Troubleshooting smt-paste

Troubleshooting
SMT Solder Paste
Problems

W.C. Heraeus SMT Applications Training Seminar Module 3: Solder Paste Troubleshooting
Printing Basics
Print speed and Squeegee pressure are inter-related
ƒ Factors effecting speed
ƒ Temperature
ƒ Cold = slow as pastes are very temperature sensitive
ƒ Squeegee Material
ƒ Polymer is faster than metal
ƒ Poly is simply a better squeegee (windshield wipers)
ƒ Poly wears very quickly, not common today
ƒ Sharpness
ƒ Not all metal squeegees are equal
ƒ Plating of Transition Automation is round and dull
ƒ Angle
ƒ 60 degree is faster than 45 degree
ƒ Watch out for MPM; long blades = low angles with high pressures

Various Metal Squeegee Blades
0.0100
0.0075
0.0095 0.0120
Stock DEK E-Blade Print-Perfect DuraGlide
0.0170
Ultraprint

Various Metal Blade Characteristics
Stock DEK E-Blade Print-Perfect DuraGlide Ultraprint
Supplier DEK AMTX
(PhotoStencil)
Transition
Automation
Photo Stencil Alpha Metals
Material Spring Stainless Nickel Spring Steel Spring Steel Spring Steel
Forming Method Machined Electroplated Machined Etched Machined
Coating None None “Permalex”
Teflon/nickel
Titanium Nitride Nickel plate
Body Thickness 7.5 mils 10 mils 9.5 mils 12 mils 17 mils
Tip Thickness (bottom
0.1”)
7.5 mils 10 mils 10 mils 11.5 mils 17 mils
Relative Sharpness** 6 8 4 7 4

F365 Print Speed Benchmarks With Various Squeegees
Stock DEK E-Blade Print-
Perfect
DuraGlide Ultraprint
Maximum speed at 3lb/in 4.25 IPS 3.75 IPS 3.25 IPS 6.00 IPS 2.75 IPS
Minimum pressure at 6IPS N/O N/O N/O 2.7 lb/in 4.3 lb/in
Stock DEK E-Blade Print-
Perfect
DuraGlide Ultraprint
Minimum pressure at 6IPS N/O N/O N/O 5.5 lb/in N/O
45 Degree Squeegee
60 Degree Squeegee

Printing Basics (continued)
ƒ Stencil Roughness
ƒ Rougher better for roll
ƒ Rougher requires more pressure for clean wipe
ƒ Etched or lasered foil is more rougher than EFAB
ƒ Paste factors
ƒ Higher viscosity usually equates to slower print speed and higher pressure.
ƒ Higher tackiness usually equates to slower print speed and higher pressure.
ƒ Squeegee pressure take away message
ƒ “Apply minimum pressure to get a clean wipe, (no solder powder) around the apertures”
ƒ This is independent of all of the other variables effecting squeegee pressure

Stencil Issues
ƒ Not all stencils release equally well
ƒ EFAB is best
ƒ Laser is OK, with electro polish or nickel plate is better
ƒ Chemical etch is worst
ƒ Stencil wiping not totally effective
ƒ Automated better than manual
ƒ Coined solder can remain and cause more needed wiping

Coined solder on bottom of stencil

Sticking to squeegee
ƒ Paste viscosity too high
ƒ Solids too high
ƒ Paste got hot in storage or shipment
ƒ Insufficient paste on the stencil
ƒ Start with 15gr/cm of squeegee length
ƒ Squeegee lift incorrect
ƒ Too high after stroke
ƒ 0.5” to 0.75” is OK
ƒ Too slow
ƒ Fast (>10mm/s) helps shear paste and leave on stencil
ƒ Pause between prints too long
ƒ Paste rheology goes back to rest viscosity
ƒ Short stroke
ƒ AOI too slow

Gasketing is critical for good paste release
ƒ Aperture must be smaller than pad in both length and width
ƒ Critical for fine pitch, small components
ƒ Legend near fine pitch is NG
ƒ Bar code or other labels near fine pitch is NG
ƒ Mask should be < pad height except over traces
ƒ HASL can be a hassle
5 10 15 20 25 30 35 40
Width (mils)
0.8
1.2
1.6
2.0
Height
(mils)
Solder Pad
Solder Mask Wall

1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Height (mils)
Single Pad HASL Anomaly

