Methoods for connectig sensitive components

[janaf]

Components like high stability resistors, references, may suffer when soldering, especially Pb-free solder. I've been thinking about what alternatives there are for connecting sensitive components in circuits. Methods that come to mind are

• Low temperature solder (like Bismuth-alloys)
• Silver / gold filled conductive epoxy
• Mechanical press
• Screw terminals
• Spot welding
• Laser / high energy welding

I have not used any of the ones mentioned below for sensitive stuff on PCBs.

Any opinions / experiences / thoughts on the topic out there?

[tautech]

Quality components have datasheets with thermal constraints, abide by them.

Solder is still the best for reliability, ease of production and future rework.

[janaf]

I was thinking of ppm levels, when component characteristics can change by tens or hundreds of ppm after soldering and show thermal shock / aging behavior that take days/moths to recover. There the solder constraints of data sheet are not of much of help.

[German_EE]

I have seen some temperature sensitive components use crimp connections, there are also screw terminals and wire wrap to consider.

[janaf]

I have seen that Peltier elements are manufactured with leads soldered on using low temperature bismuth type solder. I guess the elements simply don't stand normal soldering temperatures.

[SeanB]

Considering they are made from Bismuth alloys using the solder is the right thing, the only other alternative would be a conductive epoxy.

[Andreas]

Hello,

In AN86 of LT the good old Cadmium (60%) solder is recommended.
Which should also have a much lower melting point.
And has a very low thermal EMF against copper.

But unfortunately the damps during soldering are highly toxic.
So no way to handle without professional equipment.
And since RoHS cadmium is on the list of forbidden substances.

By the way: is there any Bismuth solder with low thermal EMF?
I think there should be a RoHS compatible replacement for the cadmium solder.

Conductive epoxy is no good solution in my eyes at least for precision resistors:
they would mainly contribute to T.C. and ageing effects.
Screws from which material (thermal EMFs)?

So for long term stability I can only imagine welding or soldering.
For hobby soldering I would prefer the good old leaded solder in this case.

With best regards

Andreas

[branadic]

Maybe spot welding could be an option. We use spot welding to connect cables with pcb traces instead of soldering. But there are several requirements to the coating of the copper wire. I.e. silver coating works great.

Regards, branadic

[janaf]

Cadmium, I won't go near it! Stays in your body for the rest of your life. Bonds to the bone and makes bone brittle. There was a big disaster in Japan decades ago, called Itai-Itai (It hurts, It hurts!)

I have not been able to find thermal EMF data on Bismuth solder. No low TEMF solder replacements for cadmium that I know of. I would appreciate if someone has more info on this.

The only substitute I know of is to put leads of a components close to each other, well thermally coupled, ie thermally well designed pcb layout.

...good old Cadmium (60%) solder is recommended.

By the way: is there any Bismuth solder with low thermal EMF?
I think there should be a RoHS compatible replacement for the cadmium solder.

Andreas

[Zad]

Generally speaking, the fewer different types of material you can use, the better and more predictable the component's performance will be. I would imagine that you would want to use the same mixture of solder as the manufacturer used to tin the component terminals. Welding can expose the component to very intense but localised mechanical and thermal shocks, I think I would avoid that if at all possible. I imagine slow thermal and mechanical "deltas" in fitting are preferable over short, sharp ones - that way fewer stresses are generated.

Depending on the component design, it may be possible to heatsink or mechanically isolate the active part of the component, so that only the relatively inert terminals are stressed.

To echo the comments above though: Read the manufacturer's information, make sure you understand it, and produce a good quality design that adheres to sound basic principles. If it is an important design, you may wish to contact the manufacturer to see if they have any examples of recommended designs.

[quarks]

For most tests I just use sockets.
For soldering I use a heatsink if possible, betwen solderpoint and component body.
Eutectic solder Sn62 Pb36 Ag2 or Sn63 Pb37 should be the best/easiest way to go for.
So far good old Sn60 Pb40 worked good for me.

