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

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


Wave Soldering - Icicling/Bridges

Hello, We recently installed a new/used UPK 650 fitted wi... - Jan 23, 2003 by hoss67  

#23132

Wave Soldering - Icicling/Bridges | 23 January, 2003

Hello,

We recently installed a new/used UPK 650 fitted with a foam fluxer using an RMA flux chemistry. We have had problems with icicling and bridging from the beginning. The bridging occurs mostly on DIPs and connectors. The icicling is on longer leads (our max length below board is about .200") and large leads. General wetting conditions and top side fillets are acceptable. We have had some success with experimenting with the following but still have problems;

Topside preheat matches flux mfgr's recommended. Even distribution of flux and correct SG. Vary conveyor speed. Pot temperature varied 480-500F Solder analysis normal. Pot dwell time ~3sec.

If you ask "why in the %&$*#* are you still using and RMA chemistry?" I have the same concern as there is no good reason for us to do so. We have plans to convert to a WS process later in the year.

The previous equipment used a wave fluxer and I'm starting to think that the flux we are using (Kester RMA 1515) could be part of the problem. If that flux worked fine in the old machine with a wave fluxer, could it be that a foam fluxer doesn't give me enough solids on the PCB to properly drain off solder on exiting the wave?

Another possibility could be the conveyor angle. We're at about 5.5 degrees. I've heard the 5-7 degrees is acceptable.

Any thoughts on this would be much appreciated.

reply »

#23133

Wave Soldering - Icicling/Bridges | 23 January, 2003

While not familiar with the flux, I'd increase the conveyor angle in a New York Minute.

reply »


RDR

#23135

Wave Soldering - Icicling/Bridges | 24 January, 2003

How are these connectors and dips oriented? If the leads are parallel as opposed to perpendicular to to the wave this can promote bridging depending on spacing and length. The long leads showing icicling are most likely either inadequate heat OR the flow of the laminar wave is not set correctly. Do you have back flow over the wave former that is the same speed as the board moving over it? This can be a pretty tricky adjustment sometimes but is worth it. I am not familiar with the flux either but if you are getting top flow, all other joints look good, etc... I would prbably look at the orientation and wave setup. Increasing the angle to 7 deg. will reduce bridging as Dave F pointed out.

Russ

reply »

#23137

Wave Soldering - Icicling/Bridges | 24 January, 2003

I sent you a couple of charts to assist with your troubleshooting. One is the Electrovert Wave Solder Troubleshooting chart and the other is Sandia Labs/Electrovert Optimum Solder Contact Time. Electrovert has a 6 deg. conveyor angle standard on the Econopak and Vectra machines. Correction: The Ultrapak (UPK) 600 and Electra models allow for adjustablity from 4 to 7 deg. The model Ultra 2000 wave machine can adjust from 4 - 6 deg.

reply »

Randy Villeneuve

#23141

Wave Soldering - Icicling/Bridges | 24 January, 2003

I am not familiar with your flux but with no-cleans (which I use) the primary cause of icicling is the amount of flux solids left on the board prior to wave and how long the board is in the wave (2.5 to 3.5 seconds should be OK). The amount of flux solids left on the board prior to the wave can vary depending on the amount of flux that was applied during fluxing or how much if any was burnt off prior to the wave due to excess board temperature. With a foam fluxer, you should be applying a good amount of flux (usually too much) but check out the fluxer in any case. Also check out your fluxer air knife settings if you�re using one. You could also try hand spraying a board and turn off the fluxer to see what difference that makes. Your top side board temperature just prior to wave should be around 100 - 110 degrees centigrade but that can very depending on board finish and flux type. Have you tried a different flux to check to see if your flux is the problem? If not, get a sample and try a different flux. Lead length can be a factor but usually it just causes bridging not icicling. Let us know what you come up with. Randy

This message was posted Add this forum to your site! Click to learn more. the Electronics Forum @

reply »

MA/NY DDave

#23146

Wave Soldering - Icicling/Bridges | 24 January, 2003

Hi,

Distance Engineering is always difficult.

I am guessing that the old machine and process ran wonderfully / acceptable.

If you still have the old machine, how about running an empty board through it and then the new machine. Without flux observe the temp with temp tabs prior to the pot to see that you are getting up to exactly the same temp.

With flux observe how you would have to adjust to get the same level of flux. If you do not have the old machine have the most familiar person help adjust the flux quantity prior to preheat to be similar, and use dryer, and wetter as test runs or alter the SG of the Flux to get the same effect.

The airknife and any brush make sure is working well.

I would look closely on the top side to make sure that all the solder fillets including those on components are just like they were on the old machine. Any solder balls? Any more rosin than normal between brickwalled R or C paks?

Please let us know when you figure it out.

YiE, DDave

This message was posted Add this forum to your site! Click to learn more. the Electronics Forum @

reply »

TriGuy

#23148

Wave Soldering - Icicling/Bridges | 24 January, 2003

RMA needs to be dry prior to the wave to work. If coverage is good and part orientation is good, then you should be okay. Pad size can be adding to the amount of material left as well. Granted it soldered before fine with .2 long leads, I would ditto the preheat settings to be correct and make sure they are dry via the knife and temps prior to wave. I would opt for a topside temp of at least 100 - 110C.

reply »

#23149

Wave Soldering - Icicling/Bridges | 24 January, 2003

DDave,

I do not have the benefit of having the old machine here for comparison unfortunately. Comparing topside wetting to old sample boards run on the older machine show similar results. No solder balls. I have a thermal profiler and have run many profiles. We have bottom side convection pre-heaters and I'm fairly certain we are not overheating or shocking the boards. Kester spec sheet recommends 82C for topside temp and I was told by their tech support to shoot for 90C.

