1. What Happens When You Send an Incomplete Specification
Every week, we receive RFQs that look like this:
> “Need a transformer. 48V input, 12V output, 5A. Please quote.”
This is not a specification — it’s a conversation starter. And it guarantees one of three outcomes:
The supplier asks 15 follow-up questions. This costs you a week of email ping-pong while your competitor moves ahead.
The supplier guesses the missing parameters. The sample arrives, doesn’t work, and you lose two weeks.
The supplier declines to quote. Either they lack the missing information to calculate cost, or they assume an inexperienced engineer means a high-risk project.
At Linkpower, we review over 200 custom transformer RFQs per month. The engineers who get accurate quotes on the first pass share one thing in common: they include the right information in the right format.
This guide shows you exactly what to include, why it matters from the manufacturer’s side, and how to avoid the seven most common spec mistakes.
2. The Seven Essential Sections of a Transformer Specification
A complete custom transformer specification has seven sections. Omitting any one of them introduces uncertainty — and uncertainty in a quote means the supplier pads the price.
| Section | What It Tells the Manufacturer | Risk if Missing |
|---|---|---|
| 1. Electrical Requirements | Turns ratio, inductance, power level | Wrong core size, wrong saturation margin |
| 2. Mechanical Drawing | Pinout, mounting, dimensions, height limit | Mechanical rework, PCB redesign |
| 3. Winding Specification | Wire gauge, layers, insulation, direction | Wrong DCR, failed Hi-Pot, poor coupling |
| 4. Core Information | Core type, material grade, gap | Wrong AL value, wrong frequency response |
| 5. Test Conditions | Frequency, voltage, temperature, load | Spec passed in test but fails in application |
| 6. Environmental Requirements | Operating temp, humidity, vibration | Field failures, warranty returns |
| 7. Certification & Compliance | UL, RoHS, REACH, flammability | Product blocked at customer receiving |
📌 link: For a deep dive into core material selection, see Magnetic Core Materials Technical Guide.
3. Section 1: Electrical Requirements — The Non-Negotiables
This is the heart of your specification. The manufacturer needs these numbers to select the core, calculate turns, and design the winding.
Minimum required parameters for any high-frequency transformer:
| Parameter | Example Value | Why It Matters to the Manufacturer |
|---|---|---|
| Input voltage range (nominal) | 36V – 75V DC | Determines duty cycle range and MOSFET stress |
| Output voltage(s) | 12V ± 5% | Determines turns ratio |
| Output current / power | 2.5A / 30W | Determines core size and wire gauge |
| Switching frequency | 200 kHz | Determines core material selection (PC44 vs PC95) |
| Topology | Flyback, forward, LLC | Determines whether the core needs an air gap and how to calculate inductance |
| Isolation voltage | 1500Vrms / 1 min | Determines insulation thickness and creepage distance |
| Efficiency target | ≥ 90% | Sets the loss budget for core + copper |
Optional but highly recommended:
| Parameter | Why It Helps |
|---|---|
| Maximum height / footprint | Prevents designing a core that won’t fit your enclosure |
| Ambient temperature | Affects thermal derating — a 40°C vs 70°C design uses different core sizes |
| Cooling method (convection / forced air / conduction) | A 30W design in still air needs 20% more core surface than one with 2m/s forced air |
| Intended standard (if applicable) | IEEE 802.3bt, AEC-Q200, UL 62368-1 — each imposes specific test requirements |
4. Section 2: Mechanical Drawing — What to Include
A mechanical drawing eliminates ambiguity. It does not need to be a CAD file — a hand sketch with three critical dimensions is better than a perfect PDF with no dimensions.
The manufacturer expects to see:
Top view: Pin numbering (1, 2, 3…) viewed from the top of the component. Pin 1 location must match the datasheet orientation.
Side view: Maximum height above PCB. This is the single most common rework reason — the transformer is 0.5mm too tall for the enclosure.
Bottom view: Pin footprint pattern — hole diameter, pad diameter, pitch between pins. For SMD parts, include land pattern dimensions.
Mechanical notes to include:
- Pin material preference (brass / copper alloy / tin-plated)
- If the part must be potted or vacuum impregnated
- Any keep-out zones on the PCB adjacent to the transformer
- Whether the assembly will be wave-soldered or reflow-soldered — this affects pin plating requirements
Manufacturer’s insight: The most common RFQ revision we request is a dimensioned drawing. If your spec arrives without one, the first email back will ask for it. Including it from the start saves 24–48 hours.
5. Section 3: Winding Specification — The Section Most Engineers Get Wrong
Engineers often specify turns ratio and inductance but leave the winding construction to the manufacturer’s discretion. This is acceptable for simple designs but dangerous for high-frequency transformers where parasitic effects dominate performance.