[tautech]

For most tests I just use sockets.
For soldering I use a heatsink if possible, betwen solderpoint and component body.
Eutectic solder Sn62 Pb36 Ag2 or Sn63 Pb37 should be the best/easiest way to go for.
So far good old Sn60 Pb40 worked good for me.

I bought a 1 pound reel of Sn62 Pb36 Ag2 from RS years ago for it's low melting temp.
It was something like $50 at the time...bloody expensive. 

It is brilliant stuff, wonderful wetting qualities and a mild flux with low amount of residue.
But not to be wasted on any dirty repair work as the flux is not active enough.
Still got a bit left that is saved for the right job. 

[-jeffB]

What about indium? It's not strong, but it's sticky -- you can solder stuff onto glass with it. Pure indium melts at 157 C, and its alloys can melt at much lower temperatures. Alloyed with gallium and tin, it can stay liquid down to -19 C. It's also a lot less toxic than lead or cadmium.

It attacks aluminum, so it's unacceptable for many uses, and it's expensive (more expensive than silver). But in a specialized application like tacking down extremely sensitive components, its stickiness, low melting point, and RoHS compatibility would seem to make it a good candidate.

[ruffy91]

For leaded components you can attach a heatsink (crocodile clip) to the lead so that the the soldering heat flows into the heatsink instead of the component.

[Pjotr]

For hand soldering Sn62 Pb36 Ag2 will do fine. But use solder as thin as you can get and a real fine soldering tip. The thinner the solder wire is, the less heat you need to melt it. I have a roll of 0.35 mm Edsin Sn62 Pb36 Ag2 for critical occasions although it does pretty well for other SMD rework also.

[janaf]

This was not about regular soldering of regular components, but very sensitive components.

I bought some Vishay hermetically sealed resistors recently, rated 0.2ppm/C, drift should be typical 1ppm/year. These can cost $100+ each so I don't want to degrade them by soldering if it can be avoided. Vishay do not suggest a method that maintain the specifications. I have asked. I guess they do all their testing with temporary clamps. 

Pleas see this thread from which the quote below is from,
https://www.eevblog.com/forum/projects/t-c-measurements-on-precision-resistors/

"I received my VHP202Z back from Vishay Precisison/ Germany, he had heated them asymmetrically to 125°C, and they were stuck at +5ppm higher resistance, until I unloaded this memory / hysteresis effect."

"Dr. Frank reports 5ppm hysteresis for a 25 deg to 125 deg excursion for a hermetically 10K VHP202Z which is the hermetically tight upgrade of Z201. Others report soldering drift (hysteresis?) of even 50 ppm."

I have got Kester SN62PB36AG02 down to 0.25mm (0.01), great for small parts.

[free_electron]

1 word : forceps

simply clamp a pair of forceps on the wire between the solder joint and the component body. The mass of the forceps  will suck away enough of the heat so it doesn't reach the component.

[quarks]

Here is what I use

https://www.eevblog.com/forum/projects/vishay-bulk-foil-drift-after-soldering/msg445317/#msg445317

[janaf]

Quarks and free_electron, thanks for practical advice! No offence to others but the "you can try this" ideas are not very helpful in this case.

I have been searching a bit for Bismuth based solder but not really found much except they should NOT be used with lead present as those form brittle alloys that easily peel off, i.e. bad bonding. No EMF data for bismuth solder yet.

But I have realized that with hermetically sealed packages ie kovar leads, it's pretty meaningless to search for low EMF solder because the EMF kovar - copper is so bad anyway, thermal management seems the only cure.....

Here is what I use

https://www.eevblog.com/forum/projects/vishay-bulk-foil-drift-after-soldering/msg445317/#msg445317

1 word : forceps

simply clamp a pair of forceps on the wire between the solder joint and the component body. The mass of the forceps  will suck away enough of the heat so it doesn't reach the component.

[janaf]

I got a non-reply from a vendor on Bismuth solder EMF, they did not have any data.