|Ask the Experts|
January 16, 2018
Problems with Insufficient Barrel Fill
We are seeing insufficient barrel fill (60-75%) during lead free wave soldering. Can you point to some reasons why we may be seeing this insufficient condition and suggest a cure?
|Expert Panel Responses|
There are a number of factors that often combine to cause this problem. First, there is usually a significant temperature gradient between the bottom of the board and the top side. Since solder likes heat, the solder is reluctant to flow up the barrel to a cold region. This can be determined by instrumenting the top of the barrel with a thermocouple during the wave soldering process. A top side heater can be added, but even then, it is not possible to eliminate the temperature difference completely. Second, internal ground planes connecting to the barrel act as a heat sink, preventing those pins from ever achieving a high enough temperature to reflow sufficiently. Thermal relief pad design for ground pins should be considered, as long as the electical performance of the grounding system is not compromised. The thicker the board, the more difficult it is to achieve adequate barrel fill sinc the issues mentioned previously are even more pronounced. Very thick boards, such as server boards, represent the ultimate challenge. Third, insufficient or inappropriate flux. Usually for through hole soldering, the barrel diameter is enlarged somewhat to facilitate flux spraying prior to wave soldering. Depending on the surface finish of the board, OSP or HASL, different fluxes with various levels of activity are available to address particular needs. One of the current trends occurring in the industry now is the elimination of wave soldering by use of pin in paste reflow techniques. In this approach, the connectors and other through hole devices are sourced as lead free reflow capable. Solder paste is printed on the through hole pad, and the through hole devices are reflow soldered at the same time as the SMT components. If the solder volume is insufficient to fill the hole, a solder preform from a tape and reel package can be automatically placed in the solder paste to augment the solder volume provided by the paste. Solder preforms in tape and reel packaging are available in standard sizes such as 0805, 0603 and 0402. All common alloys are supported, including SAC verions, and SnPb versions. Another trend is to solder the high density SMT devices on one side using typical SAC alloy solder paste, and the use a low temperature solder paste, i.e. SnBiAg to solder all the through hole components in a second reflow step on the other side of the board. The SnBiAg solder paste reflows at 138C, which is lower than SnPb, thus ensuring that all the existing through hole devices can go through the reflow oven. The low temperature of the second pass does not disturb the SAC solder.
Global Product Manager
Possible causes of insufficient fillings are:
Managing Director, Asia Pacific
Insufficient barrel fill can be caused by many aspects in a wave soldering process. Some of these can be machine parameters such as poor flux penetration through the barrel, insufficient preheat, excessive preheat, insufficient dwell time and solder temperature itself. This can also be caused or effected by assembly related issues such as tight hole to lead ratios, board thickness, pallet designs, and no thermal reliefs when needed. The above is just to name a few as the question is complex. In answering your question, let's look at the machine related parameters you have at hand that maybe able to help you. 1) Fluxer/flux. Make sure you are within the flux manufactures application specifications when using their flux. Beyond the recommended specifications for amount and thermal aspects is to make sure you are penetrating the barrel. A simple visual method to determine this is to use thermal fax paper. This type of fax paper can be used on alcohol based fluxes. If you are using VOC free you will need to acquire pH paper. The preferred method is to sandwich the paper between two bear boards to prevent lifting of the paper. You can however also tape the paper to the topside of a bear board. It is also preferred that you use the assembly you are having difficulty with. Run the assembly through the fluxer only, do not allow this into the preheat section. You may then quickly remove the paper from the board. You should see a clear defined imprint across all barrels of the assembly. The imprint should be very well defined. Also look at small via's as a process indicator. There are also test vehicles available in the industry. For example, ECD has a product called Flux-O-Meter. Care must also be taken to insure that excessive flux is not being applied to the assembly. Follow the flux manufactures recommendations for amount. most likely this will be between 850-1500 micrograms per square inch of the assembly. If the surface coating of the flux is too much the micro-droplets can not penetrate the barrel as excessive flux on the surface prevents this. This testing tool is just a simple and quick method. It's not intended to be anything more than a visual tool. 2) Considering you have completed section one, move on to the preheat. Again, use the recommended specifications from the manufacture. Most likely this will outline topside laminate temps to be between 210-240 deg F at the exit point of preheat. Consideration for any bottomside SMT being exposed to the wave should also be considered as to not thermally shock any components. Additional care should be taken to not over expose the flux to excessive time in the preheat section, a common mistake when soldering with lead free. Other process indicators of the flux being thermally over exposed will be solder webbing on the assembly mask, large globular icicles, excessive solder and bridging with what appears to be a rough surface. The best way to confirm you profile is to use a thermal profiler with thermocouples attached to both the top and bottomside of the assembly. 3) Wave contact time and immersion depth relates to the assembly thickness. On average you will want to achieve 4-7 seconds of contact time for your average .062" thick assembly. Depending on the type of system you are using care should be used with the use of a chipwave. Some older wave soldering systems had large gaps between the chipwave and the main wave. This gap can cause a lead free alloy to solidify between the two waves. If the gap is in excess of 3-4 inches and your process requires the use of a chipwave you will need to contact the equipment manufacture to see if they have a solution for this. When trying to trouble shoot contact time, some systems offer the user the ability to change the conveyor speed while the assembly is in the conveyor. Considering that the fluxer and preheat parameters are related to the conveyor speed you maybe able to hold the conveyor speed during fluxing and preheating and then just increase or reduce your conveyor speed at the wave. This might save you time in determining the best contact time needed. Of coarse you would then go back to the fluxer and the preheat and adjust those processes accordingly. 4) Alloy type being used. In historical controlled testing done in the past, some alloys have demonstrated a higher wetting ability than others in regards to barrel fill. This is especially true on thicker assemblies in excess of .093". I hope this is helpful, Good luck
Lead free alloys wet more slowly than do tin led alloys, and I suspect that this is the problem, the thicker the board the greater the problem. The problem can also be exacerbated by the pad finish especially if you are using a tin lead OSP. The best thing to do is to increase the preheat temperature (the extent that you can do this will be dependent on the flux that you are using) and to increase the time on the wave (again the extent you can do this will depend on the flux). For 0.093 thick boards I have seen contact times of up to 9 seconds.
