DEV Community: Mia

Network Transformer Incoming Inspection: LCR Measurement, Hipot, and Functional Test Procedure

Mia — Thu, 25 Jun 2026 08:58:10 +0000

Receiving a new batch of network transformer samples? Here's the complete inspection procedure — instruments, settings, pass/fail criteria, and automation ideas.
Equipment Checklist
Required:
[x] LCR meter — must support 100kHz test frequency
Budget option: Tonghui TH2816B (~$300)
Mid-range: Hioki IM3536 (~$800)
High-end: Keysight E4980A (~$3500)

[x] 4-wire milliohm meter (or LCR meter with DCR mode)

Recommended:
[~] Hipot tester — for isolation voltage verification
[~] VNA — for insertion loss / return loss (NanoVNA sufficient for 100BASE-TX)

Optional:
[~] Calipers — for dimensional / footprint verification
[~] Test PCB — transformer footprint + 100Ω termination + SMA ports
OCL Measurement (Most Critical Parameter)
python# Measurement parameters — DO NOT deviate from these values
OCL_TEST_CONFIG = {
"frequency_Hz": 100_000, # 100 kHz
"signal_V_rms": 0.1, # 0.1 V RMS — matches actual signal level
"mode": "L", # Inductance
"secondary": "open", # Secondary winding pins: floating / not connected
}

Pass / fail criteria (design to Min, not Typ)

OCL_PASS_CRITERIA = {
"10/100BASE-TX": 350e-6, # 350 µH (IEEE 802.3 Clause 40)
"1000BASE-T": 1000e-6, # 1000 µH (IEEE 802.3 Clause 25/28)
}

Common supplier trap:

Some datasheets specify OCL at 1kHz or 1V RMS → inflated reading

Always confirm: "What is your OCL test frequency and voltage?"

DCR Measurement
Winding Pins Typical Range Flag If
Primary TX+ → TX− 1–5 Ω > datasheet max
Secondary RX+ → RX− 1–5 Ω > datasheet max
CT split TX+ → CT_TX ~0.5–2.5 Ω > 60% of full DCR

Use 4-wire (Kelvin) measurement to eliminate probe contact resistance.
High DCR + high OCL = extra turns → inflated inductance, worse HF performance.
Isolation Voltage (Hipot) Test
Test voltage: 1500V AC (standard IEEE 802.3)
2500V AC (industrial / IEC 61000-4 environments)
Duration: 60 seconds
Pass criterion: No breakdown; leakage < 1 mA
Connections:
HV terminal → all primary pins shorted together
GND terminal → all secondary pins shorted together

⚠️ Safety: proper hipot enclosure + PPE + trained operator required
⚠️ Never probe hipot with standard DMM leads
Visual Inspection Checklist
bash# Per-unit checks (100% for samples ≤20 pcs; AQL II sampling for production)
grep -E "PASS|FAIL" <<'CHECKLIST'
[ ] Part number on component matches purchase order
[ ] Pin count matches datasheet
[ ] Pin pitch within ±0.05mm of datasheet (verify with calipers)
[ ] No cracked ferrite, no damaged leads/pads
[ ] Pin 1 marker clearly visible and consistent across lot
[ ] Date code / lot code recorded for traceability
CHECKLIST
Functional Test
Test PCB setup:
PHY chip (e.g., LAN8720A, DP83848) → Transformer → RJ45
Bob Smith: 75Ω + 1000pF to chassis GND on each center tap

Pass criteria:
✓ Link establishes at 100Mbps (or negotiated speed)
✓ Zero packet loss on 5-minute ping test

✓ Link stable on 50m cable (stress test for return loss)
✓ Link stable at +70°C ambient (soak test for OCL at temperature)

If functional test fails but OCL/DCR pass:
→ Check Bob Smith termination ground (chassis GND, not signal GND)
→ Check center tap pin connections in schematic
→ Verify footprint pin-out against this supplier's datasheet
Incoming Inspection Decision Matrix
Application OCL DCR Hipot Functional
────────────────────────────────────────────────────────
Eng. sample ALL ALL — 1 unit
Consumer (AQL) AQL AQL — Sampling
Industrial AQL AQL ALL ALL
PoE design AQL ALL AQL ALL
Supplier Data to Request
Before placing an order, ask:

Lot-specific OCL test data (what frequency, what voltage?)
Isolation voltage production test report (what duration?)
Datasheet with Min / Typ / Max columns clearly labeled

Voohu Technology (www.voohuele.com) provides lot test data with OCL at 100kHz/0.1V RMS. MOQ 50pcs, DHL 3–5 days.

4-Step Guide to Reading a Network Transformer Datasheet

Mia — Wed, 24 Jun 2026 08:40:33 +0000

Reading a network transformer datasheet has a learning curve. Here's a structured approach to extracting what your design actually needs.
Step 1 — Identify the Application Standard
Compliance statement to look for in the datasheet:
"Meets IEEE 802.3 Clause 40" → 10/100BASE-TX
"Meets IEEE 802.3 Clause 40/25/28" → 1000BASE-T (Gigabit)
"IEEE 802.3at PoE rated" → Power over Ethernet
"IEC 60601-1 5000V isolation" → Medical grade

If no IEEE 802.3 clause is cited → ask the supplier or find a different part
Step 2 — Decode the Electrical Specifications Table
ParameterSymbol10/100BASE-TX1000BASE-TTest ConditionOpen Circuit InductanceOCL≥ 350µH (min)≥ 1000µH (min)100kHz, 0.1V RMSInsertion LossIL< 1.0 dB (max)< 1.0 dB (max)1–100MHz, 100Ω diffReturn LossRL> 16 dB (min)> 16 dB (min)1–100MHz, 100Ω diffIsolation VoltageV_ISO≥ 1500V AC≥ 1500V AC1 min, 50/60HzTurns Ratio—1CT:1CT1CT:1CT—DC ResistanceDCR< 5Ω typ< 3Ω typPer winding
Step 3 — Read Test Conditions, Not Just Values
python# OCL test condition matters

Different conditions = different (incomparable) values

def compare_ocl_valid(spec_a, spec_b):
"""
Returns True only if both specs use the same test conditions
"""
standard_cond = {"freq_kHz": 100, "voltage_rms": 0.1}

a_valid = (spec_a["test_freq_kHz"] == standard_cond["freq_kHz"] and

           spec_a["test_voltage_rms"] == standard_cond["voltage_rms"])

b_valid = (spec_b["test_freq_kHz"] == standard_cond["freq_kHz"] and

           spec_b["test_voltage_rms"] == standard_cond["voltage_rms"])

if not (a_valid and b_valid):

    print("⚠️ Test conditions differ — values not directly comparable")

    return False

return True

Real world trap:

spec_non_standard = {"ocl_uH": 600, "test_freq_kHz": 10, "test_voltage_rms": 1.0}
spec_standard = {"ocl_uH": 380, "test_freq_kHz": 100, "test_voltage_rms": 0.1}

