Maybe OT: Why so much resistance to SMD?

[Ice-9]

Heres a nice little hands on demo using the minimum of tools to do a great job.

http://www.youtube.com/watch?v=0OIMmQkAuDQ&feature=related

Tweezers can be used to place the components and a cheap gas hot air iron from maplins is what i use instead of the expensive soldering iron and pre heater in the vid.

[markeebee]

I'm tempted to go to SMD, just because it's a different bunch of skills to learn and therefore might be interesting and rewarding.

Now I'm guessing, but on a normal stompbox size PCB I might save 10 minutes in production time?  And maybe a quid in components?

So if I make 30 boxes a year, that's a saving of about 2.5 hours and thirty quid?  But I'd need to stock up with SMD components and buy a hot air iron.

If I was only interested because of the potential savings, I don't think it stacks up.

[~arph]

Well, you can fit more in a box, that for me is the biggest benefit

[slacker]

Yeah depending on what you build i can see that either getting more in a box or being able to use a smaller box is an advantage, but for a lot of pedals through hole is small enough and things like the number of controls dictate the size of the box more than the size of the board does.

I'd agree with Mr Bee, for DIY builders the potential savings aren't that big, and through hole componets are cheap enough to start with. For most pedals the cost of the components you'd convert to SMD is only a small proportion of the total cost of the pedal, the off board hardware makes up the vast majority of the cost so per pedal, or for small amounts of pedals, the saving's not significant.
As this is esentially a leasure activity, I don't see any advantage to saving time.

[defaced]

You don't need hot air to solder SMD.  This is how I do mine and it's very very easy and I don't need an extra tool on my bench.  http://www.sparkfun.com/tutorials/101

[~arph]

That looks to me like a recipe for frying your board and components.. but I'll learn this soon enough as I will probably do the same soon. 

[DougH]

why do we gotta get all hung up on the rules , man? can't we all just do our own thing and hang out in the park like the old days?

The parks are much smaller nowadays, so SMD techniques are essential if you want a decent go on the swings.

Just make sure you epoxy your butt to the swing and you'll be okay.

[defaced]

That looks to me like a recipe for frying your board and components.. but I'll learn this soon enough as I will probably do the same soon.  

Working fast is an important thing, so is a very hot iron.  Funny, I think the same thing about hot air and ovens.  There's examples on that site where the technique with a hot air station de-laminated the FR-4.  The whole hot air/oven thing just seems totally counter intuitive to what I've been taught and experienced, but hey, it obviously works because cause that's how it's being done.  I think the important point is that as long as you control the max temperature and the time spent up at elevated (melting) temperatures, you can use whatever heating device you want.  

[frank_p]

- I'll gather that Doug's comment is that you shouldn't rely on the mechanical properties of the joint for support of the component.  Yes, brazed and soldered joint have mechanical properties.

Don't know why I didn't hear it that way...  What I understand now, is that it is not a good practice to count on those for holding everything: that's the PCB's job.

I understand the point of view now.  For brazing teeth on endmills : it's not the same *preoccupation* than securing a chip to the copper layer. A similar science might be behind, but the goals are not the same.

- The diffusion part of the process mentioned in that paper is caused by time, temperature, pressure, and alloys.

Yes

- If you remove one of those, namely time and pressure in our case, diffusion does not take place.

I am not sure that you have to melt the copper to have ion diffusion of copper into the brazing alloy.  The copper becomes soluble, it does not melt completely.  In brasing, the base metal never melts anyways.  And the interface exists.  If the interface of the copper and the tin solution exist at the eutectic point then the copper can become fluid in the tin since the eutectic point is the condition at which two (or more) metals can begin to coexist in liquid form.

- Assuming you're working with eutectic solder, which melts/solidifies at ~175c, the Cu-Sn phase diagram at that temp shows four possible phases.  However, the copper never melts.  Phase diagrams shows what happens when you take an alloy from melting and cool it under equilibrium conditions to whatever temperature indicated. 


Yes, you are describing a normal way of forming the alloys crystals from the liquid alloy.  The eutectic is the lowest temperature where there was alloy in liquid form while cooling down.  This is the standard way of describing the phase diagrams.

- It says nothing about when you take a solid (Cu) and a liquid (Pb&Sn) and warm it up.

It says at what temperature some fluid alloy can exist at the interface of the solid metals phases at equilibrium temperature.  Because they are in contact with each other at the interface copper is soluble in Tin.  It does not say that copper melts by itself.

- Thus, no intermetallic phase can be formed during our soldering operations.

Their presence is not to the advantage of the solder if the diffusion layer is too big : mechanical or electrical wise : these alloy phases seems unwanted and a problem mainly in Pb free solder.

http://www.metolit.by/imc_ysb/txt38_0.php

Scroll down to (the page is pretty long) :

EVOLUTION OF MICROSTRUCTURE IN THE BOUNDARY OF SOLDERED JOINT AFTER THERMAL INFLUENCE

The paper talks about annealing a SnAG/Cu joint at different temperatures and for different lapses of time.  OK, OK... there is some temperature and time involved, but not THAT high for both.