1.8
1.9
2.0
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
Silk Legend
Mask
Pad
All Dimensions in Mils
Fine Pitch Solder Pad
82C593 Pin 181
Cross Section
Pad
Mask

Compaction (happens only with solder paste)
ƒ ProFlow is DEK
ƒ Compaction in corners and upper head
ƒ Rheopump is MPM
ƒ Compaction is in center (print area)
ƒ Compaction is “pressured separation”
ƒ Flux squeezed out, solids rise, flow inhibited
ƒ Remedies
ƒ Thixatrope level
ƒ Solids
ƒ Powder source
ƒ Cannot be fixed in the process!
ƒ Temporary solution
ƒ Clean out head and replace with fresh material
ƒ Show ProFlow Head

Reflow Basics
Generic Profile
0 60 120 180 240
0
50
100
150
200
250
300
Seconds
45 Sec
219 C
245 C Peak
Preheat Range
1 - 2 C/ sec.
o
o
o
Liquidus

Reflow Basics (continued)
Profiling
ƒ Fast response thermocouples (thin gauge)
ƒ Tip must be welded not twisted to be accurate
ƒ Thermocouple, connectors and lead wires must all be the same type (J,K)
ƒ Type “K” most common, “J” also works
ƒ Tip must be fixed to surface to be measured
ƒ Aluminum tape or high temp solder are best
ƒ Too much Kapton tape (yellow) can “tent” thermocouple and give lower reading
ƒ Not fixed to surface can give too hot a reading
ƒ Best is high temp solder to component leads
ƒ For BGA, bury the thermocouple tip into a center ball (coldest)
ƒ Basic needs for Applications Support
ƒ Peak temp
ƒ Time above liquidus (TAL)
ƒ Time from entrance to peak
ƒ Shape of preheat

Preheat Shape
ƒ Soak or Ramp?
ƒ Soak is less preferred but necessary with certain applications
ƒ Very thick board (back panels, etc.)
ƒ Laminated metal backing or heat spreader
ƒ Assemblies in any type of tray, fixture or jig
ƒ Excessive soak time (>60s) or temp (>150C) can weaken paste activation
ƒ “Cold” looking solder joints Customers Profile

ƒ Ramp is preferred preheat shape in ALL applications where possible and ALL formulations
Soak Ramp

N2 versus O2 reflow Environment
ƒ N2 Benefits
ƒ Wetting
ƒ Lead free components
ƒ Lead free alloys
ƒ Old components
ƒ Oxidized surface finishes (Sn)
ƒ Flux appearance
ƒ More residue spread, appears like less residue
ƒ Lighter color
ƒ Better cleaning (for water cleans)
ƒ Less solder balls
ƒ Helps OSP survive more reflow excursions

ƒ N2 Disadvantages
ƒ Cost
ƒ Not all ovens or facilities are capable
ƒ Some defects are aggravated
ƒ Tombstones
ƒ Solder beads
ƒ Flux mobility
ƒ Flux will move farther from solder joint due to lower surface tension
ƒ Can get down a via to the backside of the assembly
ƒ Can contaminate wire bond finger or connection pad surface

Wetting Issues
ƒ Component solderability
ƒ Lead free components in a Sn63 process
ƒ Extend liquidus time, peak temp or both
ƒ MUST test shows slow wetting time but strong wetting forces
ƒ Extend liquidus time, peak temp or both
ƒ MUST test shows slow wetting time and weak wetting forces
ƒ Add solder volume
ƒ Replace with stronger activation paste
ƒ Use nitrogen reflow
ƒ Have customer seek replacement from component supplier or distributor
ƒ Component Coplanarity
ƒ Add solder height
ƒ Add placement depth
ƒ Replace with stronger activation paste
ƒ Use nitrogen reflow

