Crimp *and* solder a crimp connector? Good? Bad? Why?

[george graves]

For connectors that are designed from the get-go to be crimped, what's the harm to soldering them as well? I've heard the rule, and usually stick to it.  But I don't know the reason behind it, or the exceptions to the rule(if any).

I assume there is going to be a small about of solder to creep up into the strands and reduce flexibility. What's the failure mode for that?  If the flexibility isn't needed?  But actually, I ask assuming a automotive or vibration environment.  (Surely a static environment it would be fine?)

And does that answer change on the gauge of the wiring?  Say something like 22 gauge vs 0 gauge where there is almost no flexibility in a 0 gauge wire anyways?

[coppice]

There isn't going to be a little solder wicking up the strands. There is going to be a lot. You can attach a heat sink to the stranded wire as you solder, to reduce the wicking effect, but you can't really stop it. The result is a wire that breaks after a few flexes. If you have crimped properly, forming cold welds between a clean wire and a clean terminal, the joint will be more reliable that solder by some orders of magnitude. If you crimp poorly, and just squeeze everything together, don't expect good results.

[TopLoser]

I've never done it because the soldering operation melts the insulation that you have just clamped as part of the crimping operation, compromising the mechanical strength. Plus it may change the state of the insulation, making it brittle or soft. Plus the obvious wicking of the solder up the strands making the wire less flexible.

Lots of bads in my opinion.

[tautech]

There isn't going to be a little solder wicking up the strands. There is going to be a lot. You can attach a heat sink to the stranded wire as you solder, to reduce the wicking effect, but you can't really stop it. The result is a wire that breaks after a few flexes. If you have crimped properly, forming cold welds between a clean wire and a clean terminal, the joint will be more reliable that solder by some orders of magnitude. If you crimp poorly, and just squeeze everything together, don't expect good results.

+1
You nailed it perfectly coppice, a properly done compression weld can't be beat.

[penfold]

Hi,

For short term or regularly maintained equipment then its really not so much of a bother, additional soldering would guarantee a better INITIAL joint.

However, the idea behind crimping is that it should be done so that there is sufficient pressure between all of the metallic surfaces that it slowly 'cold welds'. This is not an instantaneous process and takes a little time for the processes to happen.

If you add solder to the voids (we'll assume that the actual metal-metal contact region is so tight that no solder will leach in to it), then as the assembly heats and cools (and thus expands and contracts) the solder will provide a region which expands for tends to separate the nice clean metal-metal region or disturbs the natural movement of the metals.

Crimping contacts eliminates the use of solder which will wick up wire strands and provide an excessive rigid patch which degrades the flexibility and strain tolerance of the wire. The crimp contacts are of course higher maintenance and require more cleanliness whereas the solder is just more forgiving but requires more strain relief, the world seems to have done in the direction of crimping because they are better suited in batch production scenarios and so connectors are optimised for crimped terminals and so might not provide as much strain relief where necessary

[German_EE]

Look at it this way, the wire connections in a spacecraft (which is probably the highest vibration environment you will find) are crimped and not soldered. Crimping works if you use the right tools and is ultra reliable, soldering will decrease this reliability.

[Circlotron]

Maybe the worst thing you can do is tin stranded wire with solder and then crimp it. (or clamp it in a screw terminal for that matter). It will definitely loosen.

[calzap]

For building wiring in the US and (I think) Canada, connections must be mechanically strong without soldering.  Then solder can be applied, but there is a risk an inspector won't pass it because once solder is applied, it may be impossible to know whether the connection would be strong without it.  In most jurisdictions, it is OK to use exothermic welding to join wires to ground rods or Ufer grounds; however, I've seen pictures of exothermic welds that cracked long after the work was done and inspected.

I'm not sure what rules UL or other certifying organizations apply to connections of free-standing, mains-voltage (or higher) wires inside electrical devices.   I've seen only crimp connectors or wire nuts without solder.

Mike in California

[Fraser]

In the automotive industry, crimp is king. BUT you do have to to watch out for moisture ingress causing corrosion and intermittent connections. I have also noted that the heavy load cable terminations in buildings are crimped so there must be confidence in the crimp method when subjected to very high currents. No solder to be seen on those connections.

