Circuit Board Solder Reflow

So, what factors have you ruled out as the root cause?

You have the solder, solder type, solder paste, PCB pad size and orientation, component spacing, obstructions, temperature, time, vibration, board contamination, surface oxidation, and the component leads themselves (contamination, oxidation, etc.).

Are the boards soldered by hand? Could be the operator. Do you use a solder pot or is this a reflow or wave solder machine? Are the problems always in the same areas or is this a random issue?

Has the process worked satisfactorily in the past? What has changed? Have the components changed (i.e., new batch), are the PCBs a new batch, have any environmental changes taken place? What about consumable materials, have any of those changed?

If you are using a machine to solder these boards you should consult with the company that makes it to see what their suggestions are.

Do you have pictures of the problem? You should put these under a microscope and take pictures of the problem areas so you have some documentation.

Have you defined the problem space adequately? I know these things are frustrating, but a detailed analysis of the problem helps determine the root cause. Documenting everything will also help if the problem ever shows up again. You won't have to retrace all your previous steps.

Lastly, as far as answering the question if your rework methods are okay, I would start with looking at all the component data sheets on the board. The manufactures will specify the type of solder, ROHS, pad sizes, temperature, and time. If you have questions about techniques those manufactures have field engineers that can advise you on best practices. Sometimes with mixed technology boards there can be challenges to find a process that works for all components.

Hi Anonymous Hero,

May be I didn't explain myself clearly enough, we are service electricians.

The problem was on a PLC about 8 years old, and only just started causing a problem, so I had a look at it and determined that there must be a dry on one of the circuit boards, but finding a dry joint on an SMD circuit board can be quite difficult, so I thought why not give the heat gun a go, worst case I would wreck the PLC but in any case it wasn't working anyway. Heating the board with the heat gun until the solder reflowed did the trick, and the PLC is now working.

We get a reasonable amount of electronic equipment to service and a lot of the time the fault is due to dry joints, this was the reason for my question, I was wondering the use of a heat gun to speed up these types of repair is used in the industry?

Cheers

Joe

Joe,

That clarifies the situation somewhat and modifies my response.

In your situation the heat gun is a brute force approach which may destroy the board as often as fix it, and with the board you have 'fixed' you don't know if the fix is good for long term operation and you do not know what else you may have heat damaged in the process.

I suggest an alternative:

1. Do a visual inspection and re-solder any joints that are obviously broken or suspect, if that fixes the board great!, proceed to step two, if not proceed to step two also!

2. To identify the area of the board that has a dry joint - with an insulated shaft - dry wood or plastic gently press on the PCB in various locations with it powered up - any change in behaviour indicates a dry joint, generally close to the position being pressed, or close to a mounting point - somewhere that you have introduced some flex to the PCB - that is why you press gently, to localise the flexing - then resolder joints in that area. Proceed to step 3

3. Repeat step two until either the board operates normally and doesn't misbehave no matter where you press - you have fixed it, or until the fault remains no matter what you do - the board either has a broken connection which is very difficult to find, or a faulty component - either way replace the module.

Regards

Bill

Well, you can't control the quality of your repair with just a shotgun approach. You don't know if the fix is right or just temporary or if you stressed a component to the point where it will fail.

What are you using for a heat gun? Is it designed for SMT work or is it just an industrial heat gun?

They make heat guns for SMT that can regulate the temperature correctly without damaging the components.

I won't add much to the previous post, except to say that I see no way to control the amount of heat you put into the assembly using a heat gun. And the air stream of the gun will blow the liquid solder around.

You need to find the root cause of the wetting problem and leave the heat gun for shrinking sleeving. Unfortunately, as AH has stated there could be any number of problems, depending on processes involved.

Good luck.

The right questions have already been asked, especially relating to "What has changed?".

Using a heat gun in the way you described is absolutely BAD practice for production items, but might help with diagnosis of the root cause.

Have you run KIK or PROPHET or any other temperature profiler on the soldering process?

Worked for a few years in as production Eng/supervisor in electronics manufacturing department using reflow and wave, and saw this type of problem occassionally. Each time was one of the fundamentals that had drifted.

It is unclear (to me) whether you are manufacturing these boards or working as service/repair of boards in the field and this will mean a difference.

The other question is "What class of board are these?" Class1 and Class 2 are allowed to be hand reworked with different levels of documentation expected, but class 3 MUST be reprocessed using the original intended manufacturing controls and methods with very specific record keeping expectations.

If this forum is unable to lead you to an outcome and you need "professional" help, contact Heinz Zimmermann at SUBA Engineering. SUBA have significant experience in this field with excellent hands on knowledge from commissioning electronic assembly lines. They have good representation in NZ and AUS. (I'll stop there since this might begin to sound like a commercial for them.)

Stay in touch.

With apologies to Gilbert & Sullivan " When I was a lad I learned my trade reworking solder joints on new (Tait) 2 ways"

Thirty odd years later I was still doin it (yes Tait still can't solder...) though I'm now no longer in the 2 way game, having progressed into commercial/mining equipment.

Hand reworking boards is better than the scattered heat gun treatment. If your looking for process faults, it will make it easier to do your "forensics" if you do them by hand.

As everyone else has mentioned, you have to look at your basics to determin the root cause of the problem.

Generally with SMD's it will be the oven profiles that are letting you down, and with plate through it may be a batch of flux that is not quite right and or and a wave that is running the boards through a tad too quick.

Environmentally it could be your ambient factory temperatures are pulling the heat out of your processes. I'm assuming you don't have a climate controlled environment..

If all else fails refer to the IPC manuals.

Cure or fix (there is a difference).

For fixing the odd printed circuit board (either one that failed a quality check at the factory or has come back from the field for repair) then you can carefully and uniformly heat the whole board and the components to get the solder to reflow. A Sandwich oven works well for SMD and through-hole components. Yes it is a common practice in the industry and yes it does work!