QFP Coplanarity Example

At placement, a component with some (>0) coplanarity is placed on top of the solder
deposit. The high lead, in these defects, is not touching the solder paste or barely touching the solder
paste. During the preheat portion of the reflow profile the solder paste goes from a firm paste
consistency to a hard deposit. There is a slight shrinkage due to the loss of solvents and other
volatiles but the printed pads of solder paste hold their shape rigidly. So at this point in the reflow
process there is still an air gap between the bottom of the lead and the top of the solder paste. If there
was just barely contact at placement then there most likely is now an air gap going into the reflow
portion of the profile. Without a physical bond of the lead foot bottom surface and the top of the
solder paste there is no or an inadequate thermal link between the surfaces to be soldered (lead,
solder & PCB pad). This thermal link is a basic requirement for the formation of a solder joint that
includes the lead of the component. Very shortly into the reflow portion the solder paste deposit
under the bent lead, without the thermal burden of the lead, will reach liquidus and collapse abruptly
soldering only to the pad. Shortly after that event the rest of the leads will reflow and the component
will collapse towards the board at a rate which is directly proportional to its solderability. Very
solderable parts combined with pads designs wide enough to accommodate the generation of side
fillets will collapse more vigorously. At the moment of this collapse the bent or high lead which was
originally deprived of flux at placement and through pre-heat will push into the molten solder beneath
it and solidify unsoldered as the reflow process ends. These leads may actually have a layer of flux
residue under them but this brief thermal link came too late in the reflow process to initiate the
soldering process to the lead foot.
From a solder paste standpoint the firmer the deposit the worse, this can be aggravated
by higher solids that also presents less wetness to the solder deposit which is critical to form that
early thermal bond that is critical. Increased placement depth and reduced component coplanarity
are the most powerful elements for elimination of this defect type. With most of the lead finishes
moving to lead free (Sn) these days, solderability will be compromised. A higher peak temp and/or a
longer liquidus (~90s) can help solderability with pure tin plated leads.
Foot-in-mud, Ball-in-socket, Head-in-pillow defect formation theory

ƒ Gold Plating too thick
ƒ Bulk solder looks “cold”
ƒ Solder joints have massive voiding
ƒ Solder joint is extremely brittle
ƒ > 3-5% gold in bulk solder

Gold: 0.5µ - 1.0µ Gold: 3.0µ
Customer Spec: 0.4µ - 0.9µ

Surface finish effects
ƒ ENIG is simply best
ƒ Wetting
ƒ Voiding
ƒ Defects
ƒ IPC-4552 covers this finish specifications
ƒ Electroless Nickel: 120 -240µin (3-6 microns)
ƒ Immersion Gold: 3-5µin (0.075-0.125 microns)
ƒ Immersion silver is second best
ƒ If thickness is too thick, Champaign voiding is inevitable
ƒ OSP and Immersion tin are third
ƒ OSP = limited reflow excursions
ƒ Concern of shelf life and variability with Immersion tin

2.5
4.5 4.4 4.3 4.2 3.9 3.6
Air 800 ppm 400 ppm 200 ppm 100 ppm 50 ppm 25 ppm
Reflow Atmosphere
0
5
10
15
20
Wetting
Wetting D Wetting C Wetting B Wetting A
Wetting
Entek (Cu)
690
1147
1228 1232
1149
1227 1233
Reflow Atmosphere
0
200
400
600
800
1000
1200
1400
Wetting
Wetting
Ni/Au
3.5
7.6 7.9 8.2
7.3 6.7 6.7
Reflow Atmosphere
0
5
10
15
20
Wetting
Wetting
Sn
4.5
10.8
11.8
16.7
10.8 10.5
12.5
Reflow Atmosphere
0
5
10
15
20
Wetting
Wetting
Ag

Cu
Sn
Au
Ag
Dewet

Cu
Sn
Au
Ag

171
132 125 130
117
150
129
Reflow Atmosphere
0
50
100
150
200
250
300
Solder
Defects
Tombstones Solder Beads Solder Balls
Solder Defects
Entek (Cu)
134
114 106 107 115
136
113
Reflow Atmosphere
0
50
100
150
200
250
300
Solder
Defects
Solder Defects
Ni/Au
253
221 215
248
202
226
195
Reflow Atmosphere
0
50
100
150
200
250
300
Solder
Defects
Solder Defects
Sn
131
120
143 133
110
126
82
Reflow Atmosphere
0
50
100
150
200
250
300
Solder
Defects
Solder Defects
Ag

0.5
-11 -14
-34
6
-13
-6
31
44 50
19
4
-9
41
Reflow Atmosphere
-100
-50
0
50
100
Points
X-Ray Points Preforms X-Ray Points BGA
X-Ray Voiding
Entek (Cu)
60
92
63 67 71 78 88
70
59
75
24
67
39
70
Reflow Atmosphere
-100
-50
0
50
100
Points
X-Ray Voiding
Ni/Au
-43 -41
-26 -29 -25
-43
-26
-7
-18 -12
7
-36
-46
-8
Reflow Atmosphere
-100
-50
0
50
100
Points
X-Ray Voiding
Sn
-42
4
15
-52
-13
-29
-14
-94 -90 -96 -98 -98 -97 -94
Reflow Atmosphere
-100
-50
0
50
100
Points
X-Ray Voiding
Ag

