Short answer: Uneven current appears when 0BB wire contacts do not have similar resistance. First check whether the EL defect follows one wire, whether it appears only after lamination, whether the cell height changes at the same position, and whether fill factor or series resistance has worsened. For a 24-wire cell, checking the top, center and bottom of every wire gives 72 alignment readings. Combine those readings with a nine-point flatness map, local resistance measurements and a cross-section from the same defect location.
Locate the Electrical Loss
The current path in a typical 0BB cell is:
Active cell area → fine finger → wire contact → copper interconnect → adjacent cell
Most 0BB designs remove continuous printed busbars, although some retain small pads or local solder areas. The structural differences are shown in a published comparison of MBB and 0BB solar cell technologies.
Fraunhofer ISE describes round interconnection wires with copper cores commonly ranging from 200 to 400 µm. The wires may be fixed by soft soldering or conductive adhesive.[1]
| Resistance Source | Where It Occurs | What It Changes |
|---|---|---|
| Finger resistance | Along the printed metal finger | Loss increases when current must travel farther to another wire |
| Contact resistance | Where the wire meets the finger, pad or conductive surface | Creates local voltage drop and uneven current sharing |
| Wire resistance | Along the copper interconnect | Raises loss when one wire carries more current than nearby wires |
Voltage drop = Current × Resistance
Power loss = Current² × Resistance
If current through a working contact rises from 1.0 A to 1.2 A while resistance stays unchanged, its heat loss rises by 44%:
1.2² ÷ 1.0² = 1.44
Cells in one series string still carry the same total string current. The difference is the current path inside each cell, the local voltage drop and the amount of resistive loss.
Check Every Wire Crossing
A wire can look correctly positioned from above while making poor contact underneath.
| Condition | What to Inspect | Expected Result |
|---|---|---|
| Wire above the finger | Local cell height, wire height and cross-section | Gap or incomplete compression below the wire |
| Wire touches one finger edge | Top, center and bottom alignment | Reduced overlap at one or more positions |
| Solder remains around the wire | Cross-section through a dark EL area | Incomplete solder spread onto the intended surface |
| Adhesive enters the contact | Adhesive width, spread and cured thickness | Insulating or high-resistance material below the wire |
| Particle below the wire | Surface inspection and cross-section | Wire lifted across one or several fingers |
| Finger below the wire is broken | Incoming-cell EL and finger continuity | Longer current path even when the wire is correctly placed |
One weak crossing may be supported by nearby contacts. Several neighboring weak crossings are more serious. If current normally reaches a wire after one wire pitch but must cross four pitches because three nearby contacts are weak, the finger collection distance becomes about four times longer in that area. This is a geometry example, not a universal resistance multiplier.
Measure Wire Alignment
Measure every wire near the upper edge, at the cell center and near the lower edge.
24 wires × 3 positions = 72 alignment readings per cell
Record these values:
- Distance from the intended contact line
- Wire-to-wire spacing
- Wire curvature
- Contact overlap near both cell edges
- Position before and after lamination
Use the defect pattern to narrow the cause:
- The same wire number is dark across many cells: check stringer alignment, wire tension and fixation.
- Different wire numbers fail randomly: check particles, cell shape and local solder or adhesive contact.
- The wire moves only after lamination: check encapsulant flow, temporary fixation and lamination pressure.
- The wire is visibly shifted but EL remains normal: the movement may not be causing meaningful electrical loss.
Check Solder and Adhesive
For soldered contacts, inspect whether the solder:
- Reaches the intended finger or conductive surface
- Spreads around the wire rather than remaining only on top
- Contains large voids
- Separates from the metallization
- Pulls printed paste from the cell
- Forms consistently at the center and edges
A 2024 HJT 0BB study compared soldering-based interconnection with an Integrated Film Covering process. In that study's test structure, the soldering-based samples had average contact-resistance values 161 mΩ higher on the front and 444 mΩ higher on the rear.
The calculated resistive-loss difference was 2.89 W. The measured average maximum-power difference between two groups of three modules was 3.55 W, including an estimated 1.23 W optical contribution. The measured fill-factor difference was 0.27 percentage points.[2]
These figures apply only to the tested HJT structure, materials and calculation method. They are not general 0BB acceptance limits.
For adhesive-assisted contacts, record:
- Dispensed line or dot width
- Distance from the electrical crossing
- Cured adhesive thickness
- Spread after lamination
- Electrical resistance before and after ageing
High pull strength does not prove low contact resistance. Adhesive can hold the wire firmly while the electrical interface remains poor.
Map the Laminator
Divide the usable laminator area into nine zones:
- Four corner zones
- Four edge zones
- One center zone
Compare these values by zone:
- Pressure distribution
- Actual stack temperature
- Vacuum behavior
- Wire movement after lamination
- Contact resistance
- EL defect rate
Nine zones are a practical troubleshooting layout, not an IEC requirement.