What to specify:
| Winding Parameter | Why It Matters |
|---|---|
| Number of turns per winding | The manufacturer needs this to calculate DCR, not just turns ratio |
| Wire gauge (AWG) | Specifying AWG avoids receiving a design with wire too thin (high DCR) or too thick (won’t fit the bobbin) |
| Number of strands (for bifilar / Litz) | Reduces proximity effect in high-current windings |
| Layer arrangement (e.g., sandwich: P/2 – S – P/2) | Critical for leakage inductance control |
| Interlayer insulation (tape type and layers) | Affects Hi-Pot withstand and creepage distance |
Example of a well-specified winding:
> Primary: 21T × 2 layers, AWG32, 1 layer of 3M 1298 tape between layers > Secondary: 4T × 1 layer, 2-strand AWG26 bifilar > Auxiliary: 4T × 1 layer, AWG34 > Winding order (from bobbin pin 1 side): P(1–11T) → Tape → S(1–4T) + Aux(1–4T) → Tape → P(12–21T)
Manufacturer’s insight: When the winding order is not specified, the manufacturer optimizes for cost (minimum labor), not for performance (minimum leakage inductance). If you need low leakage, specify sandwich winding explicitly.
6. Section 4: Core Information — How Specific Should You Be?
For production designs, you should specify both the core type and the core material grade.
Core type selection:
| Core Shape | Best For | Power Range | Typical Application |
|---|---|---|---|
| EE / EI | General purpose, easy to wind | 5W – 500W | Flyback, forward converters |
| PQ / LP | Low profile, good EMI shielding | 10W – 300W | PoE PDs, DC-DC modules |
| EFD | Low profile, good for planar | 5W – 150W | Flat designs, GaN converters |
| RM | Good shielding, compact | 1W – 100W | Telecom, board-mounted |
| Pot core | Best EMI shielding | 1W – 50W | Filter inductors, high-isolation |
| Toroid | Lowest EMI, highest cost | 10W – 1000W | PFC inductors, EMI chokes |
📌 link: Core shape selection interacts with winding loss — see How Transformer Winding Structure Affects Loss, EMI, and Thermal Performance.
Core material grade — be specific:
Writing “ferrite core” is not sufficient. A PC40 (3C90) core operating at 200kHz will run 3× hotter than a PC95 (3C95) core at the same flux density.
| Material Grade | Frequency Range | Core Loss @ 100kHz / 0.2T | Cost Index |
|---|---|---|---|
| PC40 / 3C90 | 20 – 100 kHz | ~600 mW/cm³ | 1.0x (baseline) |
| PC44 / 3C92 | 50 – 200 kHz | ~350 mW/cm³ | 1.3x |
| PC95 / 3C95 | 100 – 500 kHz | ~200 mW/cm³ | 1.8x |
| N49 / 3C97 | 200 – 1000 kHz | ~100 mW/cm³ | 2.5x |
Manufacturer’s insight: If you specify “PC44 or equivalent,” the manufacturer may substitute a lower-cost local material that meets the datasheet specifications at 25°C but performs 30% worse at 100°C. If the thermal performance matters, write “PC44, no substitution without written approval.”
7. Section 5: Test Conditions — The Most Overlooked Section
A transformer specification without test conditions is a recipe for confusion. Here’s a real example from our engineering team:
> Engineer specifies: Lp = 200µH ± 10%. > We measure at 100kHz, 0.25Vrms, 25°C — passes at 208µH. > Engineer measures at 10kHz, 1Vrms, 25°C — gets 185µH. > Result: Two weeks of emails trying to understand why the measurement doesn’t match.
Specify test conditions explicitly:
| Parameter | Test Condition to Include |
|---|---|
| Inductance (L) | Frequency, voltage level, DC bias current |
| DCR | Temperature (typically 25°C) |
| Turns ratio | Which windings are measured (e.g., Np:Ns1, Np:Ns2) |
| Leakage inductance | Which winding is shorted, test frequency |
| Hi-Pot | Voltage, duration, leakage current limit |
| Insulation resistance | Voltage applied, minimum resistance |
Example of well-specified test conditions:
> Lp = 180µH ± 10% measured at 100kHz, 0.5Vrms, 0A DC bias, 25°C. > Lleak = 5µH max, measured at 100kHz with secondary shorted. > Hi-Pot: 1500Vrms, 60s, leakage ≤ 1mA, between primary and secondary.
8. Section 6: Environmental Requirements — Don’t Assume Room Temperature
Transformers fail more often from environmental stress than from electrical overstress. Specifying the operating environment helps the manufacturer select the correct materials.