Same physical part — different measurement condition

600µH at 10kHz/1V ≠ 600µH at 100kHz/0.1V

Step 4 — Design to Min/Max, Not Typical
WRONG approach:
Datasheet shows OCL typ = 480µH
Design assumes 480µH is the production value ✗

CORRECT approach:
Datasheet shows OCL min = 350µH, typ = 480µH
Design must work at 350µH ✓
Simulate and verify at the min column value

Same rule applies to insertion loss:
Datasheet shows IL typ = 0.4 dB, max = 1.0 dB
Design for 1.0 dB worst case, not 0.4 dB typical
Pin-Out: The Most Common Error Source
Pin compatibility ≠ package compatibility

Two 8-pin SMD network transformers from different vendors:
Vendor A: Pin 1=TX+, Pin 2=TX−, Pin 3=CTx, Pin 4=RX+...
Vendor B: Pin 1=RX+, Pin 2=RX−, Pin 3=CTx, Pin 4=TX+...

Building footprint from Vendor A's datasheet for Vendor B's part = crossed TX/RX
Result: link never establishes, no obvious error in schematic

Always build footprint from the specific part's datasheet pin-out.
Never reuse footprints across different vendor parts.
Isolation Voltage: Check the Test Duration
Standard production test: 1500V AC for 60 seconds
Acceptable shorthand: "hipot tested at 1500V AC"

Red flag: "1500V AC momentary" or no duration stated
→ Much lower accumulated stress on insulation
→ May pass initial test but fail in service under sustained stress

For industrial (factory floor, outdoor) applications:
→ Specify 2500V AC, 60 seconds minimum
→ Ask supplier for production test report
Datasheet Checklist
[ ] IEEE 802.3 clause stated (Clause 40 / 25 / 28)
[ ] OCL tested at 100kHz, 0.1V RMS
[ ] OCL value is Min column, not Typ
[ ] IL and RL specified over 1–100MHz with 100Ω differential
[ ] Isolation voltage test duration documented
[ ] Turns ratio is 1CT:1CT (center taps confirmed)
[ ] Pin-out diagram matches your PCB footprint (verify every pin)
[ ] Operating temperature range covers your application
[ ] PoE rated if DC bias is present (PSE or PD designs)
Source
Voohu Technology (www.voohuele.com) — network transformers with complete datasheets: OCL at 100kHz/0.1V RMS, temperature-characterized specs, application circuits included. MOQ 50pcs, DHL 3–5 days.

Industrial Ethernet Transformer Requirements: 2500V Isolation, Temperature, and IEC 61000 Compliance

Mia — Tue, 23 Jun 2026 03:54:28 +0000

Industrial Ethernet deployments require transformers that exceed IEEE 802.3 minimums. Here's the complete specification delta between standard and industrial-grade requirements.
Standard vs Industrial Transformer Requirements
Parameter IEEE 802.3 Minimum Industrial Requirement
──────────────────────────────────────────────────────────────────────
Isolation voltage 1500V AC 2500V AC
Operating temperature 0°C to +70°C −40°C to +85°C
OCL @ temp extremes Not specified ≥ 350µH across full range
Vibration resistance Not specified IEC 60068-2-6 compliance
Shock resistance Not specified IEC 60068-2-27 compliance
EMC immunity CISPR 22 (emissions) IEC 61000-4 series (immunity)
Product lifecycle Consumer (3–5 yr) Industrial (10–20+ yr)
Industrial Ethernet Protocol Overview
pythonindustrial_protocols = {
"PROFINET": {
"developer": "Siemens / PI",
"physical": "100BASE-TX (RT) / GbE (IRT backbone)",
"timing": "Hard real-time — IRT cycle 250µs, jitter <1µs",
"market": "Europe, Japan, Korea (dominant in Asia)",
"cert_body": "PI (PROFIBUS International)",
"key_req": "IEC 61000-4 immunity mandatory for PI cert"
},
"EtherNet/IP": {
"developer": "Rockwell / ODVA",
"physical": "100BASE-TX / 1000BASE-T",
"timing": "Soft to hard real-time",
"market": "Americas, global",
"cert_body": "ODVA conformance test",
"key_req": "Standard 802.3 + IEC 61000-4 immunity"
},
"EtherCAT": {
"developer": "Beckhoff",
"physical": "100BASE-TX",
"timing": "Ultra-hard real-time, sub-µs sync",
"market": "Global, motion control dominant",
"cert_body": "ETG (EtherCAT Technology Group)",
"key_req": "Tight insertion loss uniformity across 1-100MHz"
},
"CC-Link IE": {
"developer": "Mitsubishi / CC-Link Partner Assoc.",
"physical": "1000BASE-T",
"timing": "Hard real-time",
"market": "Japan, Asia",
"cert_body": "CLPA",
"key_req": "Gigabit magnetics, extended temp"
}
}
IEC 61000-4 Immunity: What Each Test Means for Your Transformer
Test Standard Level Ethernet Port Impact Transformer Role
───────────────────────────────────────────────────────────────────────────
IEC 61000-4-2 8kV contact ESD at RJ45 Isolation barrier, Bob Smith
IEC 61000-4-4 1kV burst EFT on signal lines Isolation + Bob Smith CMR
IEC 61000-4-5 1kV surge Line-to-GND at port Isolation voltage critical
IEC 61000-4-6 10V RMS Conducted RF immunity Common-mode rejection (OCL)

Isolation voltage vs surge test:
1500V isolation + 1kV surge: marginal (1kV peak + system voltage)
2500V isolation + 1kV surge: comfortable margin

→ 2500V isolation recommended for all industrial Ethernet designs
Temperature Effect on OCL
pythonimport math

Typical NiZn ferrite permeability temperature coefficient

µr drops at temperature extremes

def ocl_at_temp(ocl_nominal, temp_C):
"""
Simplified OCL vs temperature model
Real behavior depends on core material grade
"""
# Approximate: ferrite µr reduces ~0.5–2% per degree from 25°C
delta_T = abs(temp_C - 25)
reduction_factor = 1 - (0.008 * (delta_T / 10)) # ~0.8% per 10°C
return ocl_nominal * max(reduction_factor, 0.75) # floor at 75% of nominal

ocl_nominal = 400e-6 # 400µH at 25°C (datasheet value)
min_required = 350e-6 # 350µH minimum for 100BASE-TX