Check these common sources of pressure variation:
- Worn laminator membrane
- Debris on the platen
- Folded or overlapping encapsulant
- Glass thickness variation
- Tape buildup
- Cell bow
- Ribbon crossings
- Uneven stack thickness
The laminator setpoint is not the temperature at every wire crossing. Place temperature sensors inside a representative stack and record:
- Time needed to reach the working range
- Time within the working range
- Maximum stack temperature
- Center-to-edge difference
- Heating and cooling rates
Encapsulant must fill gaps without moving the wires. Film thickness, viscosity, storage condition, heating rate and curing behavior all affect this movement. A discussion of encapsulant quality in solar modules provides more detail on film curing and layer adhesion.
Map Cell Flatness
Use a nine-point height map:
- Four corners
- Four edge midpoints
- One center point
Measure the same points at four stages:
- Before wire placement
- After wire attachment
- Before lamination
- After lamination and cooling
9 points × 4 stages = 36 height readings per sampled cell
Add a line scan along the affected wire when the EL defect falls between the nine standard points.
Record:
- Maximum height difference
- Center bow
- Edge lift
- Cell twist
- Shape change after wire attachment
- Residual shape after cooling
Use the same support method for every measurement. A freely supported cell and a cell held on a vacuum chuck can show different shapes.
Rule Out Cell Defects
Before blaming the wire contact, check for:
- Broken or interrupted fingers
- Thin or narrow printed sections
- High-resistance metallization paste
- Poor paste adhesion
- Cell cracks
- Uneven transparent conductive oxide, or TCO
- Dust, oil, silicone or adhesive residue
Several neighboring finger breaks can create an EL pattern similar to a weak wire contact. Mark the real wire direction, finger direction and crack position before deciding which part failed.
Use Controlled EL Images
Capture two controlled current levels before and after lamination:
2 current levels × 2 production stages = 4 comparable EL images
Keep these conditions unchanged:
- Injection current for each test level
- Module temperature
- Camera exposure
- Camera gain
- Lens position
- Ambient light
- Image correction
IEC TS 60904-13 covers module EL image capture, processing and interpretation under forward bias.[3] IEC TS 63202-2 covers EL inspection of non-encapsulated crystalline-silicon cells.[4]
| EL Result | Check Next |
|---|---|
| Defect already exists before interconnection | Cell crack, finger continuity, TCO and metallization |
| Defect appears after wire attachment | Wire position, placement force and contact formation |
| Defect appears only after lamination | Pressure, temperature, encapsulant flow and cell height |
| Defect grows after ageing | Joint fatigue, residual stress, corrosion and adhesive stability |
An article on EL testing before module installation shows how cracks, inactive areas and interconnection defects can appear in module inspection.
Compare EL with PL
Photoluminescence, or PL, excites the cell with light instead of passing current through the completed module contacts.
| Image Result | More Likely Direction |
|---|---|
| Dark in EL but relatively normal in PL | Local series resistance in the wire contact, finger, TCO or another conductive layer |
| Abnormal in both EL and PL | Crack, inactive cell area, recombination or material damage |
NREL researchers used differences between EL and PL to investigate local resistance changes in field-aged silicon heterojunction modules.[5]
Check I–V and Heat
Poor wire contact normally increases series resistance.
| I–V Change | What It Suggests |
|---|---|
| Lower fill factor | More resistive loss under operating current |
| Lower maximum power | Electrical or optical loss has increased |
| Open-circuit voltage stays nearly unchanged | Contact loss may be present because little current flows at open circuit |
| Short-circuit current changes little | Enough parallel contacts may still be working |
Fill-factor loss alone does not prove a wire-contact problem. Broken fingers, interconnect ribbons, junction-box contacts and other resistance sources can produce similar results.
IEC 60904-1 covers I–V measurement for cells, cell assemblies and modules.[6]
Use infrared imaging to find local heat. A fully open contact may remain cool because little current passes through it. Nearby working contacts may become hotter because they carry the redirected current.
A peer-reviewed dark lock-in thermography study linked failed solder joints with current crowding and point-like heating in nearby conductive regions.[7]
Measure Local Resistance
Use a four-wire, or Kelvin, method where possible. This reduces the effect of probe and lead resistance.
Control:
- Probe position
- Probe pressure
- Cell or module temperature
- Contact spacing
- Measurement direction
- Fixture repeatability
A probe can temporarily press a weak contact against the cell and make the measured resistance look lower than it is inside the laminate.
