Cracked surface mount solder joints.

[chaoticmayhem65]

Hello all, first post here. Anyhow i have a question im sure many of you skilled electronics gurus will probably be able to answer. I've been working on a lot of Chevy/GMC instrument clusters lately and they tend to all have cracked solder joints on surface mount components, mainly the resistors. My question is: What is the best way to reflow or resolder them in a timely manner? Usually i just dab a bit of flux on the joint and use the tip of my hakko soldering iron at 350 degrees Celsius and drag it across the joint and possibly add a touch more solder. What is the correct way? Keep in mind i have a hot air rework station if thats the better choice.
I've included a couple pics of the resistors. You can see the cracks only under magnification. These are 510 ohm 2010 component resistors i believe.

[free_electron]

Lemme guess. They drive the VFD display filament ? Bad design. These resistors run soo hot that they desolder themselves over time.
Ive had two fix two for a collegue of mine' (he drives a Buick)

But your technique is fine. Apply flux, regeat with soldering tip. Add a bit of fresh solder too , but use lead free as ot has a higher melting point. You may want to remove the old solder using a bit of solder wick.

[Rerouter]

glad to see another instrument technician on the forum,

rework with an irons generally the best way, just clean the joint before adding new solder, you can look up nasa's guidelines if you want some visual cues for what a good joint should look like,

from the colouration on the board the issue is they have all been running very close to there specified wattage, (yellow colouration in coating) and the resistors themselves appear to all be in parallel if I'm tracing them correctly, if so see if you can swap them out with a higher wattage through-hole alternative, so the heat is dissipated away from the joints,

[SeanB]

That will work, or if you are doing hot air reflow a syringe of solder paste that you use to add a drop to each end works as well. The better way would have been to have used a glue stencil to add a drop of a red SMD adhesive to hold the devices to the board and not rely on the solder alone in such a high vibration environment. Definitely a failure on the part of GM, they should be replacing these with new ones that are a better design free of charge under a recall notice.

[chaoticmayhem65]

Your close. Yea they do drive the VFD but its not a filament is it? And yea the buicks did have this problem too but on the 2002 and down. This is on a 2003-2006 GMC/chevy. Typically in your silverados and yukon/tahoe. But yea your right on the money. Really flawed instrument clusters. These resistors dont fall off though like they did in the buicks. They just crack until eventually they just form an open circuit causing a power problem.
And lead free solder?? I thought this was the problem to begin with? Lead free solder is known for poor thermal cycling  capabilities no? Wouldn't the best choice be to use like a 63/37 tin/lead solder?

[chaoticmayhem65]

That will work, or if you are doing hot air reflow a syringe of solder paste that you use to add a drop to each end works as well. The better way would have been to have used a glue stencil to add a drop of a red SMD adhesive to hold the devices to the board and not rely on the solder alone in such a high vibration environment. Definitely a failure on the part of GM, they should be replacing these with new ones that are a better design free of charge under a recall notice.

Red SMD adhesive? Is that like a preventative adhesive to be used after the joint has been made on the assembly line? And there was a recall notice and GM only covered them for like 80K miles or 7 years. Pretty pathetic warranty. And there is no better designed boards, only different revision stepper motors to replace recalled ones. But as for the board design there has been no fix or different revision boards. I've received Delphi Certified re-manufactured instrument clusters that have to be repaired after a certain amount of time because of poor board design.

[SeanB]

The red adhesive is a thermoset epoxy screen printed onto the board before mounting the components. Then the board is heated to cure it and the components are basically part of the board, then it goes through a wave solder bath that does the actual joints.

The adhesive keeps the components on the board, but it needs a pick and place machine and a syringe dispenser to apply it in precise drops if you are using a solder paste and then mount the components and IR reflow them.

Boards with the SMD parts glued to them do not suffer much from cracking solder joints, though the components can be physically cracked if there is poor attention during design so that there are no forces applied to the board during assembly ( like having to press on a package during assembly to mount the board in a retaining post) that will flex the components and break either the solder joint or the component itself. The adhesive ensures that the solder joint is either in compression or has zero stress on it at all times, so it does not crack.