for temp in [-40, -20, 0, 25, 70, 85]:
ocl = ocl_at_temp(ocl_nominal, temp)
status = "✅" if ocl >= min_required else "❌ BELOW SPEC"
print(f" {temp:+4d}°C: {ocl*1e6:.0f}µH {status}")
Vibration Profile Comparison
Environment Vibration Level Recommended Package
──────────────────────────────────────────────────────────────
Office / lab Negligible SMD
Light industrial IEC 68-2-6 Fc SMD + conformal coat
Heavy machinery IEC 68-2-6 Fd THT preferred
Transportation IEC 68-2-6 Fh THT + conformal coat
Robotic arm Custom profile THT mandatory
Industrial Ethernet Design Checklist
[ ] Isolation voltage: 2500V AC (not just 1500V minimum)
[ ] Operating temp: −40°C to +85°C with OCL guaranteed across range
[ ] Package: THT for vibration environments; SMD + conformal coat for moderate
[ ] Bob Smith: chassis GND connection — critical for IEC 61000-4-4/4-5
[ ] PoE: PoE-rated magnetics if industrial PoE powered (802.3af/at/bt)
[ ] Product lifecycle: confirm long-term supply availability
[ ] PROFINET: verify IEC 61000-4 immunity test compliance
[ ] EtherCAT: verify flat insertion loss uniformity 1–100MHz
[ ] CC-Link IE: Gigabit (1000BASE-T) magnetics required
Source
Voohu Technology (www.voohuele.com) — industrial Ethernet transformers: 2500V isolation, −40°C to +85°C, SMD and THT, PoE-rated. MOQ 50pcs, DHL 3–5 days.

Network Transformer Selection for IoT Ethernet: PHY Compatibility, EEE, and Low-Power Design

Mia — Mon, 22 Jun 2026 07:14:57 +0000

IoT Ethernet designs have different transformer requirements than standard networking equipment. Here's the complete technical reference for IoT-specific selection.
IoT vs Standard Ethernet: Key Differences
Parameter Standard Ethernet Equipment IoT Device
──────────────────────────────────────────────────────────────────────
Power budget Not constrained Milliwatt-level concern
Operating temp 0°C to +70°C (commercial) -40°C to +85°C (industrial)
PCB footprint Flexible Compact — often <50×50mm
Package preference Either SMD or THT SMD, smallest available
PoE usage Optional Common (cable powers sensor)
PHY type Full-featured Low-power (LAN8720A, W5500)
EEE support Common Mandatory for battery designs
BOM cost Secondary Critical (high volume)
MOQ requirement High volume Small batch (50–500 pcs)
Common IoT PHY Chips and Transformer Requirements
pythoniot_phy_transformer_requirements = {
"Microchip LAN8720A": {
"interface": "RMII",
"supply": "3.3V",
"speed": "10/100BASE-TX",
"OCL_min": "350µH @ 100kHz",
"turns_ratio": "1CT:1CT",
"EEE": True,
"notes": "Most popular IoT PHY; STM32 + LAN8720A is dominant combo"
},
"WIZnet W5500": {
"interface": "SPI (TCP/IP offload)",
"supply": "3.3V",
"speed": "10/100BASE-TX",
"OCL_min": "350µH @ 100kHz",
"turns_ratio": "1CT:1CT",
"EEE": False,
"notes": "SPI Ethernet — still needs external magnetics"
},
"TI DP83822": {
"interface": "MII/RMII",
"supply": "3.3V",
"speed": "10/100BASE-TX",
"OCL_min": "350µH @ 100kHz",
"turns_ratio": "1CT:1CT",
"EEE": True,
"notes": "Industrial grade, -40°C support"
},
"Realtek RTL8201": {
"interface": "MII/RMII",
"supply": "3.3V",
"speed": "10/100BASE-TX",
"OCL_min": "350µH @ 100kHz",
"turns_ratio": "1CT:1CT",
"EEE": True,
"notes": "Cost-optimized; high-volume IoT gateways"
}
}
Energy Efficient Ethernet (EEE / 802.3az) Impact
Normal operation: Continuous signal → transformer operates at nominal OCL
LPI (Low Power Idle): Near-zero signal activity → transformer core idles

Potential issue:
Some transformers show insertion loss spike after LPI wake-up
Core operating point must re-establish after idle period
First 1–5ms after wake: elevated insertion loss → packet errors

Verification:
Test with PHY in EEE mode at maximum LPI duration
Monitor first-packet error rate after wake-up
Compare transformers if issue found — core material matters
Temperature vs OCL: What to Verify
OCL specification must be guaranteed at operating temperature extremes:

Temperature Ferrite µr OCL impact
──────────────────────────────────────────────
-40°C Reduced OCL may drop 10–30%
+25°C Nominal Datasheet reference point
+85°C Reduced OCL may drop 5–15% (varies by grade)

⚠️ "350µH at 25°C" is NOT the same as "350µH across -40°C to +85°C"
Request temperature-characterized OCL curves for industrial IoT
IoT PoE Power Dissipation
python# PoE 802.3af power budget in IoT transformer
poe_current = 0.350 # A (802.3af max)
dcr_per_winding = 0.5 # Ω typical

10/100BASE-TX: 2 winding pairs (4 windings total)

total_dissipation = (poe_current**2) * dcr_per_winding * 4
print(f"PoE 802.3af transformer dissipation: {total_dissipation*1000:.1f} mW")

Output: 245.0 mW ← significant in compact IoT enclosure

For 0.3Ω DCR (low-power variant):

dcr_low = 0.3
total_low = (poe_current**2) * dcr_low * 4
print(f"Low-DCR variant dissipation: {total_low*1000:.1f} mW")

Output: 147.0 mW ← 40% reduction

IoT Transformer Selection Checklist
[ ] PHY reference schematic reviewed — transformer spec confirmed
[ ] OCL ≥ 350µH guaranteed across -40°C to +85°C (industrial)
[ ] SMD package ≤ 8×8mm footprint (compact IoT board)
[ ] DCR minimized (low-power variant if available)
[ ] EEE/LPI compatibility verified with PHY
[ ] PoE-rated if design receives or sources PoE power
[ ] 1500V AC isolation (IEEE 802.3 minimum, check for industrial uplift)
[ ] Supplier supports small MOQ (50–500 pcs for prototype/NPI)
Source
Voohu Technology (www.voohuele.com) — IoT-grade network transformers: compact SMD, -40°C to +85°C, PoE-rated options. MOQ 50pcs, DHL 3–5 days to Japan/Korea/Southeast Asia.