Fraser

[CatalinaWOW]

I think Penfold nailed the fundamental reliability issue.  Flexibility and insulation effects are real, but not critical in all environments. 

[C]

Wire_wrap is a type of compression connection
https://en.wikipedia.org/wiki/Wire_wrap

Wire wrap construction can produce assemblies which are more reliable than printed circuits: connections are less prone to fail due to vibration or physical stresses on the base board, and the lack of solder precludes soldering faults such as corrosion, cold joints and dry joints. The connections themselves are firmer and have lower electrical resistance due to cold welding of the wire to the terminal post at the corners.

Note that a proper wire_wrap connection has many compression points, and maintains pressure on the  compression points.

Insulation-displacement connector
[urlhttps://en.wikipedia.org/wiki/Insulation-displacement_connector][/url]

When properly made, the connector blade cold-welds to the conductor, making a theoretically reliable gas-tight connection

Notice that a lot of automotive connectors add some sealing on connector assembly.

A gas tight compression is best, followed by compression & sealing, then compression. 
Solder is way down the list. And as others have stated will increase the failure rate of a compression connection.


C

[nctnico]

Soldering after crimping is not done. Actually a solder joint is several magnitudes worse than a crimped connection! A solder joint will fall apart under vibration/mechanical stress.

[DanielS]

A solder joint will fall apart under vibration/mechanical stress.

An unsupported crimp will fail on fairly short notice too.

[CatalinaWOW]

In the military equipment industry both solder and crimp joints are used.  I don't actually know which is more reliable, but when properly implemented both can meet standards for life in extreme shock and vibration environments.  The difference in reliability may be moot.  Kind of like designing a heart pacemaker for a 200 year lifetime, such a device's life will never be exercised.   

[nctnico]

Run a high current through a solder joint and a 6.3mm spade connection and notice the difference. There is a good reason why they use anything but solder joints for wiring in equipment which needs to have a long service life.

[CatalinaWOW]

Run a high current through a solder joint and a 6.3mm spade connection and notice the difference. There is a good reason why they use anything but solder joints for wiring in equipment which needs to have a long service life.

Makes sense to me that solder joints are less satisfactory for high current applications.  Don't know if I would generalize to say that solder is not satisfactory for long service life in applications where resistive heating of the joint is unlikely to be significant.

As an example, in radios manufactured in the 1920s and 1930s, failing solder joints are way down the list of what is found wrong.  The list runs through bad tubes (valves), failed or drifted capacitors, failed or drifted resistors, bad sockets, failed insulation on wires and somewhere well after that solder joints.  90-100 year service life seems adequate for most applications.

[coppice]

In the military equipment industry both solder and crimp joints are used.  I don't actually know which is more reliable, but when properly implemented both can meet standards for life in extreme shock and vibration environments.  The difference in reliability may be moot.  Kind of like designing a heart pacemaker for a 200 year lifetime, such a device's life will never be exercised.

The military does indeed use a lot of soldering and a lot of crimping. They are, however, very particular about where each is used. You would, for example, never be allowed to solder to a multi-strand wire.

[donmr]

I agree that a good crimped connection is the best, but that assumes it was done with the proper equipment, a trained operator and that everything is calibrated and periodically tested.  For home hobby applications I don't usually have all of that and I tend to solder a lot of connections after crudely crimping them.

[nctnico]

In the military equipment industry both solder and crimp joints are used.  I don't actually know which is more reliable, but when properly implemented both can meet standards for life in extreme shock and vibration environments.  The difference in reliability may be moot.  Kind of like designing a heart pacemaker for a 200 year lifetime, such a device's life will never be exercised.

The military does indeed use a lot of soldering and a lot of crimping. They are, however, very particular about where each is used. You would, for example, never be allowed to solder to a multi-strand wire.

And appearantly for a good reason. Over the years I have restored several HP power supplies from the 80's and in many cases I could pull the multi-stranded  wires (including those carrying mains!) from the boards without much force. Part of the restoration is to fixate the wires with hot-melt as a strain relief after resoldering them.

[DanielS]

Run a high current through a solder joint and a 6.3mm spade connection and notice the difference.

What do you use to secure your spade-to-PCB connectors to the PCB side of the connection? Most likely solder.

If soldering was a no-go for high-vibration environments, there wouldn't be a ~300 pages long NASA standard about how to tin wires and soldering in general for aerospace applications.