[free_electron]

Your close. Yea they do drive the VFD but its not a filament is it? And yea the buicks did have this problem too but on the 2002 and down. This is on a 2003-2006 GMC/chevy. Typically in your silverados and yukon/tahoe. But yea your right on the money. Really flawed instrument clusters. These resistors dont fall off though like they did in the buicks. They just crack until eventually they just form an open circuit causing a power problem.
And lead free solder?? I thought this was the problem to begin with? Lead free solder is known for poor thermal cycling  capabilities no? Wouldn't the best choice be to use like a 63/37 tin/lead solder?

Yes, these drive the filament in the vfd. They are all in parallel.

A vfd uses a high voltage for the segments and a or grids. But there needs to be an electron emitter. If you carefully look at the vfd grlass you will see 4,5 or 6 very thin wires going from all the way left to all the way right. The are comnected to what looks like little springs. Actually they are. When you power up the vfd you will see the little blades extend. This is because the filament heats up and stretches a bit. When you power off it crimps .

The filament is the electron emitter.

This problem is a school case of bad engineering. Let's put some power resistors in series so we get about 4 volts across the filament. ...

[chaoticmayhem65]

Your close. Yea they do drive the VFD but its not a filament is it? And yea the buicks did have this problem too but on the 2002 and down. This is on a 2003-2006 GMC/chevy. Typically in your silverados and yukon/tahoe. But yea your right on the money. Really flawed instrument clusters. These resistors dont fall off though like they did in the buicks. They just crack until eventually they just form an open circuit causing a power problem.
And lead free solder?? I thought this was the problem to begin with? Lead free solder is known for poor thermal cycling  capabilities no? Wouldn't the best choice be to use like a 63/37 tin/lead solder?

Yes, these drive the filament in the vfd. They are all in parallel.

A vfd uses a high voltage for the segments and a or grids. But there needs to be an electron emitter. If you carefully look at the vfd grlass you will see 4,5 or 6 very thin wires going from all the way left to all the way right. The are comnected to what looks like little springs. Actually they are. When you power up the vfd you will see the little blades extend. This is because the filament heats up and stretches a bit. When you power off it crimps .

The filament is the electron emitter.

This problem is a school case of bad engineering. Let's put some power resistors in series so we get about 4 volts across the filament. ...

Ok, gotcha. It sucks though cause these new boards (2003-2006) have so many surface mount resistors and components and the majority of them crack. So just swapping out a surface mount resistor with a through hole really isn't an option. And then using my iron and some flux doesn't leave the prettiest joint in the world cause some of these resistors are in awkward positions surrounded by other components, and its hard to get the iron at the angle needed. I've included a pic below and circled the areas that fail. Maybe you guys could give me some insight on the best, most time efficient way of fixing these cracked joints. Keep in mind, I didn't put these bulbs in. This is a Delphi certified re manufactured board.

[jeroen74]

I spent quite some time finding out why my wife's clothes drier didn't work properly; the motor would sometimes not start running, or stop amidst a cycle, or have trouble starting up; the machine would beep an error code when the motor stopped. Most common cause on this machine are the relays that were not sealed against moisture in early production runs. New relays did not fix it. A good slap would get it running again for a while though.

In the end I found out it was a cracked solder joint on the gate resistor of the motor thyristor. Resoldered it, and the machine works like a charm again

On topic; I know ceramic capacitors are available with flexible terminations especially for automotive use. I'm not sure that termination option is available too for resistors. Replacing them with such one could fix the problem (basically proper component selection).

[Falcon69]

I'm guessing here, but it looks like those are 1206 size resistors, probably in 1/4 watt.

Why not unsolder them with some solder wic, and buy some new 510ohm 1/2 watt resistors. they won't cost more then 10 cents each, and may solve the problem of heat.

http://www.digikey.com/product-detail/en/CRCW1206510RFKEAHP/CRCW1206510RFKEAHPTR-ND/2227720

I know with digikey you'd have to purchase a reel of 5000, but, doing a search, i'm sure you can find another distributor somewhere where you can purchase just a handful of them. The point is, they make em!

[chaoticmayhem65]

I'm guessing here, but it looks like those are 1206 size resistors, probably in 1/4 watt.

Why not unsolder them with some solder wic, and buy some new 510ohm 1/2 watt resistors. they won't cost more then 10 cents each, and may solve the problem of heat.

http://www.digikey.com/product-detail/en/CRCW1206510RFKEAHP/CRCW1206510RFKEAHPTR-ND/2227720

I know with digikey you'd have to purchase a reel of 5000, but, doing a search, i'm sure you can find another distributor somewhere where you can purchase just a handful of them. The point is, they make em!

