Output Transformers
The critical link between tube and speaker. How impedance matching, core design, and winding geometry determine the bandwidth, power delivery, and sonic character of a vacuum tube amplifier.
Why Transformers?
Vacuum tubes are high-voltage, high-impedance devices. Speakers are low-voltage, low-impedance loads.
A typical output tube like the EL34 wants to see a load of 3.4kΩ at its plate. Your speaker presents 8Ω. Without an impedance-matching device, nearly all the signal power would be lost — the tube cannot drive the speaker directly.
The output transformer converts high-voltage, low-current energy from the tube into low-voltage, high-current energy for the speaker. It does this through electromagnetic coupling between two coils wound around a shared iron core. The turns ratio between primary and secondary determines the impedance transformation.
Unlike a resistor that dissipates energy as heat, an ideal transformer transfers 100% of the power. In practice, losses from winding resistance, core hysteresis, and eddy currents reduce efficiency to 85–95% for quality designs. The transformer is the single most expensive and most sonically important component in any tube amplifier.
The Turns Ratio
Primary impedance is the square of the turns ratio times the load impedance.
Example: A 25:1 transformer with a 8Ω speaker reflects 5.0kΩ to the plate. If the tube sees its rated plate-to-plate load, maximum power transfers to the speaker with minimum distortion.
SE vs Push-Pull
The two fundamental transformer architectures have profoundly different requirements.
DC bias current flows continuously through the primary winding, magnetizing the core. An air gap is mandatory to prevent core saturation.
The air gap reduces primary inductance, which hurts bass response. SE transformers need larger cores and more turns to compensate — making them expensive.
Even-order harmonics (2nd, 4th) are not cancelled, contributing to the characteristic "warm" SE sound.
Two tubes drive opposite halves of the primary. Their DC currents flow in opposing directions and cancel in the core — no air gap needed.
Without an air gap, full core permeability is available. This means higher primary inductance, better bass, and wider bandwidth for a given core size.
Even-order harmonics cancel in the transformer. The result is lower measured distortion, though some audiophiles prefer the harmonic content of SE designs.
| Parameter | SE | Push-Pull |
|---|---|---|
| Air gap | Required | None |
| DC in core | Full bias current | Cancels to zero |
| Even harmonics | Present | Cancelled |
| Core utilization | Poor (gap reduces L) | Excellent |
| Bass response | Limited by gap | Superior |
| Power / size | Lower | Higher |
| Cost / watt | Higher | Lower |
Frequency Response
Two inductances define the transformer's usable bandwidth: primary inductance controls bass, leakage inductance controls treble.
Primary inductance (L_p) must be large enough that its reactance exceeds the plate resistance at the lowest desired frequency. More turns and a larger core increase L_p, extending bass response. For SE transformers, the air gap dramatically reduces L_p.
Leakage inductance (L_lk) is caused by imperfect magnetic coupling between primary and secondary. It forms a low-pass filter with the load capacitance. Interleaving the windings (P-S-P-S) reduces leakage at the cost of increased inter-winding capacitance.
Screen Grid Taps
Connecting the screen grid to a tap on the primary winding blends pentode power with triode linearity.
In a standard pentode amplifier, the screen grid is connected to a fixed supply voltage. In triode mode, it is connected directly to the plate. The ultralinear connection taps the screen grid at a point between B+ and the plate on the primary winding, typically around 43% from the B+ end.
This creates a form of local voltage feedback: as the plate voltage swings, the screen voltage follows proportionally. The result is significantly lower distortion than pentode mode, with only a modest reduction in output power. The optimal tap point (around 40–45%) was determined empirically by Hafler and Keroes in their landmark 1951 paper.
Practical Specifications
Matching transformers to tubes: recommended primary impedances and expected performance.
| Tube | Type | Primary Z | Power | Bandwidth | Notes |
|---|---|---|---|---|---|
| 300B | SE | 2.5k-3.5kΩ | 7-10W | 20Hz-40kHz | Needs quality air-gapped core |
| 2A3 | SE | 2.5k-5kΩ | 3.5-4.5W | 25Hz-35kHz | Lower power, premium iron essential |
| EL34 | PP | 3.4k-6.6kΩ | 25-50W | 15Hz-50kHz | Classic ultralinear at 43% tap |
| KT88 | PP | 3.4k-6.6kΩ | 35-100W | 12Hz-60kHz | High power, needs robust core |
| 6L6 | PP | 4k-8kΩ | 20-40W | 20Hz-45kHz | Guitar amps: 6.6k typical |
| 6V6 | PP | 5k-10kΩ | 10-18W | 25Hz-40kHz | Lower power, excellent for hifi |
| EL84 | PP | 5k-10kΩ | 10-17W | 20Hz-45kHz | Compact, great UL performance |
| 45 | SE | 3k-5kΩ | 1.5-2W | 30Hz-30kHz | Low power DHT, needs best iron |
Selecting a transformer: The primary impedance should match the tube manufacturer's recommended plate-to-plate (PP) or plate-to-B+ (SE) load impedance. Using a lower impedance increases power but also distortion; a higher impedance reduces distortion but limits maximum output before clipping.
Core material matters: M6 grain-oriented silicon steel is standard. Premium transformers use nickel alloys (Permalloy, mu-metal) for the inner laminations to reduce hysteresis distortion at low levels. Amorphous and nanocrystalline cores offer the lowest losses but at significantly higher cost.
Key Relationships
Essential formulas for output transformer design and analysis.
Test Your Knowledge
Validate your understanding of output transformer design before moving on.
Why can't a tube drive a speaker directly without a transformer?