[mos6502]

If properly done, reliability of crimp and solder joints is equal.

E.g., check this out:

"Solder-style splices produce a smaller, more compact splice
termination, with significant weight reductions over crimp-style splices."

http://www.hq.nasa.gov/office/codeq/doctree/87394.pdf

Would they do that if it was unreliable?

[eneuro]

http://www.hq.nasa.gov/office/codeq/doctree/87394.pdf
Would they do that if it was unreliable?

Thanks, for this nice NASA document 

Anyway, for 230VAC mains on welder transformer primary I will have a chance to debunk this solder "disaster" myths, since one of connectors (5mm wide ~2.5mm2 strand wire) is only crimped and protected with thermal heatshrink, while another one like this below was the same of course strand wire, crimped, added thermal heatshrink, then from front a little bit of 1mm solder applied, but in a very controlled way-I mean lets say 5mm-10mm, so while it was heated from connector side, copper wire conducted heat and solder penetrated inside strand wires througth crimped connector, but of course we have full controll how much solder will go inside and while solder like heat it will fill connector side better than opposite side, i guess:



We'll see what happends at currents between 15A-20A (below home mains grid circuit breaker 25A), but I do not expect any harmfull flex and huge vibrations and I do not know, it looks like, while solder filled empty space between strand wire, shouldn't it create even lower resistance, assuming there wer eno extreme solder temperatures and whol eprocess was only a few seconds long? 
Visual inspection of this connector seams to better protect inside wire from external erosion, than when no solder joint is created from connector side... 

Maybe, to debunk this theory, will be better put such terminated wires on welder transformer secondary, than slowly increase welding current and see which fail sfirst at which current level, but I exepct that strand wire itself fails first rather than this connector with added small amount of solder 

Update: Probably not best example of controled solder filling, while seams to me that mayb ea little bit too much solder was added, but idea remains the same... other connectors made thi sway doesn't have this bulb where connector crimp ends-there were too much solder , I guess...

[KD0CAC John]

There are multiple reasons for doing or not doing something but when it comes to electrical connections , I ask the question is any chance of regular movement / flexing .
If so , then crimp , if not - soldering is OK .
This is from 50 + yrs. of automotive , truck , motorcycle etc.
What I have seen happen is there is movement , vibration or flexing , a soldered connection has a high failure rate from metal fatigue at the point of the solder .
I did a lot of mobile work and learned to try to reuse some crimp connectors battery and others , many times the dimple crimp for the most common type of crimp connector , would cut some of the strands when you get just enough pinch in the dimple to keep from pulling apart .
Then when computers enter cars I saw for the 1st time what I call a double row crimp , this is a true die where one half of the die is a half round and the other side of the die is made of 2 half rounds that are closer than half round and have a sharp point  between that does not reach all the way through that half of the die , this sharp point is meant to meet the seam of a crimp connector and each grabs and raps around half the strands .
After the lack of a picture and not quite 1,000 words
I have not seen one of these fail - if done correctly .
I ended up buy off the Snap On truck a 3 foot long crimper like this for battery connection , and welder cable .
I uncrimped a few of these by putting the crimp fitting in a vice and used a small punch to spread the 2 rap around parts open and with none of the strands captured by any of the crimp , I found that the some of the strands were acutely fused to the surface of the open crimp fitting .
So my reasoning for having enough crimpers of this style to do almost all sizes of wire and cable .   

[AF6LJ]

If properly done, reliability of crimp and solder joints is equal.

E.g., check this out:

"Solder-style splices produce a smaller, more compact splice
termination, with significant weight reductions over crimp-style splices."

http://www.hq.nasa.gov/office/codeq/doctree/87394.pdf

Would they do that if it was unreliable?

I use to have a hard copy of that document, thanks for posting that...
Back into the library it goes.

The largest single problems with crimping are;
The wrong terminal is used.
The wrong tool is used.
The technician / assembler is not properly trained in the use of the tools or preparation of the wires to be crimped.

As someone who has crimped a few miles of connections, it is in many cases as hard or harder to master than soldering.

[DanielS]

From what I remember of the NASA document I have skimmed, the main problems with solder connections are copper embrittlement from using the wrong type of solder or flux for the connections being made and contaminated solder.