Return Loss in Ethernet Magnetics: S11 Measurement, Impedance Analysis, and Layout Impact

Mia — Wed, 17 Jun 2026 09:56:49 +0000

Return loss measures impedance match quality at the transformer port. Here's the complete technical reference — formula, VNA method, physical causes, and PCB layout contribution.
Definition and Formula
pythonimport math

def reflection_coefficient(Z_load, Z_source=100):
"""
Z_load: impedance presented by transformer at measurement port (Ω)
Z_source: system characteristic impedance (100Ω for Ethernet)
Returns: reflection coefficient magnitude |Γ|
"""
gamma = abs((Z_load - Z_source) / (Z_load + Z_source))
return gamma

def return_loss_dB(gamma):
"""Convert reflection coefficient to return loss in dB"""
if gamma == 0:
return float('inf')
return -20 * math.log10(gamma)

def vswr(gamma):
"""Convert reflection coefficient to VSWR"""
return (1 + gamma) / (1 - gamma)

Reference table

print(f"{'RL (dB)':>10} | {'|Γ|':>8} | {'VSWR':>8} | {'Refl. Power':>12}")
print("-" * 50)
for rl in [10, 16, 20, 26, 30]:
g = 10 ** (-rl / 20)
print(f"{rl:>10} | {g:>8.4f} | {vswr(g):>8.3f} | {g**2*100:>10.2f}%")

Output:

RL (dB) | |Γ| | VSWR | Refl. Power

────────────────────────────────────────────

10 | 0.3162 | 1.925 | 10.00%

16 | 0.1585 | 1.377 | 2.51% ← IEEE 802.3 minimum

20 | 0.1000 | 1.222 | 1.00%

26 | 0.0501 | 1.106 | 0.25%

30 | 0.0316 | 1.065 | 0.10%

IEEE 802.3 Return Loss Requirements
StandardMin. Return LossFrequency Range100BASE-TX> 16 dB1MHz – 100MHz1000BASE-T> 16 dB (per pair)1MHz – 100MHz
Typical quality parts: 18–25 dB across passband.
VNA S11 Measurement Setup
VNA
Port 1
│
┌─────▼──────┐
│ Transformer │
│ (Primary) │
│ │
│ (Secondary)├──[100Ω]── GND
└─────────────┘

Step 1: Calibrate VNA at Port 1 reference plane (SOLT or TRL cal)
Step 2: Connect Port 1 to transformer primary with matched fixture
Step 3: Terminate secondary with 100Ω ±1% to chassis GND
Step 4: Sweep 100kHz to 200MHz, read S11 magnitude in dB
Step 5: Repeat with ports swapped (Port 1 to secondary, 100Ω on primary)

S11 reading = Return Loss at each frequency

⚠️ VNA calibration quality directly determines measurement accuracy
Uncalibrated VNA: results are meaningless for return loss
Poor fixture: adds 3–5 dB of apparent return loss degradation
Physical Causes of Degraded Return Loss
Cause Frequency Range Impedance Effect
──────────────────────────────────────────────────────────────────
Magnetizing inductance < 1 MHz Low ZL at low freq
Turns ratio deviation All frequencies Scales impedance by N²
Leakage inductance > 30 MHz Adds series Z
Inter-winding cap > 100 MHz Reduces parallel Z
PCB trace impedance All frequencies Adds in series with transformer

Turns ratio deviation example:
Ideal: N = 1.000 → Z_sec = 100Ω → RL = ∞
Error: N = 1.010 → Z_sec = 1.01² × 100 = 102Ω
Γ = (102-100)/(102+100) = 0.0099
RL = -20 × log10(0.0099) = 40.1 dB ← 1% turns error → 40dB RL
(Turns ratio deviation is rarely the dominant cause at high frequencies)
PCB Layout Contribution to System Return Loss
System RL = f(transformer RL, trace impedance, via parasitics, pad geometry)

Typical PCB contributions (100MHz):
50mm of 85Ω trace (vs 100Ω target): adds ~3–5 dB degradation
2 vias in signal path (1pF each): adds ~1–2 dB at 100MHz
Mismatched pad geometry: adds ~0.5–2 dB

Example:
Transformer intrinsic RL: 22 dB
PCB layout contribution: -5 dB
Measured system RL: 17 dB (still above 16 dB, but marginal)

→ Root cause of many "the part seems bad" return loss failures is PCB layout
Return Loss vs Insertion Loss Comparison
Parameter S-param Measures Limit (802.3)
──────────────────────────────────────────────────────────────
Return loss S11 Reflected at input > 16 dB
Insertion loss S21 Transmitted to output < 1.0 dB

Independence: IL and RL are NOT complementary
Can have: good S21 (low IL) + poor S11 (low RL)
Both must independently meet specification
Source
Voohu Technology (www.voohuele.com) — network transformers with full S11/S21 characterization. MOQ 50pcs, DHL 3–5 days.

Insertion Loss in LAN Magnetics: Measurement Method, Frequency Curves, and Datasheet Evaluation

Mia — Tue, 16 Jun 2026 08:18:01 +0000

Insertion loss is the primary signal-quality specification for LAN magnetics. Here's the complete technical reference for measurement, interpretation, and datasheet evaluation.
Definition and Formula
Insertion Loss (IL):
IL = 20 × log10(V_out / V_in) [dB]

Where:
V_in = signal voltage at transformer primary (source side)
V_out = signal voltage at transformer secondary (load side)

Example:
V_in = 100 mV
V_out = 89.1 mV → IL = 20 × log10(0.891) = -1.00 dB

Convention: expressed as magnitude (positive dB, "less than 1.0 dB")
IEEE 802.3 specifies: IL < 1.0 dB for 100BASE-TX
IEEE 802.3 Limits
StandardMax ILFrequency RangeNote10BASE-T1.0 dB100kHz–10MHz100BASE-TX1.0 dB1MHz–100MHz1000BASE-T1.1 dB1MHz–100MHzPer pair; pair balance < 0.5 dB
Frequency Response Shape
python# Conceptual insertion loss vs. frequency (3 regions)

regions = {
"Low-freq rollup": {
"range": "< 1 MHz",
"cause": "OCL too low → magnetizing branch shunts signal",
"fix": "Higher OCL spec (1000µH for 1000BASE-T)"
},
"Passband (flat)": {
"range": "1–100 MHz",
"cause": "N/A — optimal operating region",
"typical_IL": "0.3–0.6 dB for quality parts"
},
"High-freq rollup": {
"range": "> 100 MHz",
"cause": "Leakage inductance + inter-winding capacitance",
"note": "Outside 802.3 spec range but visible on VNA curve"
}
}

Frequency where 3dB rolloff begins:

Low end: f_3dB_low = R_load / (2π × OCL)

At 100Ω load, 1000µH OCL: f_3dB_low = 100 / (2π × 1000e-6) ≈ 15.9 kHz

import math
R = 100 # Ω
OCL_100M = 350e-6 # 350µH for 100BASE-TX
OCL_1G = 1000e-6 # 1000µH for 1000BASE-T

f_low_100M = R / (2 * math.pi * OCL_100M)
f_low_1G = R / (2 * math.pi * OCL_1G)

print(f"100BASE-TX low-end -3dB: {f_low_100M/1e3:.1f} kHz")
print(f"1000BASE-T low-end -3dB: {f_low_1G/1e3:.1f} kHz")

100BASE-TX: ~45.5 kHz

1000BASE-T: ~15.9 kHz (lower rolloff → supports lower signal frequencies)

