CN3085 Design Notes — Dual AA NiMH Charger
Scope
This file records the key design rationale and calculations discussed for the USB-C powered 2×AA NiMH charger using the CN3085. It is intended as a reference for future schematic/layout decisions. No schematic or PCB layout changes are implied by this document.
User Preferences / Workflow Constraints
- Do not modify the schematic unless explicitly instructed.
- Do not modify the PCB layout unless explicitly instructed.
- Calculations and design rationale should be saved to project files when they affect future design choices.
Charger Architecture Decision
The design uses a standalone CN3085 linear NiMH charger rather than a switching charger or MCU-controlled solution.
Reasons:
- The target battery format is 2×AA NiMH, which is common in consumer devices.
- A 3-cell pack would reduce the voltage drop from USB 5V and improve thermal efficiency, but is less practical because most devices use two cells.
- A 4-cell pack would be close to the available USB 5V input headroom near full charge and is not a good fit.
- Standalone switching NiMH charger ICs are scarce, often old/EOL, expensive, unavailable, or require significant external circuitry.
- MCU-based charging can be flexible but adds firmware/labor cost.
- The CN3085 is simple, low-cost, and standalone, but thermally inefficient because it is a linear charger.
Key CN3085 Behavior Used in Calculations
- USB input assumed: 5.0V.
- Fast-charge current is set by RISET.
- Precharge current is approximately 10% of fast-charge current.
- Maintenance charge current is approximately 60% of fast-charge current.
- For the 1.45V/cell maximum-voltage design procedure, the datasheet indicates roughly 40% of capacity is delivered before maintenance mode and the remaining 60% is delivered during maintenance charge.
- The maintenance timer therefore simplifies to:
because:
T = (0.6 × C_BAT) / (0.6 × I_CH)
Two-Cell Voltage Divider
The two-cell divider was selected around the CN3085 recommended ratio for 1.45V/cell operation.
Original ideal values:
- R2 = 100kΩ
- R1 ideal ≈ 140.7kΩ
- Divider ratio ≈ 1 + R1/R2 ≈ 2.407
Practical sourcing note:
- 140.7kΩ is not a common stocked value.
- A practical stocked substitute is 140kΩ, 1%, 0603.
- The resulting error is small relative to typical resistor tolerance and acceptable for this design.
Updated lower-impedance divider choice selected for the schematic:
- R1 = 51kΩ, 1%, 0603
- R2 = 36kΩ, 1%, 0603
Verification of this ratio:
R1/R2 = 51kΩ / 36kΩ ≈ 1.4167
Divider multiplier = 1 + R1/R2 ≈ 2.4167
Compared with the original ideal multiplier of approximately 2.407, this is about 0.40% high:
Error = (2.4167 - 2.407) / 2.407 ≈ 0.40%
Using the same CN3085 FB reference assumptions as the earlier calculation, this would move the approximate two-cell maximum-voltage point from about 2.90V pack to about 3.07V pack:
V_BAT,max ≈ 1.205V × 2.4167 ≈ 2.91V
V_CELL,max ≈ 2.91V / 2 ≈ 1.46V/cell
This note records the selected lower-impedance divider values. The ratio is only slightly above the original ideal target and is acceptable for this design, subject to normal tolerance review.
Precharge Threshold and Worst-Case Fast-Charge Entry Voltage
Using the CN3085 precharge FB threshold discussed:
With the selected 51kΩ / 36kΩ divider ratio:
Pack voltage at precharge exit:
V_BAT = 0.843V × 2.4167 ≈ 2.04V
Per-cell voltage:
V_CELL = 2.04V / 2 ≈ 1.02V/cell
This ~2.04V pack voltage is the key worst-case thermal point, because the charger transitions from low-current precharge into full fast charge while the battery voltage is still low.
