Best Soldering Helping Hands

[sahib]

You probably don't want a bend in those power resistor legs.
The legs on that resistor type tend to be quite thick. Bending them is likely to introduce stress or fracture internal to the component. Additionally, they often come with short legs (that may not even need trimming after soldering depending on clearance to adjacent chassis / pcb etc.
This makes bending difficult if not impossible - in reality it becomes more an issue of pushing the leg into a bend rather than 'bending'.

I am not sure if you aim that to me but I am against lead bending.

Yes - they will 'desolder themselves' if there's a fault that causes massive overheating and , if on a vertically orientated pcb, fall out.

They certainly will. But that will be because of bad design, or very exceptional circumstance at which falling off the power resistor will be the last of one's concerns, and bending the leads will not save the day. If the bonding of the solder node is compromised due to overheating then the electrical connection on that node will be compromised too and create fault whether the board is oriented vertical, horizontal or upside down.

I'll spare people the anecdote but will add that it wasn't my design !

I did not get that. Did your company have anything to do with that (Soundcraft) amplifier design?

And where mechanical stability is an issue then deliberate mechanical solutions may be preferred eg the 'kinked' legs common on larger electrolytic capacitors. And there's always the glue/mastic/gunk solution (although I have a personal aesthetic aversion to it).

Indeed, but it still does not absolve it from the fact that it is a risky solution rather than a safe one, perhaps preferred due to cost effectiveness. If I was in charge, and was able to chase off the accountants, I would specify gluing (lucky that I work for myself and do not have accountants breathing over my neck.)

 

[Newmarket]

I am not sure if you aim that to me but I am against lead bending.

No. I understand your viewpoint is "anti-bend". I could have been clearer. But basically just discussing around that and pointing out that some components just aren't suited to leg bending in any case. These tend to be the more "Power" type components with thick leads so any "pro-bend" argument wrt heat dissipation, solder resistance etc basically falls away. eg pcb mounting transformer bobbins, Bridge Rectifiers, beefy thyristors etc.

They certainly will. But that will be because of bad design, or very exceptional circumstance at which falling off the power resistor will be the last of one's concerns, and bending the leads will not save the day. If the bonding of the solder node is compromised due to overheating then the electrical connection on that node will be compromised too and create fault whether the board is oriented vertical, horizontal or upside down.

Yes. I'm not suggesting that bending the leads would be a good thing her (apart from the specific case being the type of resistor illustrated with short rigid legs). Quite the opposite in fact. If it hadn't dropped out of the pcb then things would have gotten very hot indeed and more difficult to diagnose.
In the (real) case I'm talking about the fault was due to a failure of an electromechanical relay such that the resistor was not being bypassed as required in an inrush current limiting scheme.

I did not get that. Did your company have anything to do with that (Soundcraft) amplifier design?

No - I was referring to the circuit with the failed relay mentioned above.
Not meaning to be cryptic but guessing limited interest in the PFC mains input stage of an OES spectrometer.

Indeed, but it still does not absolve it from the fact that it is a risky solution rather than a safe one, perhaps preferred due to cost effectiveness. If I was in charge, and was able to chase off the accountants, I would specify gluing (lucky that I work for myself and do not have accountants breathing over my neck.)

Not sure what you are referring to as a "Risky Solution" here - kinked legs ?
'Glue' or 'Mastic' solutions are interesting. One thing I've had to do in the past is stop people using hot melt glue guns with glue sticks where the adhesive goes 'hard/brittle' over time/heat and easily detaches from the component /FR4. But its tempting for people to go for that when alternatives are seen as "messy" or expensive eg

https://uk.farnell.com/loctite/382-20g/adhesive-382-tak-pak-bottle-20g/dp/1676330

 

[warpie]

This is one of the most important tools on my bench.

I also ned to invest in one of these. The older I get the most necessary it becomes...

Luckily (?) I only deal with THT for now,

 

[sahib]

Not sure what you are referring to as a "Risky Solution" here - kinked legs ?
'Glue' or 'Mastic' solutions are interesting. One thing I've had to do in the past is stop people using hot melt glue guns with glue sticks where the adhesive goes 'hard/brittle' over time/heat and easily detaches from the component /FR4. But its tempting for people to go for that when alternatives are seen as "messy" or expensive eg

Apols. My turn. I should have paid more attention. There is a difference between kink and bend. So, I suppose a little kink to keep the component in place would be acceptable when it is appropriate.

