Calculating transformer impedance

[Vintageaudioholic]

Hello, I need some help with transformer math...
I have these 2 transformers between my headphone amp's balanced output and unbalanced output, I would like to upgrade them but I don't know what transformers are suitable replacements in place of the originals so here we go...
Desoldering the transformers and measuring open circuit with no load, the transformers show that measuring at 1kHz, the Inductance is 185.2mH with a Q-factor of 1.18. Both ends have the same inductance value hence I believe that the transformer ratio is 1:1. I'm quite lost from this point because to my understanding, there is a formula that gives you the REACTIVE impedance of the transformer but this isn't too helpful for me as i want to know the nominal(?) impedance of a transformer (i.e. 600:600 ohm transformer), can anyone pls explain how I should proceed from this point?
Many thanks!

 

[ruffrecords]

Check this out:

Cheers

Ian

 

[Vintageaudioholic]

Check this out:

Cheers

Ian

Thanks a lot Ian!

 

[abbey road d enfer]

Hello, I need some help with transformer math...
I have these 2 transformers between my headphone amp's balanced output and unbalanced output, I would like to upgrade them but I don't know what transformers are suitable replacements in place of the originals so here we go...
Desoldering the transformers and measuring open circuit with no load, the transformers show that measuring at 1kHz, the Inductance is 185.2mH with a Q-factor of 1.18. Both ends have the same inductance value hence I believe that the transformer ratio is 1:1. I'm quite lost from this point because to my understanding, there is a formula that gives you the REACTIVE impedance of the transformer but this isn't too helpful for me as i want to know the nominal(?) impedance of a transformer (i.e. 600:600 ohm transformer), can anyone pls explain how I should proceed from this point?
Many thanks!

182mH suggests a nominal impedance of about 8 ohms.
I come to this conclusion by calculating the attenuation of about 0.5dB at 20 Hz.
The circuit is a fisrt order high-pass.
The 0.5dB point happens at a frequency 3 times the -3dB point.*
182mH and 8 ohms give a -3dB point of 7 Hz, hence a -0.5dB point at 20Hz.
Beware not every xfmr mfgr uses the -0.5dB point for their nominal.

*This is applicable to a 1st-order high-pass. For a 1st-order high-pass, the -0.5dB frequency is 1/3 the -3dB.

 

[ruffrecords]

182mH suggests a nominal impedance of about 8 ohms.
I come to this conclusion by calculating the attenuation of about 0.5dB at 20 Hz.

It might be higher than that. The OP measured the inductance at 1KHz. If the core material is iron and/or nickel based then its inductance will rise significantly at low frequencies.

Cheers

Ian

 

[CJ]

Why do you need to replace them?

A better way to figure out what you need is to post the amplifier schematic as transformers are not usually speced by their inductance.

 

[abbey road d enfer]

It might be higher than that. The OP measured the inductance at 1KHz. If the core material is iron and/or nickel based then its inductance will rise significantly at low frequencies.

Cheers

Ian

Correct, so they may actually be nominal for 32 ohms earpieces.

 

[abbey road d enfer]

Why do you need to replace them?

The lure of upgrading.

 

[Cqwet Dbdfte]

Transformer impedance specs alludes to typical signal levels, their mass and DC resistance to power levels of lowest frequency, I would test a transformer with a signal generator and and a scope to do some freq sweeps. REW could also do it.
Advertised impedance is normally the the properties of the attached circuits, not something directly measurable on the transformer.

 

[Vintageaudioholic]

Thanks a lot for the discussions and suggestions!
A little bit more information: My headphone amp is a Sony Ta-zh1es. There is a 33ohm resistor from signal out (balanced) into primary winding 0 and a 1M resistor from primary winding 1 to ground. The relevant circuit is on p.48.
The primary objective is to upgrade the transformer without messing with the circuit, although I know that transformers are reactive components, I'm pretty sure that there will be a difference in placing a 600:600 transformer vs a 50:50 transformer, etc.
I do know that sony quoted with a 32ohm load, it has 1200mw output on the balanced side and 300mw on the unbalanced side.

