Printed Circuit Board Assembly & PCB Design SMT Electronics Assembly Manufacturing Forum

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


Samtec Searay 500 pin connector solder wicking issue after reflow

Views: 2130

#84178

Samtec Searay 500 pin connector solder wicking issue after reflow | 13 January, 2020

Hi, I need help on how to address the solder wicking on the samtec searay connector after reflow resulting to open solder joints on thick and thermal heavy PCB (4.4mm and 6.3mm). I increased my solder paste thickness to 8 mils, the connector is manually placed and my reflow process is vapour phase. I don't have problems though with thinner small form factor PCBs. I was thinking it is the thermal heavy PCB ( heavy grounding ) that can result to big temperature differential. The failure mechanism is such that solder from the solder paste wicks up to the lead, an indication the conn lead is hotter than the pad. I've tried many different profiles but to no avail. Would anybody think that convection oven would address my issue, assuming it is better in handling the heavy thermal mass? Appreciate any leads.

Regards, Arnel

reply »

#84181

Samtec Searay 500 pin connector solder wicking issue after reflow | 14 January, 2020

Hi Arnel,

I've used these before but with a 0.005" stencil. The key issue here was that the pins are actually a solder preform at the base that melts into the board - sort of BGA like, so too much paste can cause issues.

Samtec have some great applications engineers so worth contacting them for advice - also that have some really helpful processing documentation like the following link:

http://suddendocs.samtec.com/literature/searay_soldering1.pdf

reply »

#84182

Samtec Searay 500 pin connector solder wicking issue after reflow | 14 January, 2020

I've worked with these quite a bit in the past as well. As the previous poster mentioned, we ran either 4/5 mil thick stencil with a pretty standard RTS profile in a 10 zone convention oven without issue.

reply »

#84183

Samtec Searay 500 pin connector solder wicking issue after reflow | 14 January, 2020

Thanks Robl. I’ve contacted samtec followed their advice but ended up empty handed. Originally my stencil thickness is 6 mils, samtec stressed the need for the paste and charge to have good contact but this is challenged by the inherent charge non-coplanarity of 9 mils so pushing for an 8 mils would address it. In the end it seems it does not matter much since my problem remains. I have to tell you again that i do not have any issue with this connector on small pcb with my process. So, it appears to me now that the problem maybe the thermal differential management caused by the very thick thermal heavy pcb.

reply »

#84184

Samtec Searay 500 pin connector solder wicking issue after reflow | 14 January, 2020

Hi Phil, thanks. I am successful using the same connector on small thinner boards <3mm. Standard process would be sufficient. Yield is 100%. But with a thicker ( 4.3-6.3mm) and heavy boards (>2-2.5kg) i cant make it work, tried all kinds of profiles with vapour phase but to no avail. Yes you read it right 4.3-6.3mm pcb thickness and 2-2.5Kg. And it is 44 layers. Im thinking there’s a lot going on thermally on the boards, lots of thermal gradient exist affecting each of the pad ( some pads are hotter than others) as it relates to the temperature of the connector leads. If colder pads meets a hotter pin then becomes a recipe for solder wicking. This is my hypothesis though. I read from another thread that most of his open joints are those that are connected to ground, it make sense to me since the ground acted as a heat sink and therefore made it cooler relatively.

reply »

#84189

Samtec Searay 500 pin connector solder wicking issue after reflow | 14 January, 2020

Avillaro,

would you say that vapor phase process was selected, because the boards are so thick and heavy? I would assume so. Have you ever tried to run them through reflow oven? To understand better your challenge, we need to understand the setup better. What is different between thinner and thicker boards? We would think the thinner board will bend more, creating more open joints. Are boards at same temperature in VP? We assume the thicker board needed more temp, so may be now the connector bends more? Are boards same/ similar shape? How about speed of moving this board up and down?

reply »

#84190

Samtec Searay 500 pin connector solder wicking issue after reflow | 14 January, 2020