RISET Calculation Examples
CN3085 fast-charge current equation used:
For 450mA:
R_ISET = 1218 / 0.45 ≈ 2706.7Ω
Practical value:
This gives:
I_CH ≈ 1218 / 2700 ≈ 451mA
Maintenance Timer Calculation
CN3085 timer equation used:
T = 2654 × R5 × C1 + 4980 × C1 × 10^3
Solving for R5:
R5 = (T/C1 - 4,980,000) / 2654
Case A — 2400mAh Battery, 450mA Charge Current
Timer target:
T = 2400mAh / 450mA = 5.33h = 19,200s
With C1 = 10µF effective:
R5 = (19200 / 10e-6 - 4,980,000) / 2654
R5 ≈ 721.6kΩ
Practical value:
R5 = 720kΩ
C1 = 10µF effective
Verification:
T = 2654 × 720000 × 10e-6 + 4980 × 10e-6 × 1000
T ≈ 19,158.6s ≈ 5.32h
Important capacitor note:
- C1 must be the effective capacitance at the RC pin operating bias, not just nominal capacitance.
- Very small high-value MLCCs may derate heavily under DC bias.
Case B — 2400mAh Battery, 1000mA Charge Current
Maintenance current is 60% of fast-charge current:
I_MAINT = 0.6 × 1000mA = 600mA
Remaining capacity after fast-charge phase:
C_REMAINING = 0.6 × 2400mAh = 1440mAh
Timer target:
T = 1440mAh / 600mA = 2.4h = 8640s
Equivalently:
T = 2400mAh / 1000mA = 2.4h
With C1 = 10µF effective:
R5 = (8640 / 10e-6 - 4,980,000) / 2654
R5 ≈ 323.7kΩ
Practical value:
R5 = 324kΩ
C1 = 10µF effective
Smaller Timing Capacitor Option for 1000mA Case
To reduce DC-bias derating and cost, a smaller nominal timing capacitor in a larger package can be preferable.
For C1 = 4.7µF effective and T = 8640s:
R5 = (8640 / 4.7e-6 - 4,980,000) / 2654
R5 ≈ 688kΩ
Practical values:
R5 = 690kΩ, 1%
C1 = 4.7µF effective, X7R, larger package preferred
Verification:
T = 2654 × 690000 × 4.7e-6 + 4980 × 4.7e-6 × 1000
T ≈ 8680s ≈ 2.41h
This is a good practical timing choice for a 2400mAh battery when I_CH = 1A, provided C1's effective capacitance is close to 4.7µF at the RC pin bias.
Thermal Limitation of the Linear Charger
The major weakness of the 2×AA-from-USB design is the voltage difference between USB 5V and the low two-cell pack voltage.
Worst-case fast-charge entry voltage:
Without added series resistance, charger IC power at 1A:
P_IC = (5.0V - 2.03V) × 1A ≈ 2.97W
This is likely too high unless the PCB thermal path is excellent. The CN3085 will likely enter thermal regulation at high current if heat is not spread effectively.
Thermal Resistance Assumption Examples
Assuming:
T_A = 25°C
T_J_LIMIT ≈ 135°C
ΔT_ALLOWED = 110°C
If effective thermal resistance is 100°C/W:
P_MAX = 110°C / 100°C/W = 1.1W
I_MAX = 1.1W / (5.0V - 2.03V) ≈ 370mA
If effective thermal resistance is 50°C/W:
P_MAX = 110°C / 50°C/W = 2.2W
I_MAX = 2.2W / (5.0V - 2.03V) ≈ 740mA
Heat-Spreading Series Resistor Concept
A series resistor can shift some dissipation away from the CN3085 into external resistor(s), similar to the approach shown in TP4056-style linear charger guidance.
