Skillet vs Oven vs whatever... What is the safest bet?

Hi All…

I have read the tutorials and a bunch of other stuff, but I just can’t decide on the better approach. I have a product I need to make and the company that was going to assemble it said they can’t get to it until June.

My PCB is 10mm X 80mm. It has a USB connector and a SD card socket. Of course it has a bunch of ICs, resistors, capacitors and so on. Its all designed to be surface mount, except for some screw terminal strips that are thru-hole. I figured I would manually add those after the reflow. I am only going to have 5 prototype PCBs made and there are some expensive components. I can’t really afford to destroy several boards and components figuring out the best approach.

So it seems with an oven I run the risk of melting parts of the SD socket and USB connector, but I can get better temp control with a controller, such as one from SparkFun.

With a sklillet I get hot spots and cool spots, and can’t really control the temp well. I could burn the bottom of the board, but I probably won’t damage the USB or SD connectors. I did see one project where there is a temp controller made for a hot plate, but that does not solvbe the hot spot problem.

So what’s the safest bet?

Has anyone tried an oven with the top heating element disconnected, or will that not get hot enough?

I’m open to any suggestions, thank you!

Just use a suitable soldering station for a small batch of boards like that. It won’t be fast, but you will be able to test the first board in stages as you put it together. Faultfinding a new design can be very difficult if it is all assembled at once; you won’t know if you have a board manufacturing fault, a design error, incorrect component placement, or poor solder joints.

In the past I used a skillet in combination with a hot air (paint stripper) gun. That way, you can observe progress and apply a bit more heat where necessary. Now I’ve got access to a Chinese reflow oven I just use that, but sometimes it doesn’t reflow the heavier components completely.

1,What is the surface finish of your board? Maybe it is the reason that affect the reflow.

2,The components are expensive, as you said, I think the bare pcb is very cheap for 10mmx80mm, maybe you can re-make the bare pcb, that is the way to find the root cause.

by the way, may i know what is the totall price for your prototype?

Thanks everyone.

I don’t think hand soldering is the best bet. Its not my best skill, and the units will have to be assembled eventually suin SMT. I think its better to put the time and money into the equipmrnt now.

bobliu1982:
1,What is the surface finish of your board? Maybe it is the reason that affect the reflow.

2,The components are expensive, as you said, I think the bare pcb is very cheap for 10mmx80mm, maybe you can re-make the bare pcb, that is the way to find the root cause.

by the way, may i know what is the totall price for your prototype?

Hm, I don’t know the surface finish. I have not actually tried this yet, I was hoping to get some advice on the best method to try first.

With the board and the parts, each board will probably end up around $110.00 USD. I know it does not sond like much, but if I screw up three or fou of them its real money.

A simple solution to hotspots in a skillet could be cured with a thick piece of aluminum in the skillet with your board set on top of that.

http://www.onlinemetals.com/ would have what you need or give you a good idea on what the price would be like.

Chagrin:
A simple solution to hotspots in a skillet could be cured with a thick piece of aluminum in the skillet with your board set on top of that.

That will certainly help, but certain parts of the board need more than others (ie, where there are heavy components). I found using a heatgun allowed me to apply this extra heat where needed, without overheating the rest of the board.

Thanks everyone.

Would it make sense to try a toaster oven with the top element disabled?

Also, if I use a lowe temp paste, might that solve the problem of damaging USB and SD connectors?

I suppose another option is to hand solder anything that could be damaged. Actually, that’s just the SD socket. The USB, header pin strips and screw terminals are all thru hole so I need to do those manually anyhow.

As I think about it, what I like about the oven is that it seems its easier to measure and control the temp.

jarcher:
Also, if I use a low temp paste, might that solve the problem of damaging USB and SD connectors?

If you're using lead-free paste, then you should by some proper solder (ie, containing lead), especially when you're prototyping. Lead-free solder is very problematic, for a number of reasons.

MichaelN:

jarcher:
Also, if I use a low temp paste, might that solve the problem of damaging USB and SD connectors?

If you're using lead-free paste, then you should by some proper solder (ie, containing lead), especially when you're prototyping. Lead-free solder is very problematic, for a number of reasons.

Oh! Thanks for the tip, I didn’t know that. My problem is i need to ship some of these prototypes to Europ and Australia, where it needs to be RoHS compliant. My understanding, which may be wrong, is that it needs to be completely lead free.

jarcher:
My problem is i need to ship some of these prototypes to Europ and Australia, where it needs to be RoHS compliant. My understanding, which may be wrong, is that it needs to be completely lead free.

I'm not up-to-date with the legislation unfortunately. Having said that, I've designed products for a number of companies in Australia, and we always used lead-based solder for our products (which were low-volume, specialist items).

Lead-free is proving to be a nightmare for reliability. First, it requires higher soldering temperatures, which puts more themal stress on components. Second, solder joints are more brittle and fail earlier. Third, “tin whiskers” are a major problem, as they cause short-circuits on PCBs (even when encapsulation / conformal coatings are used).

It’s only a matter of time before this tree-hugging madness will bring down an airliner (if it hasn’t already)…

MichaelN:
It’s only a matter of time before this tree-hugging madness will bring down an airliner (if it hasn’t already)…

No argument there. I’ll look into the requirements more carefully, but if I have to do it, well I have to do it. But maybe I can make different runs for the parts of the world where lead paste is allowed. All the components i specified are compliant, and i was planning to have the boards made compliant as well.

Will the boards be less reliable if they are lead free?

Thanks very much…

jarcher:
Will the boards be less reliable if they are lead free?

It depends. Gold or silver finished boards are much better than tin, but the tin plating on the components, as well as the tin in the lead-free solder could still cause problems.

For lead free boards (PCB, components, solder), you have the following things to take care of:

  • - The board need to be high TG, for example TG170 because it needs to go through higher temperature which leaded process doesn’t need to. This is important as for lower TG material, the laminate will not be able to withstand the higher temperature.

  • - The surface finish needs to be suitable with the lead-free process. You cannot use the normal leaded HASL. You can either choose immersion gold, immersion tin, OSP, and some other finish which I can’t recall.

  • - The components must be lead-free. They should be able to withstand the higher reflow temperature at least once.
  • You might want to try using this:

    http://www.reflow-kit.com/rkuk/order_pr … fe64f39427

    Okay, thanks again for the information. Apparently I have a lot of reading to do about building things.

    I agree with 48X24X48X, one more thing to say, lead free hasl is also meet ROHS complain.

    The high tg170 materials will expensive 30% of normal TG FR4 material.

    I’ll run the numbers, but the bottom line is that if i want to ship these tot he EU I need to comply. It might be worth doing different PCBs and paste for the EU, but probably not.