Mic input transformer question. Principle low frequency design element?

I'm not sure how best describe this question/issue.

Is there a principle design feature, e.g. Turns count, Ratio, Core size, Metal type, Wire size... etc. THAT,

would have the most impact on low frequency headroom and fidelity?

Case in point: All other things being the same between two transformers, the one that has  larger EI lams would give better low frequency response?

Generally use more iron and copper... but I am not the transformer expert here.

JR

Hi mkiijam
You may find that this link may help with some of your questions.
Duke

http://www.hollywoodsapphiregroup.com/insandouts.pdf

mkiijam said:

Case in point: All other things being the same between two transformers, the one that has  larger EI lams would give better low frequency response?

Click to expand...

Not necessarily. Iron size determines the rated power. Frequency response is a complex issue: LF extension depends essentially on inductance, the higher the inductance the lower the LF cut-off, but OTOH increasing inductance tends to also increase leakage inductance, which reduces HF response.
However, increasing the size of the core allows increasing the inductance without increasing the leakage, that's why input transformers are often much bigger than their rated power demands.
Nickel or Permalloy lams result in increased inductance but lower saturation (less headroom).
Output transformers have a different optimization, depending on the circuit that drives them.

LF response depends principally on low frequency inductance which depends on core material and number of primary turns.

Maximum level before serious distortion  ( often quoted at 1%)  depends flux density and core material. Flux density depends on core cross sectional area and number of turns.

Cheers

Ian

phase shift at low frequencies is something to be considered,  not too many transformers put that spec in their data sheets, Jensen does,

DCR is the main culprit for low freq phase shift,

this article is dope>

and thanks for that other link above! 

https://babel.hathitrust.org/cgi/pt?id=uc1.b4154646;view=1up;seq=1

remember, music first, transformers second.

https://www.youtube.com/playlist?list=PLSXEWU-h31MdpqcnIvG_QnEl3UeKbEBwk

note: if the transformer is to be used inside an automobile, frequency response should extend above and below tire and engine noise, but you must balance conflicting factors,

CJ said:

phase shift at low frequencies is something to be considered,  not too many transformers put that spec in their data sheets, Jensen does,

Click to expand...

Transformers are Minimum-Phase devices, hence the phase response is biunivocally related to frequency response; knowing one is knowing the other.
LF response is not an inherent parameter in a xfmr, because it is dependant on the source and load impedances. As a consequence, phase-shift cannot be represented by a single figure.

DCR is the main culprit for low freq phase shift,

Click to expand...

LF Phase-shift is the result of the High-Pass constituted by the global DCr of the source (source Z+DCr of primary) and the primary inductance. DCr and inductance are equally responsible.

Mic input transformers are probably an exception due to the specificity of use, but as a trend, comparing high quality transformers of different eras, you will find very old transformers with inductance so high that it takes a source Z higher than rated to start losing low response.  Many (most?) modern transformers are designed with bridging drive in mind, with the source having very low Z relative to nominal primary impedance, and low frequency losses are a given with anything other than 10x lower drive, matching condition being totally unacceptable.  This is more a matter of translation, the expectations having shifted, the constraints for deriving the rating changed over decades of time, not a failure of design.  You have to know this, because if you compare old apples to new apples as if the same, you will have problems. 

As Abbey stated, there is a direct correlation between phase and frequency response.  On the top, it appears to be as simple as x degree of phase shift = x dB of loss.  The few screen captures I can pull up at the moment appear to be 90º = -6dB. 

EmRR said:

As Abbey stated, there is a direct correlation between phase and frequency response.  On the top, it appears to be as simple as x degree of phase shift = x dB of loss.  The few screen captures I can pull up at the moment appear to be 90º = -6dB.

Click to expand...

That's because the LF response is a 1st-order Hi-Pass, R-L filter. The -3dB roll-off happens when the reactance of the inductance is equal to the resistance. Both vectors being equal but 90° apart, the resultant is at 45°. But -6dB is about 60°.
90° is an asymptote that is approached only at VLF.

"..LF response is not an inherent parameter in a xfmr, because it is dependant on the source and load impedances. .."

source and load impedance are usually specified ,
'

CJ said:

"..LF response is not an inherent parameter in a xfmr, because it is dependant on the source and load impedances. .."

source and load impedance are usually specified ,
'

Click to expand...

A mic input specified for 150 ohms may be connected to a Neumann TLM103 (Z=50 ohms) or a Shure SM 57 (Z=310 ohms)