Senior Applications Chemist
As with tin/lead wave soldering lack of barrel fill is almost always heat related although poorhole solderability is a possibility. More preheat and/or more time on the wave is probably the answer. Are the holes connected to ground / earth planes or heavy mass components? This would be another symptom of this being related to the thermal mass.
Principal Engineer - CMA Lab
Good soldering needs 3 things: Solder, a "wetable" surface(s), and heat. I suspect you have the solder, so that's not it. "Wetable" surfaces are those that will wet with solder and are free from contamination. I assume the component leads and PCB barrel are of wetable materials like tin or copper (there are others like gold, silver, nickel, etc). This leaves oxides and other contamination which your flux is designed to remove, provided the flux reaches in the barrel and to the top side of the board. You should confirm this. Finally, heat or temperature. Solder will not wet a surface that is not hot enough. You need to run a thermal profile of the area where the barrels do not fill, measuring the top side of the board to make sure it's hot enough for your solder alloy. If it is not, the wetable, flux cleaned, surfaces will not draw the solder up through the barrel because it's simply is not at the right temperature.
Senior Project Engineer
Electronic Controls Design Inc
There are several reasons for insufficient fill of the PTH during wave soldering. The major reason is poor wetting that can be due to oxidation on the leads, oxidation in the barrel, not enough flux, etc. One way to increase the barrel fill is to use an inert atmosphere which will increase the wetting of the solder into the barrel and in turn allow your current flux chemistry to work more efficiently. You can also use a more aggressive flux which will require intensive cleaning post wave. Your PTH plating maybe poor, however this can be reviewed with your board vendor to make sure that your specs are being adhered to.
Global Segment Manager
Air Products & Chemicals, Inc.
George Oxx and I published the results of testing we performed on Lead-Free Intrusive Reflow which showed good barrel fill. This was published in SMT magazine Nov/Dec 2007 and titled "Intrusive Reflow of Lead-Free Solder Paste".
Vice President Technology
Comment: Check solder bathtime and temperature and ensure you use a flow accelerator to improve thesolders ability to flow well. I would always use a Lead Free Solder at 275C pottemperature and generally have an immersion time of 2-3 seconds (yes the same asLEADED). Often extended (over 4 seconds dwell time) will kill most fluxes. Lookfor a typical top board temperature of 105 - 120C. If the solder is ayellow/Gold discoloration and sluggish then please contact us
Technical Sales Manager
BLT Circuit Services Ltd
One of CTQ aspect if not covered are under initial DFM - PTH connected Layer count,weight (oz). We have proved with derived formula with 90%+ success rate on initial assessment of lower fillet height on this. Thermal value PTH= No of spokes X width of spokes X Layer Thickness if greater than 800mil.sqr will sure have problem of fillet height in wave soldering in wave optimized spec. >1500mil.sqr can trigger problem even in selective nozzle type wave soldering.
Supplier Quality Leader
There are two main causes for insufficient holefillMallikarjun, Jabil, India
All good comments. One more, check for soldermask in the thru-holes. Masking can bleed into the holes preventing the solder fromflowing up the holes completely.Jerry Wiatrowski, General Dynamics
I would to add two specific points to this greatdiscussion. Did you check by x-section analysis on PCB plated hole for somesamples to see if you have a proper copper plating thickness (or copper holewall for OSP) inside the hole to be filled by solder? I faced a similarsituation times ago and, for my surprise, after do x-section in three or foursamples of not used PCBs, the hole walls had not enough thickness or, in somecases, simply the hole copper wall did not exist! I do not know who your PCBsupplier is, but you should consider this too. In the case you have a PCB quality issue, as it was supplied, you willnot see the problem in an x-ray image, since most of times you cannot see thedetails of copper/plated walls in a x-ray image. Second, since 2006 when IPCreleased IPC-610D, including lead-free aspects for soldering inspection andsolder acceptance, this subject is controversial. Now we have IPC-610F and, inmy opinion, the IPC-610 committee was not still able to make barrel fillingrequirements fully clear so far.Glayson Figueiredo, Philips Medical Systems, Brazil
With limited information it is hard to pointthe exact reason why this happened. However, based on a project that I havedesignated and completed years ago to address exactly this issue I have foundout that on the 98.7% of the assemblies that exhibited this type of issue, themain root cause was the belt speed – faster/more is not necessarily betterall the time. On the rest of them 1.3% there were some temperature adjustmentsthat were also needed.
Engineering and Operations Management
In addition to ALL of the feedback above that includes poor design, poor thermal profile, contamination and incorrect or old flux, the most important thing is to use an ENIG finish. This is the one thing alone that can insure a higher quality solder joint even in the worst possible manufacturing environment.Jeffrey Lewis, Display Graphics
|Submit A Comment|
Free Newsletter Subscription
Circuitnet is built for professionals who bear the responsibility of looking ahead, imagining the future, and preparing for it.
Insert Your Email Address