Physical Loss Mechanisms
Source Frequency Dependence Effect
──────────────────────────────────────────────────────────────────
Winding DCR Flat (rises with temp) IL floor at all freqs
Core hysteresis Increases with freq Mid/HF loss
Eddy currents f² dependence HF loss
Leakage inductance Increases with freq HF rolloff
Inter-winding cap. Increases with freq HF bypass / resonance

Temperature effect on DCR:
ΔR/ΔT = +0.39%/°C (copper)
At +85°C vs +25°C: DCR increases ~23.4%
→ IL increases at elevated temperature
→ Critical for PoE designs with self-heating
VNA Measurement Setup
Vector Network Analyzer (VNA) S21 measurement:

VNA Port 1 ──[100Ω]── Transformer Primary ──┐
│ (magnetic coupling)
VNA Port 2 ──[100Ω]── Transformer Secondary ─┘

S21 (dB) = forward voltage transmission = Insertion Loss

Sweep range: 100 kHz – 200 MHz
Termination: 100Ω differential at both ports (simulates cable/PHY impedance)
Level: 0.1V RMS or per manufacturer spec (keep below saturation)

Without 100Ω termination: results are meaningless
→ Floating secondary = very low insertion loss (no load, no loss path)
→ Mismatch at primary = resonances not visible in real circuit
1000BASE-T: Pair Balance Requirement
All 4 pairs must be measured individually:

Pair 1: IL = 0.45 dB
Pair 2: IL = 0.48 dB
Pair 3: IL = 0.51 dB
Pair 4: IL = 0.44 dB

Max variation: 0.51 - 0.44 = 0.07 dB ✅ (well within 0.5 dB limit)

Why it matters:
PAM-5 receiver sees different signal amplitude on each pair
→ PHY DSP compensates, consuming noise margin
→ Large imbalance → reduced link margin → packet errors at max cable length
Datasheet Evaluation Checklist
✅ Insertion loss vs. frequency curve (not just single-frequency value)
✅ Maximum guaranteed value (not just "typical")
✅ Test conditions stated: frequency range, termination impedance
✅ Temperature range (25°C? full operating range?)
✅ Pair-to-pair balance data (1000BASE-T)
❌ Red flag: "IL < 1.0 dB typical" with no curve, no conditions
❌ Red flag: Specified at one frequency only (e.g., "< 1.0 dB @ 10MHz")
❌ Red flag: No maximum value — only typical
Source
Voohu Technology (www.voohuele.com) — LAN magnetics and network transformers with full insertion loss characterization. MOQ 50pcs, DHL 3–5 days.

Network Transformer PCB Placement: Rules, Measurements, and Layout Patterns

Mia — Mon, 15 Jun 2026 09:22:52 +0000

Transformer placement and routing determine EMC test outcomes more than any other single PCB decision in an Ethernet design. Here are the rules with specific measurements.
Optimal Layout Pattern
Board edge (left) ─────────────────────────────── Board interior (right)

┌──────┐ ┌─────┐ ┌─────────────────┐ ┌─────────┐
│ RJ45 │───│ CMC │───│ Transformer │───│ PHY │
│ │ │(opt)│ │ (secondary) │ │ chip │
└──────┘ └─────┘ │ (primary) → │ └─────────┘
└─────────────────┘
│
┌───────────────┐
│ Bob Smith │
│ 75Ω + 1000pF │
└───────┬───────┘
│
CHASSIS GND pour

Signal flow: PHY → Transformer → CMC → RJ45 → Cable
Layout: interior ──────────────────────→ board edge
Placement Distance Rules
Connection Target Maximum Impact if exceeded
──────────────────────────────────────────────────────────────────────
Transformer secondary to RJ45 < 10mm 20mm EMC cable coupling
Bob Smith to center tap pins < 5mm 10mm HF bypass degraded
CMC to RJ45 < 15mm 25mm Re-couples CM after CMC
PHY to transformer primary < 30mm 50mm SI trace length increases
Transformer to switching node > 10mm 5mm min Switching noise coupling
Transformer to crystal/clock > 10mm 5mm min Clock coupling into pairs
Isolation Zone (Ground Plane Moat)
Per IEC 62368-1 @ 1500V, Pollution Degree 2:

Creepage (along PCB surface): ≥ 6.4mm
Clearance (through air): ≥ 4.0mm

Moat implementation:
┌──────────────────┬──────────────────────┐
│ PRIMARY side │ SECONDARY side │
│ (Signal GND) │ (Chassis GND) │
│ PHY, decoupling │ Bob Smith, RJ45 │
└──────────────────┤ shield, mounting │
← 4mm+ → │ │
└──────────────────────┘

Rules:
✅ No copper crosses moat on any layer
✅ No via passes through moat
✅ Only connection: Bob Smith 1000pF → chassis GND (star point)
❌ Do NOT bridge moat with ground pour
❌ Do NOT run differential traces across moat boundary
Differential Pair Routing Checklist
differential_pair_rules = {
# Electrical
"impedance": "100Ω differential (±10%)",
"length_match_100M": "±2mm between pairs",
"length_match_1000M": "±1mm between pairs",
"via_count": "Minimize — zero preferred on secondary side",

# Routing
"pair_spacing":         "Consistent throughout run (2× trace width typical)",
"layer_transitions":    "Use differential via pairs — never single-ended via",
"parallel_run_limit":   "< 5mm alongside switching traces or clocks",

# Layer strategy
"preferred_layer":      "Top layer (short runs), avoid buried via transitions",
"reference_plane":      "Continuous ground reference beneath entire run",

}

routing_mistakes = [
"Single via on one trace of differential pair (impedance imbalance)",
"Routing pair across isolation moat (violates galvanic isolation)",
"90° corners on differential pairs (use 45° or curved bends)",
"Length mismatch > 2mm (skew degrades common-mode rejection)",
"Running pairs adjacent to switching regulator traces > 5mm"
]
Multi-Port Layout
For boards with N Ethernet ports:

Port 1: [RJ45] ─ [CMC] ─ [XFMR1] ─ [PHY1] ← isolated moat 1
Port 2: [RJ45] ─ [CMC] ─ [XFMR2] ─ [PHY2] ← isolated moat 2
Port N: [RJ45] ─ [CMC] ─ [XFMRN] ─ [PHYN] ← isolated moat N

Each port has its own:

Isolation moat
Bob Smith termination
Chassis GND connection to chassis pour

Do NOT share moat between ports.
Each transformer's secondary chassis GND connects independently to chassis copper.
Layer Stack Recommendation
Layer 1 (Top): Components + signal routing
Layer 2: Ground plane (solid — no splits except moat)
Layer 3: Power planes
Layer 4 (Bot): Signal routing + chassis GND pour (secondary side)

Moat in ground plane (Layer 2):
≥ 4mm gap under transformer body between primary and secondary copper
Source
Voohu Technology (www.voohuele.com) — network transformers with reference footprints and layout guidelines. MOQ 50pcs, DHL 3–5 days.

