Dear visitor,
There are probably more details in audio transformers, for these to “sound good”, than most people actually think.
When I first started building these a few years ago, I strived for the best electrical parameters possible. Believing simple measurements, such as minimizing low and high frequency distortion and phase shift, I started making prototypes with incredible bandwidth, like a 3Hz to 150kHz at -3dB output transformer with a grounded secondary.
Some time later, learning the hard way, I realized bandwidth and distortion had hardly as much significance on the sound of a transformer as the materials and the way they’re combined.
In the recent years, choices of exotic magnetic materials for transformer cores have become popular. They all have their impact on the sound signature, and to be honest with you folks, there’s a taste for everyone. The most popular ones are grain oriented silicon steel (aka HiB), amorphous, nanocrystalline and high permeability nickel alloys.
It lies in the hands and mind of the transformer engineer and winder to combine the best materials with the chosen core.
Popular dielectrics for transformer windings are Paper, Mylar, Teflon, Nomex, Nomex laminate (aka NMN) and Polyethylene. Every dielectric has an unique sound signature and there is no best. Although honestly, paper sounds the most neutral, but it is hard to use it elsewhere in most applications.
There are combinations to be avoided, for example specific dielectrics with a specific core material, such as mylar sheet and nanocrystalline core. But sometimes we have to “work around” the materials, because we’re compelled to obey certain electrical parameters, such as dielectric strength and overall parasitic capacitance. This is where the term “transformer tuning” or “voicing” comes to mind. I’ve heard this quite a lot – “voicing an amplifier”. The same is vallable of audio transformers.
In fact, I’ve been able to “voice” a pair of output transformers to sound to an extremely similar way to another pair with very different interleaving and electrical parameters, adding specific materials to strategic spots.
My products are always tuned to the “best for the moment” materials and I’m constantly searching for further perfection.
Sadly, material audibility is significant in all applications – signal transformers, power supplies chokes and even power transformers. I doubt the possible explanation of this phenomena is electrical, but more about mechanical reasons. In fact, scraping against a sheet of insulation and listening to the resulting noise can predict a lot of the material’s fidelity influence on the final product.