SMD Laser Soldering at Home - Your thoughts?

[Syntax Error]

Straight question: Has anyone on the EEVBlog Forum experience of soldering SMD parts using a laser?

I was pondering using one of those mini home laser engraving rigs to melt the solder paste of SMD parts onto the PCB. Last summer, I tried soldering with sunlight focussed through a large magnifying glass. This certainly melted the solder paste... and incinerated the SMD parts! Especially those blacker than the paste. Only white ceramic caps survived the ride. Although significantly impractical, in principle this method proved that heat rays work.

Home laser engraving rigs have a 0.2mm precision; which suggests the paste on the pads could be heated to the necessary +180 to +200C, without frying the adjacent component. Theory. So is a 1500mW thru 3000mW laser that scorches wood, hot enough to melt SAC305? This concept would work on 1206, 0805 and SOT23 packages. Dpak cases might be too ambitious.

Any thoughts?

[JustMeHere]

I don't see how you'd sufficiently follow the solder profile of the parts.  How would you melt solder under parts?

You will be better off with hot air and/or a hot plate.  These are designed to help you follow the reflow profile.

There's also more than just "melting the solder"  You also need to heat up the metal you want to solder.

[S. Petrukhin]

If you are talking about mechanically moving the laser head and soldering the pads one at a time, then you risk getting tombstones on components with two leads - the molten solder on only one side will pull them on itself. And in this case, there will be no alignment of the components. If you use a complex and fast beam deflection system so that all the sites are illuminated on the PCB at once, it is quite likely. But even in this case, it is not possible to melt the contacts hidden under the case.

[wraper]

Not possible with any sort of consistency without very advanced tech with beam forming and temperature control.

[S. Petrukhin]

Not possible with any sort of consistency without very advanced tech with beam forming and temperature control.

Everything was once impossible. 

It is possible that we are discussing the birth of a new technology. Imagine: the head of the installation machine is equipped with a laser and after installing the component, it immediately seals it. That would be nice.

There may even be a way to solder hidden pads and THR.

[wraper]

Everything was once impossible. 

It is possible that we are discussing the birth of a new technology. Imagine: the head of the installation machine is equipped with a laser and after installing the component, it immediately seals it. That would be nice.

There may even be a way to solder hidden pads and THR.

Laser soldering does exist and works just fine. What I wrote impossible, is doing it with a simple laser with any sort of consistency. Solder will either not melt or everything will burn, or both at the same time. Simply timing exposure will not work as every solder joint is different. In theory you could do this with just a laser, but you will need to calibrate every solder joint. After hundreds of destroyed PCBs, you will probably get it somewhat right... For particular PCB from particular batch.

[S. Petrukhin]

Everything was once impossible. 

It is possible that we are discussing the birth of a new technology. Imagine: the head of the installation machine is equipped with a laser and after installing the component, it immediately seals it. That would be nice.

There may even be a way to solder hidden pads and THR.

Laser soldering does exist and works just fine. What I wrote impossible, is doing it with a simple laser with any sort of consistency. Solder will either not melt or everything will burn, or both at the same time. Simply timing exposure will not work as every solder joint is different. In theory you could do this with just a laser, but you will need to calibrate every solder joint. After hundreds of destroyed PCBs, you will probably get it somewhat right... For particular PCB from particular batch.

Why do you find this difficult? Even a simple toy laser is well dosed and makes a fairly accurate burning on wood or leather. I suppose you find it difficult to dose the laser due to the different heat output of different pads?

[wraper]

Why do you find this difficult? I suppose you find it difficult to dose the laser due to the different heat output of different pads?

You have no clue about lasers (I have $3k CO₂ cutter/engraver) and they are not accurate at all. It's a one thing when you need to burn through, and just use more power than actually needed for some margin. And normally used on homogenous materials. Engraving as well, there is not much difference if you deliver 20% more or less power. Once you need to deliver exact amount of power, it becomes a big problem. I've played enough with removing material to certain depth, even made countersunk holes. It's a black magic and pain in the ass. Each pad will be different. Even if pads are the same by themselves, traces connected to them are not. Imagine as if you are using a soldering iron with 2000^oC tip temperature, and need to find exact time for each solder joint so it does not burn. Let's assume you spent a zillion of man-hours and hundred of destroyed PCBs to calibrate each pad. When you receive the next batch of PCBs and especially if from different manufacturer, likely it will not be the same. Nor laser will keep it's power exactly as it was before, unless it's some weak semiconductor type operated in stable conditions.

Even a simple toy laser is well dosed and makes a fairly accurate burning on wood or leather.

If your intent is burning PCBs, it can be done easily. BTW there is not much use from a few W lasers for soldering, too weak.

[S. Petrukhin]

I really didn't know that the control energy of the laser is so inaccurate. But what about tightly recorded DVDs? What about laser scalpels in medicine? Or is hair removal practically a toy device? Are there complex structures, and not ordinary semiconductors?

[S. Petrukhin]

But again-now we can't control the laser precisely, 20 years ago we didn't have a laser at all in everyday conditions. What will happen in 10 years?

[wraper]

But what about tightly recorded DVDs?

It's easier at low power and does not need exact precision. Discs are written with ones and zeroes.

What about laser scalpels in medicine?

You are cutting basically homogenous matter, and it's not like you make a deep cut in single pass.
Again, you need a closed loop system for laser soldering. Think about using opamps without a negative feedback.

Or is hair removal practically a toy device?

Hair removal is simply blasting your skin with fiber laser. It's nowhere precise. It's just that hair is way more affected by it than the rest of skin.