They are actually 2010  510ohm resistors and i believe they are 1/2 watt. I've already identified them with digikey and purchased some. Digikey's tech support is great. They were very prompt with a reply and it was even on a sunday. And actually digikey didn't make me purchase 5000. I only purchased 25.
 

[WuWuZat]

That will work, or if you are doing hot air reflow a syringe of solder paste that you use to add a drop to each end works as well. The better way would have been to have used a glue stencil to add a drop of a red SMD adhesive to hold the devices to the board and not rely on the solder alone in such a high vibration environment. Definitely a failure on the part of GM, they should be replacing these with new ones that are a better design free of charge under a recall notice.

Red SMD adhesive? Is that like a preventative adhesive to be used after the joint has been made on the assembly line? And there was a recall notice and GM only covered them for like 80K miles or 7 years. Pretty pathetic warranty. And there is no better designed boards, only different revision stepper motors to replace recalled ones. But as for the board design there has been no fix or different revision boards. I've received Delphi Certified re-manufactured instrument clusters that have to be repaired after a certain amount of time because of poor board design.

I have actually found several missing resistors over the course of the past year. 
Have you had issues with any other SMDs on these?  I have replaced several with wired components in the field with a butane iron when the component is missing from the board/cluster completely (not a great option with the limited amount of space when the cluster is re-assembled.
Pretty much an electronics caveman here though.  I have one right now that is missing several of the tiny brown resistors (I assume they are resistors) that are attached around the plug module.

[Rerouter]

If it is only the joints that are cracked, and not the components, some gel flux, and a hot air gun with a narrow nozzle, point it so the air flows between the movement and the plug rather than pointing straight down or at either, the flux will flow then a little later the joints will reflow better than ever.

The key trick is noticing when the joints begin to flow, as that means you can move on to the next component, for that many components, once the hot air gun is up to temp, would imagine no longer than about 2 minutes to reflow all of them,

[ion54]

All is about cost. GM is well known for cutting cost any way they can and driving suppliers into bankruptcy. If you can, get the following parts from Digikey:
http://www.digikey.ca/product-detail/en/CRCW2010510RFKEFHP/541-510PCT-ND/2222686
Remove all old ones and replace with the Digikey parts.
If done like that from the beginning it will cost GM about $0.05/car and the bean counters are multiplying that with the number of cars they make (let's say 1,000,000/year as it goes in multiple car lines) so they've got "savings" of $50,000 a year. Someone in purchasing will get a bonus for that. I know as I work in the industry.

[ConKbot]

Why all the suggestions for hot air and even solder paste?  The less heat you put into the board, the better, especially on an aging board.   A dab of no clean or rma flux from a pen, and 5 seconds with a soldering iron at most, and it's done. Even for a group of resistors together, hot air would take longer, and heat up the board a lot more.  Plus if you're using solder paste manually, and you put too much down, you end up with solder balls on the loose.

flux pen and a soldering iron would definitely be my go-to for repairing cracked smd joints unless it was a qfn or bga.

[Rerouter]

On this type board, the traces rarely lift, its just the soldermask that badly discolours from extended heating, so using a hot air gun is A-OK, and allows easier inspection of joints while soldering, (its a cramped area between 2 raised objects).

you use the right nozzle shape to direct the heat where you want it to go, and if your concerned about what else it is heating up, you turn the flow rate down, in reality you can hit closely packed resistors like that with hot air gun in a very similar time frame to an iron.

Also i agree solder paste would be unrecommended without first removing what is there, mixing 2 different types of solders tends to just make it more prone to failure in the future, but gel flux is fine.

[KG7AMV]

I know it is an Old Thread...

I have a GM cluster issue... Anyone know the cause of the Highbeam LED to stay illuminated with the IPC is powered?

[Gyro]

I spent quite some time finding out why my wife's clothes drier didn't work properly;

There speaks a true (edit: male) engineer! 

[KL27x]

I wonder if it would help at all to replace the 1x 2010 510R resistor with a stack 2x 2010 resistors of 1020R.

[KG7AMV]

All fixed up... No More Blue High Beam LED! Bad Solder on a Resistor... Center of Photo By Light DS19!