Minimum environmental specification:
| Requirement | Typical Value | Material Impact |
|---|---|---|
| Operating temperature range | -40°C to +105°C | Determines insulation class (Class B, F, H) |
| Storage temperature range | -40°C to +125°C | Affects varnish and tape selection |
| Humidity | 85% RH, non-condensing | Affects pin plating corrosion resistance |
| Vibration | 10 – 500Hz, 2G | Affects core bonding (gapped cores need adhesive) |
| Altitude | ≤ 3000m | Affects creepage distance (derating above 2000m) |
Manufacturer’s insight: If your product requires the Double 85 test (85°C / 85% RH), specify it explicitly. Not all standard transformer materials pass Double 85 — the varnish system and pin plating must be selected specifically for this condition.
📌 link: For details on the Double 85 test methodology, see our Enameled Wire Pinhole Test guide.
9. Section 7: Certification Requirements — Specify Early, Not at Shipment
The most expensive change request we receive is: “We need UL certification” — received the day before shipment.
UL certification of a custom transformer is not a simple label. It requires:
- UL-recognized materials (core, bobbin, tape, wire, varnish, sleeving)
- UL-tracked production line (separate from non-UL production)
- Periodic UL witness testing (audit cost, sample cost)
- Documentation of every material change
Specify certification requirements in the RFQ:
| Certification | Lead Time Impact | Cost Impact |
|---|---|---|
| UL 1446 (Insulation System) | 2–4 weeks added | Moderate |
| UL 62368-1 (End-product recognition) | 4–8 weeks | High |
| RoHS / REACH (declaration only) | None | Minimal |
| Conflict Minerals reporting | None | Minimal |
| AEC-Q200 (Automotive) | 8–12 weeks | Very high |
Manufacturer’s insight: If you are targeting UL recognition, ask the manufacturer for their UL Insulation System class (Class B 130°C / Class F 155°C / Class H 180°C) before the design starts. Changing from Class B to Class F materials after the prototype is built often requires a complete winding redesign.
📌 link: Transformer UL certification classes are explained in our Transformer Insulation System guide.
10. The Seven Most Common RFQ Mistakes — From a Manufacturer’s Perspective
After reviewing thousands of RFQs, here are the mistakes that cost engineers the most time:
Mistake 1 — Specifying inductance without DC bias current. A gapped ferrite core rated for 200µH at 0A may measure 140µH at 2A DC bias. Always specify: “Lp = 200µH ± 10% at 0A, minimum 170µH at 2A DC bias.”
Mistake 2 — Giving turns ratio but not the number of turns. “Turns ratio 5:1:1” could mean 50:10:10 or 25:5:5. These give very different DCR values and winding costs. Always specify absolute turns.
Mistake 3 — Omitting the test frequency for inductance. Ferrite permeability drops with frequency. A core measuring 200µH at 1kHz may read 160µH at 100kHz. Specify both the test frequency and the operating frequency.
Mistake 4 — Not specifying maximum leakage inductance. Without a leakage limit, the manufacturer optimizes for cost (simpler winding = higher leakage) rather than performance. Result: the transformer passes all tests but the power supply fails EMI.
Mistake 5 — Specifying height limit but no pin clearance. A 12mm height limit with 5mm pin length leaves only 7mm for the core — which may be too thin for the required power. Specify “Maximum component height above PCB: 12mm” rather than “Transformer height: 12mm.”
Mistake 6 — Requesting “military grade” without specifying a standard. Military grade could mean MIL-STD-981, MIL-PRF-27, or nothing at all. If you need compliance with a standard, cite the full document number and revision.
Mistake 7 — Sending a PDF datasheet of a competitor’s part number. A PDF tells us the performance of the competitor’s transformer, but not your specific requirements. We cannot guarantee a direct replacement without understanding your circuit conditions. Instead, provide a parametric specification with your target values.
11. Conclusion
Writing a good custom transformer specification is not about filling more pages — it’s about including the right information in the right format. A clear RFQ with well-defined parameters, a dimensioned sketch, and explicit test conditions is worth more than a twenty-page datasheet dump.
Every missing parameter forces the manufacturer to make an assumption. Assumptions lead to incorrect quotes, wrong samples, and delayed schedules.
At Linkpower, our engineering team reviews every RFQ against a standardized checklist before quoting. If your specification includes all seven sections outlined in this guide, we can typically provide a firm quote and prototype lead time within 24 hours.
Ready to get started? Send your transformer specification to our engineering team at https://lpema.com/contact/ — we’ll review it and provide a detailed quote, usually within one business day.
📌 link: Interested in how fast we can turn prototypes? See how Linkpower delivers custom transformer samples in 24 hours.