General resistor equation used:
R_S = (V_USB - V_BAT - P_IC,max / I_CHG) / I_CHG
For 1A charge current, 50°C/W IC thermal assumption, and P_IC,max = 2.2W:
R_S = (5.0V - 2.03V - 2.2W/1A) / 1A
R_S ≈ 0.77Ω
A practical design option considered:
Two 2Ω resistors in parallel → R_S = 1Ω total
At 1A:
V_R = 1A × 1Ω = 1V
P_R,total = 1^2 × 1Ω = 1W
Each 2Ω resistor carries 0.5A:
P_each = 0.5^2 × 2Ω = 0.5W
With R_S = 1Ω, IC dissipation at the worst-case fast-charge entry point:
P_IC = (5.0V - 2.03V - 1.0V) × 1A
P_IC ≈ 1.97W
At 50°C/W:
ΔT ≈ 1.97W × 50°C/W = 98.5°C
T_J ≈ 25°C + 98.5°C = 123.5°C
This leaves roughly 11.5°C margin below a 135°C thermal regulation point under the stated assumptions.
Resistor Power Rating Discussion
For two 2Ω resistors in parallel at 1A:
- Each resistor dissipates approximately 0.5W.
- 2W parts are thermally comfortable but may be much more expensive.
- 1W parts are a likely cost/performance compromise: each runs at about 50% of rating.
- 0.5W parts are not preferred because they operate at their full rating during the hottest operating point.
PCB Thermal Assumptions
Planned PCB:
- 45mm × 80mm board
- 2 layers
- Copper pours on both sides
- Vias connecting pours between sides
Board area:
45mm × 80mm = 3600mm² = 36cm²
Both sides together:
≈72cm² copper/board surface area, if effectively used
Rough natural-convection estimate using h ≈ 10W/m²K:
A = 0.0072m²
G = h × A = 10 × 0.0072 = 0.072W/K
θ_BA ≈ 1 / 0.072 ≈ 13.9°C/W
This is optimistic because not all board area is equally hot and heat must spread from the IC into copper and through vias.
Practical working estimate for the charger IC with good exposed-pad soldering, large copper pours, and via stitching:
θ_JA,effective ≈ 35–60°C/W
A conservative planning value used:
If thermal pad connection is poor or copper/via stitching is sparse, the effective thermal resistance could be much worse, possibly 70–100°C/W.
Layout Guidance for Thermal Success
No layout changes are authorized by this document, but future layout should consider:
- Solder CN3085 exposed pad well to GND copper.
- Use large continuous copper pours connected to the exposed pad.
- Stitch top and bottom copper with many vias near the IC thermal pad.
- Keep heat-sharing resistors physically away from the CN3085 so resistor heat does not directly heat-soak the IC.
- Give the series resistors their own copper area for heat spreading.
- Avoid thermally isolating the IC pad with narrow spokes if maximum heat spreading is needed.
- Use test pads or accessible measurement points to validate charge current and thermal behavior during bring-up.
Overall Design Assessment
The CN3085 + external heat-sharing resistor approach is a defensible low-cost compromise for a standalone 2×AA NiMH USB charger.
Advantages:
- Low component/labor cost.
- No firmware or MCU programming.
- Uses the common 2×AA battery format.
- Large PCB area can help with heat spreading.
- External resistors can move a meaningful portion of heat out of the charger IC.
Risks / limitations:
- Still thermally marginal at 1A unless PCB copper and via stitching are good.
- Linear topology is inefficient when charging low-voltage 2-cell packs from 5V USB.
- Practical maximum current may be determined by thermal regulation rather than RISET alone.
- High-capacity modern AA NiMH cells make low-current charging slow.
Most realistic current range:
- Around 350–700mA without aggressive heat spreading, depending on actual thermal resistance.
- 1A may be plausible with ~1Ω series heat-sharing resistance, good PCB copper, and careful thermal layout, but should be validated empirically.
Open Validation Items
Before finalizing production values:
- Confirm selected R5/C1 values against the exact final charge-current setting.
- Confirm C1 effective capacitance at RC pin bias using the capacitor datasheet.
- Confirm RISET value for the desired charge current.
- Thermal-test the assembled PCB at worst-case battery voltage near the precharge-to-fast-charge transition.
- Measure whether the CN3085 enters thermal regulation at the chosen charge current.
- Decide final series resistor value and power rating based on measured temperature rise.