Sample contacts from:
- The front and rear
- The cell center and edges
- The first, middle and last wires
- Different stringer lanes
- Different laminator zones
- Before and after ageing
Fraunhofer ISE uses a GridTouch-based setup with 10, 20 or 30 wires so a busbarless-cell measurement can better match the intended module layout.[8]
Published research also shows that the temporary contact arrangement can change measured fill factor and series-resistance-related results in busbarless cells.[9]
Do not report only the average. Five example results of 8, 8, 8, 8 and 18 mΩ have an average of 10 mΩ, but the 18 mΩ contact has 2.25 times the resistance of each 8 mΩ contact. These numbers are an example, not an acceptance limit.
Match Pull Force with Resistance
Record average, minimum, maximum, range and failure location.
Five example pull results of 1.1, 1.0, 1.0, 1.0 and 0.4 in the same relative force unit have an average of 0.9. The average hides one joint that is less than half the strength of most samples.
| Resistance | Pull Strength | Likely Meaning |
|---|---|---|
| Low | High | Good initial electrical and mechanical result |
| High | High | Strong fixation but poor electrical interface |
| Low | Low | Good initial contact with weak mechanical support |
| High | Low | Possible incomplete, contaminated or damaged joint |
Record whether failure occurs inside the solder, adhesive or printed paste, or at an interface between two materials.
Cut the Defect Location
Do not cut random contacts. Mark the dark EL position before sectioning.
Check for:
- Direct wire-to-finger contact
- Solder spread
- Void size
- Adhesive below the wire
- Wire deformation
- Finger damage
- Layer separation
- Cracks below the contact
Compare at least one dark EL area with one bright area from the same cell. EL, resistance, height and cross-section results must refer to the same physical crossing.
Use the Defect Pattern
| Observed Pattern | Check First | Likely Cause |
|---|---|---|
| Dark band following one wire | Wire position and contact continuity | Shifted, lifted or poorly bonded wire |
| Dark area between two wires | Finger continuity and nearby crossings | Broken fingers or several weak contacts |
| Dark edges on many cells | Local cell height and edge pressure | Cell bow, edge lift or low pressure |
| Random dark spots | Surface cleanliness and cross-section | Particles, adhesive intrusion or poor solder wetting |
| Defects concentrated in one laminator zone | Pressure and temperature maps | Equipment or stack variation |
| Normal before lamination, abnormal afterward | Wire movement, pressure and contact formation | Lamination or encapsulant problem |
| Normal before cycling, abnormal afterward | Resistance, pull force and cross-section | Joint fatigue or residual stress |
| Hot points beside a dark area | Nearby open or resistive contacts | Current crowding into working joints |
Track Production Data
Group results by:
- Stringer and lane
- Laminator and platen zone
- Production shift
- Cell batch
- Wire batch
- Solder or adhesive batch
- Encapsulant batch
If 12 of 600 inspected cells show the same dark band:
12 ÷ 600 × 100% = 2.0% defect rate
If 10 of those 12 defects come from one stringer lane:
10 ÷ 12 × 100% = 83.3% of the defects from one lane
The 2.0% total defect rate shows the scale. The 83.3% lane concentration gives the stronger clue.
A high-efficiency module range includes different 0BB formats. Inspection limits should therefore be linked to the exact model, cell structure and bill of materials.
Set Model-Specific Limits
Do not use one limit for every 0BB design. Establish separate limits for:
- EL brightness variation
- Wire movement
- Contact resistance
- Pull force
- Cell bow and twist
- Lamination pressure
- Stack temperature
Build each limit from:
- Known-good modules
- Known-failed modules
- Measurement repeatability
- Ageing results
- Power and field-performance correlation
Track the minimum, maximum, range and outliers, not only the average.
Check Ageing Results
During thermal, humidity and mechanical testing, record:
- Growth of EL dark areas
- Change in contact and total series resistance
- Fill-factor loss
- Minimum pull force
- Wire lift or movement
- Crack growth near contacts
- New thermal patterns
IEC 61215-1-1 covers design qualification for crystalline-silicon modules, but the standard states that qualification results are not a quantitative prediction of module lifetime.[10]
Compare intermediate EL and resistance results rather than checking only the initial and final power values. A guide to comparing solar module reliability also explains why test conditions and bills of materials must match before degradation figures are compared.
Finally
Start with location-matched evidence. For a 24-wire cell, collect 72 alignment readings and 36 flatness readings, then compare four controlled EL images taken before and after lamination. A repeated defect on the same wire points to alignment, tension or bonding. A repeated defect in the same laminator zone points to pressure, temperature or stack thickness. Random defects point more often to particles, broken fingers or isolated contact gaps. Confirm the cause with local resistance and a cross-section from the same point. Published HJT testing shows that contact differences can add 2.89 W of calculated resistive loss, so small defects should not be judged only by visual appearance.