- The solder cracking paper is discussing what happens in the solder joint itself, and the addition of elements to the solder which form intermetallics within the joint - not forintermetallics between the joint and the base metal - to prevent cracking.
It's a grain size issue, same thing happens in welds, particularly with ferritic stainless steels.

http://www.absoluteastronomy.com/topics/Solder
Search for section :
Intermetallics in solders

Also:

http://www.dfrsolutions.com/uploads/publications/2001_IEEE_Kinetics_IM_Alloys.pdf

*While the frequency of use of no-Pb solder continues to
increase, electronic solder joints are still generally fabricated
using Pb-Sn eutectic solder reflowed in contact with the
metallic surfaces to be joined. Upon reflow new intermetallic
alloys grow at the interface solder/pad and ensure bonding.
For instance for Cu metallizations, Cu6Sn5 with some Cu3Sn
forms at the Sn/Cu interface [1,2,3]. If these alloy layers are
too thick, they have deleterious effects on the mechanical
reliability of the joints [4].  Therefore, over the years, the
composition of electronic leads has changed from a single
metal, usually Cu, to multi-layered structures with two or
more metallic layers stacked on top of each other. A thin
coating of a noble metal such as Au or Pd is frequently used to protect the surface from oxidation, while a second layer, commonly Ni, is used as a diffusion barrier to prevent the Cu
underneath it from interacting with solder. A number of
investigators have shown that Ni/solder intermetallics grow
more slowly than Cu/solder intermetallics during reflow
[5,6,7].*

------------------------------------------------------------------------------------------------------------------------------
http://en.wikipedia.org/wiki/Brazing

*Interaction with base metals

For successful wetting, the base metal has to be at least partially soluble in at least one component of the brazing alloy. The molten alloy therefore tends to attack the base metal and dissolve it, slightly change its composition in process. The composition change is reflected in the change of the alloy's melting point and the corresponding change of fluidity. For example, some alloys dissolve both silver and copper; dissolved silver lowers their melting point and increases fluidity, copper has the opposite effect.

The melting point change can be exploited. As the remelt temperature can be increased by enriching the alloy with dissolved base metal, step brazing using the same braze can be possible.

Alloys that do not significantly attack the base metals are more suitable for brazing thin sections.

Nonhomogenous microstructure of the braze may cause non-uniform melting and localized erosions of the base metal.*

-----------------------------------------------------------------------------------------------

It seems diffusion is necessary for bonding but not too much of it.

Using non-trough holes SMD components and lead free solder seems much less rugged than leaded ones and using standard Pb containing solder.

[amptramp]

Are people still using gold to prevent oxidation on component leads?  I thought the experiences we and everyone else had on spacecraft would have deep-sixed that long ago.  We had 2N2222A transistors with leads that were gold over kovar with the leads rusting away due to the porous gold layer forming an electrolytic cell with the kovar.  Add to that, the gold forms intermetallic compounds known as "purple plague" because of its appearance.

I don't foresee much need for SMD in stompboxes except where components like microcontrollers may only be available in SMD.  The main savings is not having to drill holes in the board as holes are expensive - the board wears down drill bits quickly, especially with epoxy-glass boards.

[CodeMonk]

I did alot of SMD back in the late 80s while I was working at JPL.
We soldered all that by hand with the aid of a microscope (the typical one you might find in a high school lab).
I had no problems with it, but it does take some time getting used to.
We didn't have hot air irons, just really small soldering iron tips. Although I modified mine to heat each side of an SMD component at the same time.

Would I build pedals, etc. with SMD stuff today. Maybe if I had a microscope. Eyes aren't what they used to be.

[DougH]

Yeah, we have a microscope in our lab too. Not for  doing smd but for inspecting boards and other hw, and doing detailed work. One of my issues with doing anything super tiny anymore is the eyesight.

[DougH]

continuing last post (on the phone)- I don't see a lot of practical use for smd in analog stompboxery. Seems to be an attitude of let's do stuff just to do it. Fiine for learning but I don't usually grab a hammer from the toolbox unless I really need one.

[defaced]

For successful wetting, the base metal has to be at least partially soluble in at least one component of the brazing alloy. The molten alloy therefore tends to attack the base metal and dissolve it, slightly change its composition in process. The composition change is reflected in the change of the alloy's melting point and the corresponding change of fluidity. For example, some alloys dissolve both silver and copper; dissolved silver lowers their melting point and increases fluidity, copper has the opposite effect.

and also from Wiki

Lead-tin solder layers on copper leads can form copper-tin intermetallic layers; the solder alloy is then locally depleted of tin and form a lead-rich layer. The Sn-Cu intermetallics then can get exposed to oxidation, resulting in impaired solderability.[78]
...
# Cu6Sn5 – common on solder-copper interface, forms preferentially when excess of tin is available; in presence of nickel (Cu,Ni)6Sn5 compound can be formed
# Cu3Sn – common on solder-copper interface, forms preferentially when excess of copper is available, more thermally stable than Cu6Sn5, often present when higher-temperature soldering occurred

Ok, intermetallics are formed because of some diffusion/solubility between the solid Cu and liquid Sn/Pb mixture.  I don't fully understand how, they kinda skipped over that stuff in my phase diagrams class apparently, and I was more interested in focusing on welding metallurgy and arc physics than brazing/soldering/solid state joining processes.  Now the phrasing of the notes about the different intermetallics makes me want to know more about the particulars of what's happening.

Just out of curiosity, what's your background?  I gather you're probably a welding engineer or a really practical metallurgist or something of that ilk. 

[frank_p]

Just out of curiosity, what's your background?  I gather you're probably a welding engineer or a really practical metallurgist or something of that ilk. 

I did Mech. Eng.. But the reason why I was relatively interested in that kind of stuff is because I worked in R&D for securing blades on hi-speed wood shaping tools for industrial cutting machines.   I had to do some resistance testings protocols and a bit of microscope inspecting on the carbide brazings with a lab assistant at the university.  The guy was autistic, but for I don't know for what reason, he had memorised all the metallurgy encyclopedias he had in his lab and he could identify and calculate the % of each phases of metal samples just by looking rapidly in the microscope. Then he would talk endlessly about grains while playing in his nose and scratching his balls  .    But that was 10 years ago... and all I remember about brazing is that the science behind is quite complicated and mysterious.