However, I have taken the opportunity today and did some tests which I hope bring something tangible results into the discussion.

Test1:

This involved soldering a capacitor of 20mm diameter and 40mm height with leads. The lead diameters were 0.8mm.. The aim was to subject the capacitor to an external impact using a hammer.



I have not even done a proper Pete Townshend on it. Just a modest whack. However, an average person can exert a force of around 10-20Kg when leaning over a, say, table. So, we can take an average and assume that I subjected the capacitor to 15Kg force (though I should express that vectorially to be proper).

The result was that the capacitor electrodes/leads gave way before the solder nodes. The solder nodes did not even register that.



Test 2:

This involved soldering (straight) a piece of copper wire of 0.8mm diameter to the same proto board to represent the capacitor lead. The aim was to estimate the lead and solder node strength and to see which one would give way first.



As can be seen from the picture the solder node is a standard 2mm diameter pad with 1mm plated through hole.

I clamped the proto board to a vice so that I could get a grip. I tied the open end of the copper wire to a weight of 13Kg (box of power transformers), and attempted to lift it to see if the solder node would give way. The copper wire snapped before even I could lift it. The solder node did not even yawn.



From this experiment we can establish that bending a component lead has no relevance to component stability. As I mentioned before the weight of the components in comparison to the bonding strength of the solder node does not even enter the realm of significance.

Below is for Abbey..

The camera I have is called Enhanced Vision. It is an older version as the video connection is through 15 pin D for VGA. But performs superb. You can select the image type, colour, B&W, or negative and other few options. The small Sony monitor is again given by a close friend. It is old but razor sharp.



It comes mounted on a right angle bracket with a swivel head. Initially I had mounted it on a piece of laminated MDF board. But later on another close friend, who is a photographer, gave me a proper stand with backlighting and railhead for fixing the camera. At the moment the camera is still on its old stand but I will be making a suitable bracket to fix it onto the actual railhead..





Mega thanks again to my good friends Paul (camera), George (monitor) and Donald (stand).

Finally, I though this might be of some interest. This is one of my workbenches. I have another identical one where I have my computers and the PCB inspection bench next to it.

 

[moamps]

From this experiment we can establish that bending a component lead has no relevance to component stability. As I mentioned before the weight of the components in comparison to the bonding strength of the solder node does not even enter the realm of significance.

What you have shown by your experiment is that a straight lead newly soldered in a metallized hole will withstand considerable tensile force. But you didn’t even get an approximate value of what the maximum tensile force is. You didn’t experiment with bended lead to be able to compare it. So you can’t conclude that bending lead has no significance in stability because you haven’t even tested it. If you really want to play with it, try to find a resistor on a single side PCB that has been heating to 70-80% of maximum dissipation for ten years and repeat the experiment.

It is clear to me that many do not like to repair PCBs where components have bent pins, I also often do not bend pins in my projects, but to claim that this is bad is plain wrong. Maybe someone designs their PCBs in such a way that the soldering points on the PCBs are small in diameter, that the ground plains are placed too close to the pins and that any bending of the pins can lead to an unwanted joint. This should then be emphasized to the person soldering the components, not claiming that this is a general rule.

The following figures show excerpts from NASA Student Handbook for Hand Soldering.



Apart from the fact that the pins can be bent, it also shows that the excess lead is cut to the exact prescribed height BEFORE soldering.

Some will say these instructions are old-fashioned etc., but maybe that’s one of the reasons some lucky ones are still chatting with Voyagers.

 

[moamps]

Here is one PCB from SONY RGB video monitor that can serve as a curiosity because the solder space for bending the leads of capacitors is provided on the design of the PCB in soldering mask.

 

[sahib]

I am sorry and I do not mean to be rude or offensive but you are hanging on to your false claims just for the sake of hanging on. What you are saying is utterly a nonsense.

In terms of the stability of a component soldered onto a PCB, there are two initial joints come into play. There will be a third one for components like say inductors but we are primarily considering resistors and capacitors here.

