Updated April 2026 — What the cell grading system actually means, why one seller's "A-grade" is another's "B-grade," and how to choose the right grade for your budget and project.
If you've shopped for SunPower IBC cells on the secondary market, you've seen the labels: A-grade, B-grade, cosmetic B, cut cells, pulls, surplus. The problem is that these terms aren't standardized. One seller's A-grade might match another's B-grade, and some sellers simply make up the label. This guide explains how cells are actually graded at the factory level, what defects drop a cell from A to B, and which grade makes sense for which application.
How Cells Are Graded at the Factory
After wafer fabrication, every cell goes through an inline tester that measures electrical parameters at STC (Standard Test Conditions): Voc, Isc, Vmp, Imp, fill factor, and efficiency. It also passes through an electroluminescence (EL) imager that reveals microcracks, finger interruptions, and contact defects invisible to the eye.
Based on this data, the manufacturer sorts cells into bins:
- A-grade: Meets all datasheet specifications. No visible defects. EL image clean. These go into branded modules sold with full warranties.
- B-grade: Slightly below spec on one or more parameters — typically 1–3% lower efficiency, or minor cosmetic defects like edge chips, slight color variation, or small contact interruptions that don't significantly affect output. These may be sold to third-party assemblers or enter the secondary market.
- C-grade / scrap: Clearly defective — cracked, severe discoloration, or electrical parameters far below spec. Usually recycled or sold to recyclers for silicon recovery.
The key point: the factory grading is based on electrical performance, not just looks. A cell with a tiny edge chip but full electrical output is still A-grade for most manufacturers. A cell that looks perfect but has a 5% efficiency drop is B-grade.
A-Grade Cells: When You Need Consistency
A-grade IBC cells are the ones that met every parameter on the manufacturer's datasheet. If you're building a panel where every cell must perform identically — a research experiment, a commercial product, or a module you plan to sell — A-grade is the only choice.
What you get with A-grade:
- Documented electrical parameters (Voc, Isc, Vmp, Imp, FF) that match the datasheet within ±2%
- No microcracks visible on EL imaging
- Uniform color and surface appearance
- Tighter batch-to-batch consistency for series string matching
The trade-off is price. A-grade cells cost 30–50% more than B-grade on the secondary market. For a one-off DIY project, that premium might not be justified. For a thesis experiment where data quality matters, it is.
B-Grade Cells: The Hidden Value
B-grade cells are the industry's best-kept secret for budget-conscious builders. The defects that relegate a cell to B-grade are often irrelevant for small custom panels:
| Defect type | Typical impact on output | Relevance for DIY/small builds |
|---|---|---|
| Edge chip (<2 mm) | 0–1% power loss | Negligible — the cell still works at near-full output |
| Color variation (light/dark patches) | 0–2% power loss | Cosmetic only; doesn't affect electrical performance |
| Minor finger interruption (rear contact) | 1–3% power loss | Slight reduction, often unnoticeable in single-cell measurements |
| Efficiency 1–2% below datasheet | 1–2% power loss | For a 4W cell, that's 0.04–0.08W — usually acceptable |
| Small scratch on anti-reflective coating | <1% power loss | Cosmetic; irrelevant unless you're doing optics research |
We've used B-grade cells in our own custom mini panel production where the slightly lower efficiency is offset by significantly lower cost. The key is testing: every B-grade cell should still be individually checked for Voc and Isc before it's built into a product. A "B-grade" label from the factory doesn't mean "bad" — it means "not perfect." Some B-grade cells test at 98% of A-grade output.
Cut Cells: A Different Category
Cut cells aren't a grade — they're a form factor. Full cells are diced into smaller pieces for specific applications. The cutting process itself introduces a yield loss: not every full cell survives being cut. Some develop microcracks at the cut edge that propagate into the active area.
Reputable suppliers test cut cells after dicing and discard the ones that cracked. Less reputable suppliers ship everything and let the buyer discover the failures. When buying cut cells, ask:
- Was the cell tested after cutting?
- What percentage of cut pieces fail quality control?