Hi Evtimov, Thanks for your response. Yes, the vapour phase is suppose to handle big and heavy boards, at least thats the intention. Well our thin board is like around 3mm which is not really thin. In our application the thicker the boards means more layers, more copper and as you may know if they are not balance you get a lot of issues thermally. Running this in a convection oven is the next step we’re pursuing and this means subcontracting this since we only have vapour phase in house. The temperature of our boards at any 2 points across the connector have thermal gradients spanning ~10 deg C, this is big tenperature differential. On thinner smaller board, we can manage it to ~4 deg C. That’s how the thermal mass heavy board affects the temperature on the board. My vapour phase reflow process seemed pushed to the limit. Our vapour phase movement from one level to another is constant ( no adjustment ). If i want to quickly ramp the temp to reflow i just bring the assembly nearest to the boiling sump and manage the time. I’ve tried a profile with this but did not address my issue. Im running out of option with my process. I can entertain any ingenious ideas though, i read in one of the threads of the same issue that changing the solder paste resolved his solder wicking issue ( im using water soluble paste, will RMA paste can make a diffrenece?). It has something to do with RMA flux handling oxidation during soak well than watersoluble paste. What do you think?

reply »

#84194

Samtec Searay 500 pin connector solder wicking issue after reflow | 15 January, 2020

Having 10c thermal differential is a lot. Trying different paste might help - usually thermal expansion makes component detach from PCB, then joint is formed and solidified, then component goes back but is too late to form a joint. Cooling speed might be the key as well. One approach is to have your peak temp before last zone to give more time for forming this joint. The other approach might be having a fixture/top hat that keeps connector flat with the board during reflow/VP process.

reply »

#84195

Samtec Searay 500 pin connector solder wicking issue after reflow | 15 January, 2020

Hi Evtimov, Thanks again! The solder wicking/open solder joint is not actually gross rejection, it is only a few and appeared random (difficult to trend the location of the opens), what is perflexing is that i have this 1 or 2 open solder joints surrounded by many good solder joints ( inspection by xray ). So the root cause must be localized than general. I also have a weight on top of the connector ( tried 30g and 100g), failed to mention this earlier. I have to always go back to my argument that i have great success using this connector on small thinner pcb. So it seems obvious that the thermal heavy pcb is the culprit but do not have that certainty. Changing the board thermal design would be a challenge technically, expensive and will throw our schedule off. I wanted to exhaust all possibilities before we go to that path. I’ll try the rma solder paste.

reply »

#84196

Samtec Searay 500 pin connector solder wicking issue after reflow | 15 January, 2020

As far as the large thermal gradient I have on my thick and thermal heavy board, I've done quite a bit to minimize the gradient but not successful. it's like one pad is connected to a signal while the other pad is connected to ground, you can't bring the temperature to be close ( within 4 deg C ), even at soak because of the thermal sinking the one connected to ground will always be colder and the one connected to a signal will be hotter during ramp up, cooling will be the opposite, the pad connected to ground will be hotter (thermal storage) than the pad connected to a signal.This is same when you ramp the profile from soak to peak. Nothing much we can do on the profile. Convection oven in my mind will be more of a challenge, since heat transfer is "convection", vapour phase is direct heating as the hot vapour condense s on the board surface.Can a convection oven handle the temperature gradient to 4 deg C on a very thick thermal heavy board? What can it be for the profile design?

reply »

#84197

Samtec Searay 500 pin connector solder wicking issue after reflow | 15 January, 2020

Avillaro, I don't have hands on experience with Vapor Phase equipment, but my guess is that it will work better for thick boards like yours. Trying another chemistry is a logical move. May be it is worth trying the fixture for connector flatness as next resource(in case chemistry doesn't help). May be not just a weight over connector, but something that supports connector and PCB flat together.

reply »

IPC Certification Training Schedule

Reflow Oven