Bob Smith Termination: Circuit Analysis, Layout Rules, and Common Mistakes

Mia — Fri, 12 Jun 2026 06:06:50 +0000

Bob Smith termination is in every Ethernet schematic. Here's the complete technical breakdown — component values, impedance analysis, chassis ground vs signal ground, and layout rules.
Circuit Topology
Transformer center taps:

CT_PRI ──── 75Ω ──┐
CT_SEC ──── 75Ω ──┴── CTAP_NODE ──── 1000pF ──── CHASSIS_GND

For 1000BASE-T (4 pairs):

CT_PAIR1 ──── 75Ω ──┐
CT_PAIR2 ──── 75Ω ──┤
CT_PAIR3 ──── 75Ω ──┼── CTAP_NODE ──── 1000pF ──── CHASSIS_GND
CT_PAIR4 ──── 75Ω ──┘
Impedance Analysis: Why 1000pF
import math

C = 1000e-12 # 1000 pF

frequencies = [1e6, 10e6, 30e6, 100e6, 300e6]

print(f"{'Frequency':>12} | {'Zc (Ω)':>10} | {'Note'}")
print("-" * 50)
for f in frequencies:
Zc = 1 / (2 * math.pi * f * C)
note = ""
if f <= 1e6: note = "High Z — DC/LF isolation maintained"
if f >= 30e6: note = "Low Z — effective common-mode bypass"
if f >= 100e6: note = "Near-short — excellent at cable radiation freq"
print(f"{f/1e6:>10.0f}MHz | {Zc:>10.2f} | {note}")

Output:

1MHz | 159.15 | High Z — DC/LF isolation maintained

10MHz | 15.92 |

30MHz | 5.31 | Low Z — effective common-mode bypass

100MHz | 1.59 | Near-short — excellent at cable radiation freq

300MHz | 0.53 | Near-short — excellent at cable radiation freq

Chassis GND vs Signal GND: Critical Difference
CORRECT: CT → 75Ω → CTAP → 1000pF → CHASSIS_GND
✅ Common-mode discharges to chassis (large, clean plane)
✅ No coupling to PHY reference ground
✅ ESD energy absorbed by chassis
✅ Radiated emissions reduced

WRONG: CT → 75Ω → CTAP → 1000pF → SIGNAL_GND (DGND)
❌ Common-mode injects noise into PHY reference ground
❌ Common-mode → differential-mode conversion at PHY input
❌ Radiated emissions may INCREASE
❌ ESD stresses PHY ground reference
ESD Discharge Path Analysis
ESD pulse characteristics:
Rise time: ~1 ns (IEC 61000-4-2 Contact discharge)
Spectral content: significant energy to ~300 MHz

At 300 MHz, Zc of 1000pF ≈ 0.53Ω

Discharge path comparison at ESD frequencies:
Via Bob Smith 1000pF to chassis: ~0.53Ω + trace ← ESD prefers this path
Via transformer winding to PHY: hundreds of Ω ← ESD avoids this path

→ Bob Smith network provides first-stage ESD diversion
→ TVS diode handles residual large-energy events
PCB Layout Rules
Rule Target Why
─────────────────────────────────────────────────────────────────────
Trace: CT pin to 75Ω < 5mm 1mm ≈ 1nH; 10nH = 18.8Ω @ 300MHz
Trace: common node to 1000pF < 3mm Same inductance concern
Chassis ground pour Under transformer Low-impedance reference plane
Star ground point 1 controlled only Avoid ground loops
75Ω resistor placement Adjacent to CT pins Minimize series inductance
1000pF capacitor type C0G/NP0 preferred Stable at HF, low ESR/ESL
Common Implementation Mistakes
Mistake Consequence
──────────────────────────────────────────────────────────────────────
Cap to signal GND instead of chassis Noise injection into PHY
Missing center taps (10/100 ref for GbE) 2 of 4 pairs unterminated → EMC fail
Long traces to 75Ω resistors High series inductance → HF bypassing fails
No chassis ground pour No low-Z return path for common-mode
Cap omitted (only 75Ω to GND) DC ground loop possible; no HF bypass
75Ω omitted (only cap to chassis) No resistive termination; resonances possible
Values: Standard vs Variations
Standard (IEEE 802.3 reference):
75Ω + 1000pF to chassis GND

Acceptable variations:
49.9Ω or 51Ω (lower impedance, application-specific)
2× 1000pF parallel = 2000pF (lower Zc at lower frequencies)

TVS diode in parallel with cap (enhanced ESD clamping)

Not recommended:
0Ω (no resistor): removes damping, can cause resonance
Signal GND connection: as above
Source
Voohu Technology (www.voohuele.com) — 1CT:1CT network transformers with center taps accessible on both windings. 50pcs MOQ, DHL 3–5 days.

Common Mode Choke vs Network Transformer: EMC Stack Design for Ethernet

Mia — Thu, 11 Jun 2026 07:39:45 +0000

Common mode chokes (CMC) and network transformers are frequently confused. They perform different functions and are often used together. Here's the complete technical breakdown.
Functional Comparison
Component Network Transformer Common Mode Choke
────────────────────────────────────────────────────────────────────
Galvanic isolation Yes (1500V AC) No
Impedance match Yes (1:1 for Ethernet) No
Signal coupling Across isolation barrier Conducted (no barrier)
Differential loss < 1.1 dB (spec) < 1 dB (target)
Common-mode reject 40–50 dB (balance + BST) High ZCM (90Ω–1kΩ+)
DC blocking Yes No
PoE support Yes (PoE-rated) Requires PoE-rated CMC
IEEE 802.3 mandate Mandatory Optional
Signal Path Architecture
PHY chip
│ (differential signal, 100Ω)
▼
Network Transformer (1CT:1CT)
│ Provides: isolation, impedance match, differential coupling
│ Bob Smith termination → chassis GND (common-mode drain)
▼
Common Mode Choke ← placed HERE, between transformer and RJ45
│ Provides: second-stage common-mode suppression
│ ZCM >> ZDiff (transparent to data, blocks common-mode)
▼
RJ45 Connector
│
▼
Cable (potential antenna for common-mode radiation)
Why Transformer Alone May Fail Class B

Simplified radiated emissions estimation

Cable radiation from common-mode current

freq_MHz = 100 # Frequency of concern
cable_length_m = 1.0 # Ethernet cable length
I_cm_uA = 5 # Residual common-mode current after transformer (µA)

Field strength at 10m (simplified dipole model, worst case)

E ≈ 60 × π × I × L × f / (c × d) [V/m]

import math
c = 3e8 # speed of light m/s
d = 10 # measurement distance m
f = freq_MHz * 1e6

E_Vm = 60 * math.pi * (I_cm_uA * 1e-6) * cable_length_m * f / (c * d)
E_dBuVm = 20 * math.log10(E_Vm * 1e6)

print(f"Estimated field: {E_dBuVm:.1f} dBµV/m @ {freq_MHz}MHz, {d}m")
print(f"CISPR 22 Class B limit: 40.0 dBµV/m @ 100MHz")