A heat gun is not (generally) recommended as it has a tendency to a) not be hot enough and b) blow solder off on to adjacent pads (especially with SMD chips). I managed to blow an entire chip (and a capacitor or two) clean off the PCB using this method. Additionally there is a chance that the heat gun will not be able to properly heat up and reflow the solder (potentially creating a short-term fix that may fail in the future), and too much heat on one spot could cause thermal stress and a potential reduction in component reliability.

A soldering iron or SMD rework soldering station work well and in a timely manner (but only if you know where the fault is or are good at guessing). There is still space for the sandwich oven next to the other two, with the heat gun relegated to thermal cycling fault finding duties.

HI Jack of all trades, a sandwich oven why didn't I think of that. How would this work when the cct board contains control relays wouldn't they melt? What temperature would you suggest we bake the board at?

My other thought was to use a plastic welder heat gun as these are temperature controlled and the heat is more focused, whats your opinion?

We still have to catch up some time, I am only around the corner, pop around some time for a coffee or better still a beer.

Cheers

Joe

How would this work when the cct board contains control relays wouldn't they melt?

Plastic (and other heat sensitive) parts would need to be removed from the board before baking, so the technique is not suitable for all situations (the chain is only as strong as its weakest link after all), but does work quite well for populating a PCB with SMDs.

What temperature would you suggest we bake the board at?

I cannot remember, it has been years since I needed to use the technique. Medium/Medium High I think, carefully ramping up the temperature over a period of time to heat everything evenly. You will need more heat if lead-free solder (higher melting temperature) has been used however.

My other thought was to use a plastic welder heat gun as these are temperature controlled and the heat is more focused, whats your opinion?

It may work, but I think you are still better off with either a soldering iron or SMD hot air soldering station. For small quantity work a soldering iron is fine (even though it may take a while to reheat all the potentially suspect solder joints), it is what I use for small quantity repairwork.

A standard or temperature-controlled heat gun just has too many potential problems (and too high or low non-adjustable fan speeds) that will more likely waste than save you time in the long run (especially with SMDs that can move in the heat flow), and an SMD hot air soldering station is quite an expense that you may not be able to justify for the quantity of work you are doing.

We still have to catch up some time, I am only around the corner, pop around some time for a coffee or better still a beer.

Sound like a plan.

Joe,
It's unclear whether all the troubled PCBs are coming from one source. Beyond repairing PCBs it could be valuable to determine if they are being manufactured using a dodgy reflow oven.

For board prototyping we invested in a desktop sized T-962 reflow oven from www.tech168.cn and it is atrociously variable. The geographic temperature distribution is so deviant it is embarrassing. It has a "sweet spot" around the centre, but boards that extend beyond the sweet spot always require inspection, test and manual rework. We do that with an iron.

Mark Bingham
Relativity PL

As previous posters have stated,

1. a heat gun is a bad idea - you will overheat some areas and not heat others enough to achieve reflow - results will be inconsistent and probably heat damage some components and/or the PCB.

2. from your description it sounds like a process problem.

3. if the dry joints are general, spread over the board, then once the process is fixed they can be passed through the reflow oven again (assuming you are using a reflow process). - this will apply the correct amount of heat evenly across the PCB - but remember to re-flux the PCB first and the clean it afterwards.

4. alternately there are single board vapour phase reflow ovens available for rework situations (expensive but available). - again remember to re-flux the PCB first and the clean it afterwards.

5. if the above options are not available hand rework with either a hot tip soldering iron or a hot air soldering iron(these have very small nozzles (1-2mm dia) for precise rework), but be careful because with these because it is easy to overheat the component and/or the PCB, and even with a low air flow rate it is easy to blow small components off the PCB.

Regards

Bill

Hi Contributors, thanks for your good advise, all I was trying to do was to find an effective way to repair circuit boards (any make or model) I quite like "Jack of all trades" suggestion to use a sandwich oven or something similar that we can just put the boards in and let the heat to the work.

I don't have a technical qualification in electronics, apart from Advanced Trade papers C and D, but that is more theory than hands on training.

I suppose that I was just trying to find an easier way to service the dry joint problems in one quick hit and I thought there had to be a better way than re-soldering each joint by hand, also sometimes it can be rather difficult to do this process by one joint at a time also it is not cost effective.

Thanks for all your help

Best Regards

Joe

You will find anything with a plastic cover or has a plastic component like connectors and relay will be "toast" if you place it on an oven.

The best way is the old way I'm afraid. That is to rework the offending area with an appropriate soldering iron and solder.

It's not only SMD joints that can be insidious but PTH (plate through) ones as well. often these dry joints are heat stress/fatigue related if its in older equipment.

I know it's painful, but it is a commercial environment your operating in. If you don't want to repair it let the customers bean counter decide whether he wants to pay for a repair or to replace/upgrade the equipment. Especially if it is equipment that is well past it's prime.. Just make the replacement more "attractive" than the repair.

IR oven (or toaster) in not a good idea for most finished boards. Heat sensitive parts are added after factory solder process (IR reflow or wave) and will be damaged.

I suggest a good magnifying viewer (very useful for 0204 class parts) and a good soldering iron (cheap and simple) or one of these hot air reflow workstations (expensive, but sometimes the ONLY solution)...

http://www.smtmax.com/category.php?id=26

We successfully used both methods in a factory production and a service repair environment.

I think you are referring to what we call "cold" solder joints. If that's the case, the best solution is to apply a small amount of rosin flux to the joint and remelt it with a soldering iron. It should look shiny after it cools. If it doesn't, the solder is likely contaminated and should be wicked away with an iron and some de-soldering braid. A new joint can then be made with fresh solder.