1. The joint where the lead is bonded to the component.
2. The joint where the lead is bonded to the PCB. Here I mean straight lead bonding/solder.

The experiment carried out proved my point that the stability is dictated by the joint where the lead bonds to the component, as this is the weakest point. The joint on the solder node is irrelevant. You can bend the lead beyond the straight solder node, put a chain on it or tie it to a bollard if you wanted to. It will make zero difference.

The experiment also proved that, in the event of some miracle that the bond on the component side was able to withstand the force, then this time the lead will break. Again, anything beyond the straight lead solder node is irrelevant.

If you are not able to accept that then, as I said you are dragging this thing on probably because of your pride but I have no time for that. If you can not see it then you must be blind.

You have already been rude by saying that your claims come from your years of experience of dealing with electronics while ignoring the experience your opposition may have, or even not acknowledging when pointed out.

 

[moamps]

I am sorry and I do not mean to be rude or offensive but you are hanging on to your false claims just for the sake of hanging on. What you are saying is utterly a nonsense.

Certainly, I spent some time studying how scientific experiments are performed and obviously I learned nothing.

You have already been rude by saying that your claims come from your years of experience of dealing with electronics while ignoring the experience your opposition may have, or even not acknowledging when pointed out.

I'm sorry, but it's not clear to me how mentioning my experience means rudeness to you? Can't I mention it?
And I mentioned it only because this experience is not only from the audio field, but also from the field of telecommunications and industrial electronics, where, very very often you can see designs where the components have bent leads.
And I did not allude to your experience, I'm sorry if you understood it that way. For the rest of what I said, these are all arguments and I have nothing to apologize for.

BTW, at the moment I have a Liebert UPS on my desk that is full of components with bent legs. And I’m not optimistic that I’ll be able to fix it. I will end up beating him up with this, in the name of science.

 

[abbey road d enfer]

OK, I think you two should bury the war hatchet. The divide won't move. We all have out favorite way of doing things, that may be criticized by others. As the saying goes, there are many ways to skin a cat.

 

[sahib]

All good on this side. This in no way would diminish my respect for Moamps and I enjoy his posts.

However, I have actually done one more test (I can here you all go, oh no!). I know the forum is generally about electronics design but hopefully this discussion will have some value.

In the previous test you may have seen from the picture that I wrapped the wire round the white cord that held the weight (box of transformers). You may have also noticed that the wire snapped literally from the beginning of the bend. The reason for this is that, by bending the wire we introduced fracture to the crystal structure of the copper wire. So, strictly speaking there could be an argument that the test was not complete.

So, this time I clamped the wire to an aluminium link. The other side of the link where the cord was tied to carry the weight. Now the wire is completely straight with no fracture, that at least we know of. The applied weight was again 13Kg.

This time the wire snapped on the PCB side and immediately before the solder node. You can see that in the below picture. The tip of the wire is elongated. This is a typical signature of what happens before the material breaks. It stretches. It might be that by coincidence this point turned out to be the weakest in wire's structure.

This again supports the previous argument.

Incidentally, in the case of a resistor with bendable leads, we introduce the same fracture to the lead structure when we bend it to solder into its position. So, assume that the resistor itself and the lead joint at the resistor side were able to withstand the impact, the lead would snap.

But in any case the bonding strength of the solder node is vastly above the breaking point of the component lead.

 

[Newmarket]

What you have shown by your experiment is that a straight lead newly soldered in a metallized hole will withstand considerable tensile force. But you didn’t even get an approximate value of what the maximum tensile force is.

Although reluctant to prolong this discussion...I think there's a point being missed with the way the discussion has turned.
The "Pull Test" discussed is basic tensile testing yes (and fwiw I used to work in that field although it was applied more to stuff like steel bars and textiles). But I don't see that it particularly relates to modes of failure in electronic equipment (audio or otherwise).
ime mechanical failure is likely to be due to vibrations over time - the obvious being transportation or, to get audio specific, a mobile facility but also from acoustic sources eg amplifiers in active loudspeakers. Leading to weakening / fatigue over time until there's an open connection.
Environmental tests for those hazards require an different method - eg "Shaker Table".
I'm staying neutral on lead bending