Even 5µA common-mode on a 1m cable can approach Class B limit

CMC Key Specifications
**
Parameter Target Value Notes
──────────────────────────────────────────────────────────────────
ZCM @ 100MHz > 90Ω (Class B) Higher = more suppression
> 600Ω (industrial/medical)
Differential ins. loss < 1 dB (1–100 MHz) Must not degrade data signal
Rated current ≥ PoE class if PoE Standard CMC saturates under PoE DC
Frequency curve Flat 10–200 MHz Check manufacturer's impedance plot
**PoE + CMC: The Saturation Problem
Standard CMC under PoE DC bias:
DC flows as common-mode current through both CMC windings
→ Flux adds in core (not cancels — it's common-mode)
→ Core saturates
→ Differential insertion loss rises
→ Signal quality degrades

Solution options:

PoE-rated CMC (air-gapped core, rated for DC bias current)
Integrated magnetic jack with PoE-compatible magnetics throughout
Place CMC before PoE power injection point (topology-dependent) EMC Design Decision Tree Need CE marking / FCC Class B? ├── YES → Include CMC footprint (populate if needed after EMC test) └── NO (Class A or industrial only) → Transformer + Bob Smith may suffice

PoE enabled?
├── YES → PoE-rated CMC or integrated PoE magnetic jack
└── NO → Standard CMC

Emissions failure above 100MHz at 10m?
└── Almost certainly cable common-mode → add or upgrade CMC
Source
Voohu Technology (www.voohuele.com) — network transformers (standard + PoE-rated), BOM sourcing support including CMC cross-reference. MOQ 50pcs, DHL 3–5 days.

1CT:1CT Network Transformer: Turns Ratio, Center Tap Functions, and PoE Architecture

Mia — Wed, 10 Jun 2026 06:18:23 +0000

The 1CT:1CT notation on every Ethernet transformer
datasheet encodes three distinct electrical features. Here's the complete technical breakdown.
Notation Decoded
1CT : 1CT
│ │
│ └── Secondary winding: 1 turn, with Center Tap
└───────── Primary winding: 1 turn, with Center Tap

Meaning:
Turns ratio: 1:1 (no voltage transformation)
Impedance ratio: 1:1 (Z_sec = Z_pri × (1/1)² = Z_pri)
Center taps: Both primary and secondary — BOTH accessible
Why 1:1 Is Fixed for Ethernet

Transformer impedance transformation

Z_secondary = Z_primary × (N_sec / N_pri)^2

Z_primary = 100 # Ω — Ethernet differential impedance

ratios = [1.0, 1.5, 2.0]
for r in ratios:
Z_sec = Z_primary * (r ** 2)
match = "✅ matched" if Z_sec == 100 else f"❌ mismatch → {Z_sec}Ω"
print(f" Ratio 1:{r} → Z_secondary = {Z_sec}Ω {match}")

Output:

Ratio 1:1.0 → Z_secondary = 100.0Ω ✅ matched

Ratio 1:1.5 → Z_secondary = 225.0Ω ❌ mismatch → 225.0Ω

Ratio 1:2.0 → Z_secondary = 400.0Ω ❌ mismatch → 400.0Ω

Any ratio other than 1:1 breaks 100Ω impedance continuity → return loss degradation → link errors.
Center Tap: Three Functions
Function 1 — Bob Smith Termination
Primary side center tap (PHY side):
CT_PRI ──── 75Ω ──┐
├── CTAP_NODE ──── 1000pF ──── CHASSIS GND
CT_SEC ──── 75Ω ──┘ (cable side)

Effect on signal:
Differential mode: CT is at 0V differential → termination invisible to data
Common mode: CT collects common-mode voltage → discharged to chassis

Missing termination consequences:
❌ Radiated emissions fail at 30m cable test
❌ ESD events reach PHY chip directly
Function 2 — PoE Alternative A Power Injection
Alternative A PoE current path (802.3af/at):

PSE (+) ── cable pair conductor 1 ──┐
│ (DC superimposed on data)
PSE (−) ── cable pair conductor 2 ──┘

DC flows: CT_SEC (cable) → through winding conductors → CT_PRI (PHY side)

Why CT makes this work:
CT at differential zero → DC doesn't modulate differential signal
CT provides the return path for DC current flow

Without CT: Alt A PoE is architecturally impossible
Function 3 — Winding Balance Test Reference
Balance measurement:
Apply common-mode voltage to input pair
Measure differential output voltage at CT reference

Longitudinal Balance (LB) = 20 × log10(Vcm_in / Vdiff_out)

IEEE 802.3 requirement: LB > 40 dB (1–100 MHz)
Good transformers: LB > 50 dB

Imbalance effect:
External common-mode noise → partial conversion to differential noise
→ Degrades receiver SNR without obvious cause
1000BASE-T: Four Independent 1CT:1CT Sets
Package contains: 4 × 1CT:1CT winding pairs (one per Ethernet pair)

Each set:
Primary: TX+, TX−, CT_PRI
Secondary: RX+, RX−, CT_SEC

Total center taps in Gigabit transformer: 8 (4 primary + 4 secondary)
Bob Smith termination networks required: 4 (one per pair CT)

Common mistake:
Using 10/100 reference (2 CTs) for Gigabit PHY (needs 4 CTs)
→ 2 pairs unterminated → EMC failure
Datasheet Red Flag: "1:1" Without CT
"1:1 ratio" (no CT notation)
→ Center tap may NOT be accessible
→ Bob Smith termination impossible
→ PoE Alt A impossible
→ Will fail EMC compliance

Always verify: "1CT:1CT" or explicit statement of center-tapped windings
Source
Voohu Technology (www.voohuele.com) — 1CT:1CT transformers for 10/100 and 1000BASE-T, SMD and THT, standard and PoE-rated. MOQ 50pcs, DHL 3–5 days.

