Distortion in audio transformers is not as straightforward as it is in amplifiers — it often only appears under specific frequencies, levels, or DC conditions. Some find the low end becomes "muddy" after swapping transformers, others notice the midrange "congests" at high volume, and some measure THD rising abnormally with frequency. To solve these issues, you first need to identify whether the distortion originates from the core, bias, or winding.
Three Main Distortion Types
| Type | Cause | Audible Symptom | Mitigation |
|---|---|---|---|
| Core nonlinearity | Curved magnetization curve, hysteresis | Mid-low frequencies sound rough, increased even-order harmonics | Enlarge core, choose high-permeability material, reduce operating level |
| DC bias saturation | DC component in the primary winding | Low end collapses first, more noticeable at higher volumes | Add air gap, increase saturation margin, block DC |
| Winding asymmetry | Uneven leakage inductance and distributed capacitance | High-frequency roll-off, phase shift, degraded common-mode rejection | Layer and section winding, symmetrical structure, optimized shielding |
Core Nonlinearity
The magnetization curve of a core is not linear — the larger the signal, the closer the operating point moves to the curved region, and distortion rises accordingly. Permalloy offers excellent linearity at low signal levels but has a relatively low saturation flux density; silicon steel has a high saturation point but limited permeability at small signals. When selecting materials, look at where the actual operating level falls on the curve rather than just the initial permeability. Reducing the operating level, increasing the core cross-section, or adding turns can all pull the operating point back into the linear region.
DC Bias Saturation
When single-ended circuits, biasing networks, or phantom power isolation are not fully effective, DC current flows through the primary. The DC pushes the core toward unilateral saturation, causing inductance to drop, low frequencies to attenuate prematurely, and in severe cases, noticeable distortion. Solutions include adding an air gap, choosing materials with higher saturation flux density, or blocking DC at the circuit level. An air gap lowers effective permeability, so turns and inductance must be recalculated. For related selection logic, refer to the Audio Transformer Selection Guide.
Winding Asymmetry
Leakage inductance determines high-frequency roll-off, while distributed capacitance determines high-frequency resonance — both are linked to the winding structure. Simple random winding results in higher leakage inductance and inconsistent distributed capacitance, manifesting as high-frequency attenuation, phase shift, and even peaks or dips at certain frequencies. Layer winding, section winding, and interleaved wire routing can improve these issues but add to manufacturing cost. The placement and grounding of shielding layers also affect high-frequency performance — see the Shielding Scheme Comparison for more details.
Other Easily Overlooked Factors
- Source impedance and load mismatch. This can cause premature low-frequency roll-off or high-frequency resonance, which may be mistaken for distortion.
- Mounting location. Placing the transformer near a power transformer or magnetic chassis can introduce hum and extra distortion via stray flux.
- Potting material. Curing shrinkage alters distributed parameters — re-test after potting.
- Temperature. Core parameters shift with temperature, and saturation margin decreases at high temperatures.
Measurement and Troubleshooting Methods
- Set a 1kHz reference level, measure THD versus frequency to identify the band where distortion rises.
- Fix the frequency, sweep the level, and check whether distortion approaches saturation.
- Re-measure with added DC bias to confirm its impact on the low end.
- Compare no-load and loaded results to rule out load matching issues.
- Test multiple samples of the same spec from different batches to determine consistency issues.
Frequently Asked Questions
Does low THD always mean it sounds good?
Not necessarily. THD only reflects the total amount, not the harmonic distribution. Even-order and odd-order harmonics sound very different, and frequency response and transient behavior must also be considered.
Why is it normal at low volume but distorts at high volume?
This is typically core approaching saturation or DC bias shifting the operating point. You need to calculate the flux density at maximum signal and leave adequate margin.
Will a larger core solve all distortion issues?
It can improve saturation-related distortion, but it does little for high-frequency problems caused by winding asymmetry. Winding structure must also be optimized in parallel.
Distortion increased after potting
Potting material and the curing process change distributed capacitance and mechanical stress. Re-measure frequency response and THD after potting, and adjust winding parameters if necessary.
If your sample shows abnormal distortion in a specific frequency band, send us the test conditions, schematic, and batch number for further analysis: Contact the Technical Team. Related products can be viewed on the Audio Transformer Product Page.