Immersion Tin Immersion Gold

Immersion Silver Immersion Gold
Copper Preform
No Preform

Old Board Lot New Board Lot

Common Reflow Defects
ƒ Solder Beads
ƒ Higher solids paste can help
ƒ Paste with low or no hot slump can help
ƒ Air reflow (if nitrogen) can help
ƒ Home plate aperture design (effective)
ƒ Crown aperture design (most effective)

10% Inside Reduction Home Plate Crown
X-Rays after placement but before reflow

Preferred Solution

ƒ Solder Balls
ƒ Nitrogen reflow can minimize
ƒ Poor miss-print board cleaning?
ƒ Look for solder balls in the vias
ƒ Water clean
ƒ Pre cleaning process solder balls are normal
ƒ Poor stencil to pad gasketing?
ƒ Profile not optimal
ƒ Ramp-to-spike preferred

ƒ Tombstones
ƒ Paste not registered pads properly
ƒ Component not centered on pads
ƒ Inter-pad gap too big (common in European designs)
ƒ Shrink gap
ƒ One pad tied to thermal drain such as a via
ƒ Variable component end term solderability
ƒ Air reflow (assumes nitrogen reflow)
ƒ Anti-tombstone alloy
ƒ Blurs reflow event to help form both solder joints together
ƒ Post reflow alloy is the same

ƒ Voiding
ƒ Voiding is mainly due to expanding gas during reflow (liquidus)
ƒ Water vapor
ƒ Solvents
ƒ Gaseous reaction byproducts
ƒ Main source is paste formulation
ƒ Generally more active formulations = more voiding
ƒ Generally Sn63 is better than Lead Free alloys
ƒ Secondary source is surface finish
ƒ ENIG is best, OSP is worst
ƒ Tertiary source is component type
ƒ BGA's give central large void
ƒ Leadless and SOT devices have large parallel surfaces to trap voids
ƒ Last (and least powerful) is profile
ƒ Ramp is better than soak in preheat

Cleaning Basics
Most no cleans are “Not Cleans”
Water Clean basics
ƒ Water should be de-ionized (DI)
ƒ Water should be heated (120-160F)
ƒ Impinging spray is best
ƒ Gets under components
ƒ Adds mechanical cleaning energy
ƒ Drying is last step in cleaning process
ƒ Certain high surface area dark solder masks will retain “flux stains”
ƒ Cosmetic and very difficult to remove
ƒ Nitrogen reflow helps
ƒ Cleaning right after reflow helps
ƒ Lowering peak reflow temp helps
ƒ Clean after every reflow !

Cleaning Basics (continued)
Case Study: SiP Module “White Ring”

Cleaning Basics (continued)
ƒ SiP reflow/clean experiment conclusions
ƒ White ring visible with both light microscope and electron microscope
ƒ White ring is tin oxide
ƒ Nitrogen minimizes ring
ƒ All test coupons pass ion chromatography
ƒ Blank boards failed for calcium ion
ƒ White ring is cosmetic

Lead Free Basics
Common issues
ƒ Printing is equal to or better than tin/lead
ƒ Hot slump is typically more than tin/lead
ƒ Wetting can be less than tin/lead
ƒ Nitrogen reflow is helpful
ƒ Surface finish of board very powerful
ƒ Voiding can be more than tin/lead

Lead Free Basics (continued)
ƒ RoHS Challenges
ƒ “We cannot defend our position as RoHS compliant from a customers solder joint”
ƒ Amount of material is too small for a quantitative assay
ƒ Solder in solder joint is a blend of board finish, paste and component finish
ƒ Ownership of the assembly process is the customers.
ƒ “Last word” testing is an ICP assay of bulk metal
ƒ Requires 5-10grams reflowed in clean crucible
ƒ Hand held XRF is useful and accurate for paste samples
ƒ Borderline readings must be refereed by ICP
ƒ SEM EDS is only qualitative

Troubleshooting smt-paste

Recommended

Recommended

More Related Content

What's hot

What's hot (20)

Similar to Troubleshooting smt-paste

Similar to Troubleshooting smt-paste (20)

Recently uploaded

Recently uploaded (20)

Troubleshooting smt-paste