Bifilar Line Input Transformer?

[Clbraddock]

I am in the process of sourcing parts for a P2P pultec build and saw that vintagewindings.com makes two different versions of the HS-56. One version is a clone of the original. The other one is bifilar wound and claims, "Bifilar wound like a Peerless K-241-D using the original Triad 56 data."

I asked the maker about it and they sent me:

"Hi, Thanks for your question. Here is an excerpt from a paper that I'm working on currently:
When a standard shell type transformer is wound, a Triad HS-52 for example, one half of the low side is wound, followed by one half of the high side, followed by the second half of the high side, and finally, the second half of the low side is wound. The result of this arrangement means that the dcr of the first half of each winding will be lower than the dcr of the second halves. This is due to the simple fact that the coil grows larger as the winding progresses which means more wire is used on the later windings. Additionally, the coupling of the windings to the core will be slightly different because each winding has a different orientation to it. Parallel winding solves all of this. By winding the low side together, side by side, and the high side, pie wound at the same time with each side having the exact same orientation to the core. This is what Peerless means by “balanced capacities” in the 20 -20 line. It is a superior way to wind a coil. It is also more difficult to do which is why Triad never did it. Parallel winding also keeps the voltage differential between the layers of the high side lower as there are fewer windings in each layer and this is one thing that reduces capacitive effects. Thanks again and have a fun day!, Chris Vintagewindings"

Unfortunately, I don't really understand what this means in the context of an audio signal. What would the purpose/benefit of a bifilar line input transformer be (i.e. better distortion spec?, more flat frequency response?)

Thanks!

 

[ruffrecords]

A lot depends on whether the transformer is used for input or output. In general bifilar winding is more suitable for output transformers.

Cheers

Ian

 

[Clbraddock]

A lot depends on whether the transformer is used for input or output. In general bifilar winding is more suitable for output transformers.

Cheers

Ian

In this instance it would be a line level input transformer.

 

[Walter66]

Bifilar is a superior way of winding. It can result in greater bandwith due to improved magnetic coupling of the signals. Thats what I would resume it.
In general, the Triad engineers were not dumber than the Peerless ones. Both have different sound characters. Both are excellent in their original design.

 

[ruffrecords]

In this instance it would be a line level input transformer.

Bifilar winding reduces leakage inductance and hence improves high frequency response. However, bifilar winding increase the capacitive coupling between primary and secondary windings. This means common mode interference is more easily coupled to the secondary. To minimise this, many input transformers incorporate a Faraday shield between primary and secondary. Bifilar winding does not allow the use of a Faraday screen. As with all engineering it is a balance and a compromise. Bifilar windings are commonly used in audio output transformers. Most audio input transformers are not bifilar wound.

Cheers

Ian

 

[CJ]

well that guy was using an output transformer to compare with your question on an input transformer so his response was a little garbled for what you are asking,

on the K-241-D, Peerless used bi-fi on the primary only,

you can get away with this with a good nickel core as your turns count will be low enough to prevent excess capacitance between the adjacent primary wire lengths used,

you can use a faraday shield with bi fi windings in the case of the k-241-d as the structure is pri - sec pri

so you have 2 copper shields , one between the inner pri and sec and one between the sec and outer pri

 

[LoudSpeakerPro]

I am in the process of sourcing parts for a P2P pultec build and saw that vintagewindings.com makes two different versions of the HS-56. One version is a clone of the original. The other one is bifilar wound and claims, "Bifilar wound like a Peerless K-241-D using the original Triad 56 data."

I asked the maker about it and they sent me:

"Hi, Thanks for your question. Here is an excerpt from a paper that I'm working on currently:
When a standard shell type transformer is wound, a Triad HS-52 for example, one half of the low side is wound, followed by one half of the high side, followed by the second half of the high side, and finally, the second half of the low side is wound. The result of this arrangement means that the dcr of the first half of each winding will be lower than the dcr of the second halves. This is due to the simple fact that the coil grows larger as the winding progresses which means more wire is used on the later windings. Additionally, the coupling of the windings to the core will be slightly different because each winding has a different orientation to it. Parallel winding solves all of this. By winding the low side together, side by side, and the high side, pie wound at the same time with each side having the exact same orientation to the core. This is what Peerless means by “balanced capacities” in the 20 -20 line. It is a superior way to wind a coil. It is also more difficult to do which is why Triad never did it. Parallel winding also keeps the voltage differential between the layers of the high side lower as there are fewer windings in each layer and this is one thing that reduces capacitive effects. Thanks again and have a fun day!, Chris Vintagewindings"

Unfortunately, I don't really understand what this means in the context of an audio signal. What would the purpose/benefit of a bifilar line input transformer be (i.e. better distortion spec?, more flat frequency response?)

Thanks!

One of the tricks in the tool-box of transformer manufacturers is choice of wire gauge. In a split winding transformer, it's highly desirable to have both windings have the same number of turns ... but the second winding, wound over the first takes a longer length of wire because of the increasing diameter. This results in a higher DCR. The solution is to use for example 45 gauge wire on the inner winding, and 44 gauge wire on the outer winding. The number of turns stay the same, but the DCR can be more closely matched.

The other thing to be concerned about is transformer termination. Impedance ratings are a easy way to pick transformers, but they are almost never represent the optimum loading for the transformer. I can personally attest to the fact that changing input and output loading on good transformers can result in orders of magnitude improvement in harmonic distortion. That's moving the decimal point two places!

 

[gyraf]

Q: Why/when would it be important to closely match DC resistances of windings? I find it hard to think up a real-world scenario where it becomes significant?

/Jakob E.