We've adjusted our conveyor angle (this morning) from 6 up to 7 degrees per DaveF's response above. Only a few boards run so far but no noticeable difference in quality. Looks like I may be back to focusing on the flux.

A suggestion above mentioned wave dynamics. The exit wing on the lambda wave is flowing at conveyor speed and we do run lev checks with a glass plate.

Let me know if you have any other suggestions.

Thanks,

Joe

reply »

#23150

Wave Soldering - Icicling/Bridges | 24 January, 2003

Correction: The Ultrapak (UPK) 600 and Electra models allow for conveyor incline adjustablity from 4 to 7 deg.

reply »

MA/NY DDave

#23162

Wave Soldering - Icicling/Bridges | 24 January, 2003

Hi

Again Distance is tough, besides far less exciting.. You got the panther at your breast, screaming!!

Somehow you have to make this go the other way.

Rather than solder flowing any old place it wants to, when it wants to. Chasing the heat, Chasing wetting and doing all this funny bridging stuff.

Reduce Heat, Speed up the conveyor, reduce solders mobility in some way are maybe ideas. Yes reduce flux so everything doesn't want to solder.

By the Way, I could be all wet.

Without being there it could be something totally unrelated, like the components leads are different.

YiE, DDave

P.S. OF NO VALUE: Years ago an engineer I know put washers yes washers, in a soldering bath to clip icicling. He won rave reviews. As I think about your problem I think he was making the solder attract and cling to the washers instead of the leads. Now .200 is long compared to .125

This message was posted Add this forum to your site! Click to learn more. the Electronics Forum @

reply »

#23182

Wave Soldering - Icicling/Bridges | 28 January, 2003

I agree with most of the other suggestions here, but 90 C for a topside preheat temp seems low to me. I remember aiming for 98 to 104 C for topside temp just before hitting the solder. Boards that had large ground/power planes or a lot of layers I would aim for the 104 end as I would have less problems with the solder solidifying before it got to the top side.

You might try raising the temp of the solder pot to 505 F for this board. That extra few degrees may give the solder a chance to drip off the leads before it cools and becomes an icicle. The downside of a higher temp solder pot is you will generate more dross, but it may save you rework time.

Let us know which change(s) work.

Mike F

reply »

sajulius

#23183

Wave Soldering - Icicling/Bridges | 28 January, 2003

I've seen this problem on more than one wave,What i've done is to run the conveyor a little faster,use a solder temp of 505 degrees F,and most important is to make sure that the solder is flowing over the back. this will keep the thin buildup of dross on the wave to flow away from the board, And the surface tension will help stop the snap back of solder to the leads. Please write back and let me know how it worked!

reply »

#23188

Wave Soldering - Icicling/Bridges | 28 January, 2003

Mike,

I've tried the high side of the flux preheat spec with no success. I did just get off the phone with Electrovert tech support......

Get ready. This machine is configured with a Contour Plus N2 nozzle. We are not running N2. According to Mark Parrick (nice guy), anyone who runs this nozzle in normal atmosphere has the problems I've described in the post above. Because the Contour Plus was designed for N2 using an airknive, little consideration was given to applications not using it. Dross formation of the "Christ...It's huge(with Scottish accent)" variety is another result.

In summary, I now have to figure out whether to retrofit a standard air-type lambda nozzle or setup for N2. I'm going to look at the costs associated with both and see where the bang is. Considering the low-medium complexity of our boards I'm not sure N2 makes sense.

Thanks to you and everyone who responded to my original post. My problem isn't solved but at least I know that we were looking at the right things.

Joe.

reply »

MA/NY DDave

#23215

Wave Soldering - Icicling/Bridges | 31 January, 2003

Hi

Wow, I can't imagine this set up. No wonder some of us were coming from opposing directions.

My thoughts would be to buy another pot if needed and not go to N2.

Do you have a site that we could look at the dynamics of this N2 air knife/ solder nozzle set up.

YiE, DDave

This message was posted Add this forum to your site! Click to learn more. the Electronics Forum @

reply »

#23216

Wave Soldering - Icicling/Bridges | 31 January, 2003

Dave,

I do have the mechanical drawings but no electronic versions. After looking at the cost of setting up for N2 I decided against it for now, BTW. If you think about it, having an air knife on this nozzle would tend to negate the negative effects (icicling/bridging) for the "not designed for" wave dynamics. A crutch if you will. That's my semi-educated guess to the question of why.

At any rate, I'm looking for a standard lambda nozzle/flow duct to retrofit to my existing pot and we'll take it from there.

Thanks,

Joe.

reply »


RDR

#23217

Wave Soldering - Icicling/Bridges | 31 January, 2003

I would go for the laminar flow as opposed to N2. The contour nozzle doesn't really work all that well even with N2. I knew I should of asked you if you had a contour, I was thinking laminar flow hence the dynamics input.

Russ

reply »


RDR

#23218

Wave Soldering - Icicling/Bridges | 31 January, 2003

Hoss, if you have trouble with finding the parts you need this guy can probably help. He's our local rep for speedline/electrovert.

Jeff Howe - AMTI 303-748-4242 jeff@amtireps.com

Russ

reply »

#23219

Wave Soldering - Icicling/Bridges | 31 January, 2003

Russ,

Roger that.

Thanks.

reply »

reflow oven profiler