SMD vs THT Network Transformers: Package Selection Guide with Trade-off Analysis

Mia — Tue, 09 Jun 2026 07:22:44 +0000

Choosing between SMD and THT for your network transformer affects assembly process, mechanical reliability, and serviceability. Here's the complete technical breakdown.
Package Overview
SMD (Surface Mount Device):
Mounting: Soldered to board surface (reflow)
Footprint: Pads only, no drilled holes required
Assembly: Automated pick-and-place + reflow oven
Typical size: 8-pin (10/100) or 16-pin (Gigabit) SOIC/SOP

THT (Through-Hole Technology):
Mounting: Leads through drilled/plated holes (wave or hand solder)
Footprint: Plated through-holes + body clearance
Assembly: AI machine or hand-insertion + wave/selective solder
Typical size: DIP or SIP, 2.54mm lead pitch
Mechanical Strength Comparison
Joint type Pull strength Shear strength Vibration rating
────────────────────────────────────────────────────────────────────────
SMD solder pad Low–Moderate Low IPC Class 1–2
THT plated hole High High IPC Class 2–3
THT joints are anchored through the board. Under sustained vibration (IEC 60068-2-6) or mechanical shock (IEC 60068-2-27), THT joints outperform SMD across all standard test profiles.
Assembly Process Requirements
assembly_requirements = {
"SMD": {
"process": "Screen print → Pick & Place → Reflow",
"equipment": ["stencil printer", "pick-and-place", "reflow oven"],
"volume": "Best at medium–high volume",
"prototype": "Requires SMT equipment (or hot plate/hot air)",
"rework": "Hot air station + specialized tools"
},
"THT": {
"process": "Insert → Wave solder / Hand solder",
"equipment": ["wave solder machine", "OR soldering iron only"],
"volume": "Viable at any volume; higher cost at scale",
"prototype": "Hand-solderable with standard iron",
"rework": "Standard soldering iron + solder pump"
}
}
PCB Layout Impact
Parameter SMD THT
──────────────────────────────────────────────────────────────
Board real estate Low (surface only) Higher (hole + keepout)
Component density High Lower
Double-sided loading Yes (both sides) Limited (hole access needed)
Minimum footprint ~5×5mm (10/100 8-pin) ~12×8mm (DIP equivalent)
Height profile Low (<4mm typical) Higher (leads + body)
Lead inductance Near zero Small (~1–3nH per lead)
Application Selection Guide
Application Recommended Package
──────────────────────────────────────────────────────────
Consumer electronics / IoT SMD
PCBA modules (standard assembly) SMD
Industrial gateway (DIN rail) THT preferred
Transportation / automotive THT preferred
Medical device (long service life) THT (field serviceability)
Outdoor enclosure (thermal cycling) THT preferred
Engineering prototype (no SMT equip) THT
High-volume production (cost focus) SMD
Lead-Free Compliance Check
Both packages available in:
HASL-LF (Hot Air Solder Level, Lead-Free)
NiAu / ENIG (Electroless Nickel Immersion Gold)
Sn finish

Solder compatibility:
SMD reflow: SAC305 (Sn96.5/Ag3/Cu0.5), peak 245–260°C
THT wave: SAC305 or SN100C, pot temp 260–270°C

Verify: transformer body rated for reflow temperature if SMD
(should be ≥ 260°C peak, check Tmax in datasheet)
Mixed Technology Strategy
Common approach for ruggedized Ethernet ports:

RJ45 connector: THT → mechanical strength for plug cycles
Network transformer: SMD → signal performance, density
PHY chip: SMD → no THT option available

Assembly flow:
1. SMT → reflow (PHY + transformer)
2. THT → selective solder or hand (RJ45 connector)
Source
Voohu Technology (www.voohuele.com) — network transformers in both SMD and THT packages, 10/100 and Gigabit, standard and PoE-rated. MOQ 50pcs, DHL 3–5 days.

PoE Transformer Specs: DC Bias, OCL Derating, and Thermal Design

Mia — Mon, 08 Jun 2026 06:57:29 +0000

Standard network transformers fail in PoE circuits because DC bias saturates the core. Here's the complete technical breakdown of what to specify and how to calculate thermal performance.
Why Standard Transformers Fail Under DC Bias
Standard transformer: designed for AC signals only
→ No DC current → no DC flux → core stays in linear region
→ OCL stable at rated value (e.g., 1000µH for 1000BASE-T)

PoE transformer: DC bias current from PoE PSE/PD
→ DC current → DC flux consumes core's flux capacity
→ AC signal flux has less headroom
→ Core saturates → OCL collapses → link fails

Typical OCL derating in a standard transformer under 600mA DC bias:
Rated OCL (0mA bias): 1000µH
Effective OCL (600mA): ~150–300µH ← BELOW 1000BASE-T minimum
Result: link unstable, drops under load
Key PoE Transformer Specifications
Spec to check What you need
──────────────────────────────────────────────────────
OCL at rated DC bias ≥ 1000µH (1000BASE-T) under max PoE current
DCR per winding As low as possible — drives thermal load
Current rating ≥ max PoE class current per winding
Core type Air-gapped (prevents saturation under DC bias)
Isolation voltage ≥ 1500V AC (unchanged from standard)
Operating temperature Match to deployment environment
OCL Under Bias: What to Look for in a Datasheet
Good PoE transformer datasheet shows:
OCL @ 0mA: 1200µH (headroom above minimum)
OCL @ 350mA: 1050µH (still above 1000µH → valid for 802.3af)
OCL @ 600mA: 1010µH (barely above minimum → acceptable for 802.3at)
OCL @ 960mA: 980µH (below minimum → NOT suitable for 802.3bt Type 4)

If the datasheet only shows zero-bias OCL: ask for the bias curve,
or test it yourself at rated current before committing to the part.
DCR Thermal Calculation

PoE transformer self-heating calculator

def poe_thermal_dissipation(current_A, dcr_ohms, num_pairs):
"""
current_A: DC bias current per pair (A)
dcr_ohms: DCR per winding (Ω)
num_pairs: number of pairs carrying DC (2 for af/at, 4 for bt)
"""
power_per_winding = current_A ** 2 * dcr_ohms
# Each pair has 2 windings (primary + secondary)
total_power = power_per_winding * 2 * num_pairs
return total_power

Examples:

print(poe_thermal_dissipation(0.35, 0.5, 2)) # 802.3af: 0.245 W
print(poe_thermal_dissipation(0.60, 0.5, 2)) # 802.3at: 0.720 W
print(poe_thermal_dissipation(0.96, 0.5, 4)) # 802.3bt T4: 3.686 W
Alt A vs Alt B Power Delivery
Alternative A: DC on data pairs (1-2, 3-6) via center taps
→ Transformer windings carry both data signal AND DC power
→ Center tap used for both Bob Smith termination AND PoE injection
→ Most common in modern PoE equipment

Alternative B: DC on spare pairs (4-5, 7-8)
→ Data transformer windings carry signal only (no DC bias)
→ Separate magnetics or choke may be used on spare pairs
→ Not usable in 1000BASE-T (no spare pairs)
Design Checklist for PoE Magnetics
[ ] Transformer is explicitly PoE-rated (not just "1000BASE-T rated")
[ ] OCL at rated DC bias ≥ 1000µH (get the bias curve)
[ ] Current rating per winding ≥ PoE class maximum
[ ] DCR thermal load calculated and within package rating
[ ] PoE class matches transformer rating (don't underrate)
[ ] For 802.3bt: all 4 pairs PoE-rated
[ ] Standard 1500V AC isolation still verified
Source
Voohu Technology (www.voohuele.com) — PoE-rated 1000BASE-T transformers for 802.3af/at/bt. OCL-under-bias specs provided. 50pcs MOQ, DHL 3–5 days.