- Are failed pieces replaced, or do you pay for them?
From our cutting operation, roughly 5–8% of diced pieces develop edge cracks that reduce output below acceptable thresholds. We discard these and only ship pieces that pass a post-cut Voc test. If a supplier quotes a cut price with no mention of yield or testing, assume you're paying for the failures too.
Pulls and Decommissioned Cells
"Pulls" are cells removed from decommissioned modules — often from warranty replacements, storm damage, or end-of-life recycling. They're not graded by the original manufacturer; they're graded by whoever pulled them.
The risk with pulls is degradation. A cell that's been in a module for 5–10 years has experienced thermal cycling, UV exposure, and potential moisture ingress. Its output may be 10–20% below the original datasheet values. The cell might still work fine for a low-stakes DIY project, but it's not suitable for research or products where you need known parameters.
How to identify pulls:
- Evidence of prior soldering on the rear contacts (residual solder, flux stains)
- EVA residue on the edges from prior encapsulation
- Slightly lower Voc than new cells from the same platform
- Sometimes, the original module's serial number is laser-etched nearby
How to Verify Grade Claims Before Buying
| Seller claim | What to ask for | Red flags |
|---|---|---|
| "A-grade" | Individual cell test data; EL images | No data provided; price suspiciously low |
| "B-grade" | Specific defect list; sample test data | Defects not described; cells look salvaged |
| "Cut cells" | Post-cut test data; yield percentage | No post-cut testing; rough, unpolished edges |
| "New surplus" | Date code; factory traceability | No date code; mixed lots in one shipment |
| "Matched set" | Tolerance specification (±X%) | No tolerance stated; measured values vary wildly |
Pro tip: If a seller claims "A-grade" but the price is within 10% of B-grade pricing, you're probably looking at relabeled B-grade cells. Genuine A-grade commands a clear premium because the supply is smaller and the testing overhead is higher.
Which Grade for Which Project?
| Project type | Recommended grade | Why |
|---|---|---|
| University research / thesis | A-grade, matched set | Reproducible data; documented parameters for publications |
| Commercial product prototype | A-grade | Consistent performance for customer expectations |
| Personal DIY panel | B-grade or tested cuts | Cost-effective; slight output variation doesn't matter |
| Educational demo / classroom | B-grade or pulls | Low cost; students learn on cells where perfection isn't required |
| IoT / embedded prototype | Tested custom cuts | Size matters more than peak efficiency |
| Module assembly for resale | A-grade, matched | Warranty implications; customer performance expectations |
Need graded cells with test data? We stock A-grade and B-grade 125 mm and 166 mm IBC cells, all tested for Voc and Isc before shipping. Custom cuts available with post-dice verification.
FAQ
Can B-grade cells be used in series strings?
Yes, but with caution. The key is matching current output. If your B-grade cells vary by ±5% in Isc, the lowest-current cell limits the entire string. For best results, measure Isc on every B-grade cell and group them into matched strings. Don't mix a 5.5A cell with a 6.2A cell in the same series string — the 5.5A cell acts as a bottleneck.
Do cut cells have lower efficiency than full cells?
The silicon itself doesn't change efficiency when cut. However, the cut edge can introduce defects (microcracks, shunts) that reduce the effective output of that specific piece. A well-cut, tested piece should deliver output proportional to its remaining area. A 50% cut should deliver ~50% of the full cell's power. If it delivers significantly less, the cut damaged the cell.
How long do pulls last compared to new cells?
It depends on how the original module was treated. Cells from a module that failed due to junction box corrosion (common after 10+ years) may still have 85–90% of original output and decades of remaining life. Cells from a module damaged by hail or impact may have hidden microcracks that worsen over time. Test pulls individually and discard any with Voc below 0.60V or visible crack patterns on EL imaging.
Is there such a thing as "A+ grade"?
No. The factory grading systems used by Maxeon and major manufacturers don't include an A+ tier. "A+" is a marketing term used by some secondary sellers to justify higher prices. It has no standardized meaning. If a seller claims A+, ask them to define it — and provide the test data to back it up.
