going to do a whole lot.

20

Annealing is not the same metallurgical process as stress relief.

Do you have any source or data to indicate that?

Stress relaxation/stress relief is a significant design challenge when using copper, even around 100°C. Definitely worth some searching and reading.

1 Like