AN-104
Transformer Sonics and Lamination Influences
A common class of questions we receive is about transformer sonics. Can the transformer deliver additional qualities to the audio? If so, what can it really do? What gets sacrificed by doing a particular thing? etc. Usually this is described subjectively, such as “fatter bass response”, or “more brilliance”, or “richer sound.” The descriptions are numerous. The solutions in the transformer engineering and application are very many and continuously evolving. Winding topologies and core materials work together to provide solutions.
Audio signal level transformers can be broken down into a few categories. The contributions to the signal path vary inside each of these classifications.
High quality audio transformers must perform over an extreme dynamic range and over a fairly wide frequency range. Human hearing reaches a serious pain level which can be dangerous to hearing at 120dB(A), but we can hear beyond that level. DON’T LISTEN TO ANYTHING THAT LOUD UNLESS YOU WANT TO ENDANGER YOUR HEARING! LIVE CONCERTS CAN EXCEED THAT. It is commonly accepted that people have a dynamic range of 130dB, with 0dB being the threshold. Anechoic chambers can be as quiet as -20dB(A). Some people can hear below 0dB. Audio transformers work within the dynamic range, some offering nonlinear response characteristics which must be pleasing to the design engineer as well as the people who use what he makes.
Audiologists say that we have a hearing range of 20Hz-20kHz. Most of us lose the far ends of this spectrum as we grow older. For reasons that will be explained later, frequency response beyond 20kHz actually is audible. We cannot escape the laws of physics.
Direct Box Transformers:
Direct box transformers take a high level signal and reduce it to a low level, most often to drive a microphone preamplifier. They can take a high impedance signal and reduce the impedance to a useful one. Transformers transform the impedance as the square of the turns ratio. If the transformer has a turns ratio of 10:1, the impedance on the secondary is 1/100th of the signal driving the primary.
A good direct box transformer does not change the character of the signal feeding it. Low distortion and excellent bandwidth are necessary here. They should not add to or subtract from the signal.
Lamination quality and the winding structure determine how well these perform. Because they need to work well with high impedance signal sources, the core material must have very high permeability. For this reason, only high-nickel alloys (80% nickel) work well in this application. CineMag only uses this class of alloy. It is precision alloyed. The processing recipe is proprietary, providing maximum permeability and minimal distortion.
Microphone Input Transformers:
Microphone input transformers (“mic input transformers”) boost the signal level for preamplifiers. As explained above, the impedance changes as the square of the turns ratio. Amplifiers only amplify the voltage. Hence, the power of the signal from the microphone (P=E*I) is up-converted to the E component which is what the preamplifier works on. This provides “free” voltage gain which helps with the signal-to-noise ratio of the system.
Mic input transformers can have a meaningful contribution to the character of the signal. This mostly happens with bandwidth and phase shift. For example, some designs will have a peak at the high frequency portion of their range. For example, the CM-75101A peaks at around 60kHz. Because of this, its frequency response slowly rises about 08.dBu from 20Hz to 20kHz. The peaking at 60kHz makes it sound a touch bright.
Phase shift at high frequency can cause sound stage image “smear.” That means that the apparent size and placement of the sound source will be slightly less precise and focused.
These transformers interact with the microphone. Most microphones will change their acoustics depending upon the loading. Audio transformers present a complex reactance which alter the signal.
Mic input transformers should have low THD throughout their operating signal level range. They need high permeability and low distortion laminations for optimal performance.
Microphone Output Transformers:
Microphone output transformers do add a significant “voice” to the audio signal, being second in importance to the element which turns the mechanical energy of the sound wave into an electrical signal.
The transformer interacts with what is driving it, be it an active electronic circuit, a dynamic microphone element, or a ribbon. The configuration of the windings as well as the makeup of the laminations work together to create this voice.
Often, the laminations are all high-nickel alloy. Sometimes, grain oriented silicon steel is introduced. Cobalt can also be used in the lamination structure.
The style of stacking in the laminations also can be used to craft the transformer’s character. For example, alternating groups of different materials can be placed. The positioning of each of these groups and the amounts in the groups do affect the sonics.
Line Input & Splitter Transformers:
Line input transformers are designed to not add anything to the signal. This is because the variability of the drive impedance. Repeatability is important. Without that, unexpected performance issues would be difficult to find.
Only high-nickel laminations are used for these transformers.
Line Output Transformers:
Line output transformers operate at low impedance on the secondary. They can take high voltage signals from vacuum tube stages or line level signals from solid stage circuits. These characteristics give this type of transformer a wide range of sonic possibilities.
One of the first considerations is the type of core material to use:
- High-nickel has very low distortion and the nickel costs a lot.
- High-nickel can be mixed with steel for some sonic character. The steel component has higher distortion than the nickel. some transformer manufacturers use a 50% nickel alloy. We have determined that a mix of high-nickel and steel is superior.
- Grain oriented silicon steel has the most overtones. It is the least costly option and is the most popular, not because of cost but because of its harmonics.
The winding structure plays an important part in the harmonic structure of the signal. Keep in mind that the objective can be to create more harmonics. Some designs will stop the harmonics in the low mid-range, while others will allow the harmonics to continue to the high range of what we can hear.
There are myths about the harmonic structure of audio transformers. A common one is that transformers predominately have even harmonics. Typically, they have both even and odd harmonics.
This FFT shows the harmonic content for the CMOB-2S (steel laminations):
Frequency Response Beyond Hearing Range:
Most adults cannot hear beyond 18kHz. Frequencies beyond 20kHz are audible. The laws of physics are inescapable.
Sound in the real analog world we inhabit consists of many different frequencies of constantly varying amplitudes. Any waveform can be broken down into discrete sine waves. [This mathematical concept was established by French mathematician and physicist Joseph Fourier in 1807. Equations based on his work are use to calculate approximations of a chosen accuracy and are called Fast Fourier Transform. Hence the term “FFT”.]
These sine waves vary in amplitude, sometimes very rapidly. Consider one of these sinewaves at 22kHz, varying in amplitude 5,000 times per second. The fundamental frequency is being amplitude modulated. Hence, there are two sidebands. The upper sideband is at 27kHz, which we cannot hear. The lower sideband is at 17kHz which we do hear.
4487 Ish Drive, Simi Valley, CA 93063
(818) 993-4644
cinemag@cinemag.com