 

[Cqwet Dbdfte]

The 600 ohm number probably has its origin in telephone line characteristic impedance, where miles of cable was the norm. It also implies a power level, i.e. not lighting up stoves.
If you can find 50 ohm transformer it would, depending on its mass, be an RF device.
The transformer has two primary roles, isolation and voltage conversion. Equal ratio implies isolation, center tapping is for differential signalling used for noise cancelling, or push-pull driving.
No transformer is "perfect" and has design tradeoffs. Audio transformers may have a desirable, or not, sonic signature, but surrounding circuits may swamp these.
Your projects device is a complicated device with large BGA chips offering much opportunity to "color" the sound.
Various reviews of it are very divided. The transformer may play a very minor role. Proceed with caution.

 

[Vintageaudioholic]

The 600 ohm number probably has its origin in telephone line characteristic impedance, where miles of cable was the norm. It also implies a power level, i.e. not lighting up stoves.
If you can find 50 ohm transformer it would, depending on its mass, be an RF device.
The transformer has two primary roles, isolation and voltage conversion. Equal ratio implies isolation, center tapping is for differential signalling used for noise cancelling, or push-pull driving.
No transformer is "perfect" and has design tradeoffs. Audio transformers may have a desirable, or not, sonic signature, but surrounding circuits may swamp these.
Your projects device is a complicated device with large BGA chips offering much opportunity to "color" the sound.
Various reviews of it are very divided. The transformer may play a very minor role. Proceed with caution.

Colouring the sound is the primary reason why I'm attempting to change the transformers
If you compare the theoretical sound quality of balanced vs unbalanced in this amplifier, the unbalanced output would have an extra bottleneck where it sees the resistors and transformers. Not only that, it would be quite a waste to make both outputs sound the same because that would effectively lead to one output being the primary output. What I decided to do was to modify the unbalanced side to make it sound more old school. For now, I have modified the 33 and 1M ohm resistors to Allen Bradley resistors, it makes quite an effective cure against "Digititis" or when I would want to have a warmer, less resolving sound signature.
Ultimately, I do hope the transformers are a direct swap instead of a modification because that defeats the purpose of keeping the circuit original.

 

[ruffrecords]

Thanks a lot for the discussions and suggestions!
A little bit more information: My headphone amp is a Sony Ta-zh1es. There is a 33ohm resistor from signal out (balanced) into primary winding 0 and a 1M resistor from primary winding 1 to ground. The relevant circuit is on p.48.

This seems to be a relay switching board. The input signals are at the top right hand corner and come from the digital board and labeled HPOUT. But I cannot find what device on the digital board drives them.

Cheers

Ian

 

[Vintageaudioholic]

This seems to be a relay switching board. The input signals are at the top right hand corner and come from the digital board and labeled HPOUT. But I cannot find what device on the digital board drives them.

Cheers

Ian

Hi Ian, p.47 should be the output that you are looking for, it goes through the relays in p.48 and gets outputted

 

[abbey road d enfer]

It might be higher than that. The OP measured the inductance at 1KHz. If the core material is iron and/or nickel based then its inductance will rise significantly at low frequencies.

Cheers

Ian

[HIJACK]
After your comment, out of curiosity I tried to find a reference article about this, which anyone who has worked with audio transformers has noticed.
The only reference I found was in a Stack exchange thread, where the OP exposed his findings, hoping to have confirmation, but he didn't get satisfactory answers.
Apart from that, I could find no reference to permeability varying with frequency and level at low frequencies.
I believe it's a disseminated practical knowledge, but it's not been published.
Even Bill Whitlock's paper does not mention it.
There are many articles about permeability variations at high frequencies, because it's a major factor in smps efficiency, though. [\HIJACK]

 

[abbey road d enfer]

This seems to be a relay switching board. The input signals are at the top right hand corner and come from the digital board and labeled HPOUT. But I cannot find what device on the digital board drives them.

Cheers

Ian

It comes from a composite using an NJL8068 with a BUF634.
Interesting fact: the output goes through PhotoMOS SS relays, one in each leg.

 

[ruffrecords]

It comes from a composite using an NJL8068 with a BUF634.

Thanks for doing that. My old eyes were hurting looking at those schematics

I was not aware of the BUF634. I am surprised it is not used a lot in pro audio output stages.

Cheers

Ian

 

[amplexus]

Thanks for doing that. My old eyes were hurting looking at those schematics

I was not aware of the BUF634. I am surprised it is not used a lot in pro audio output stages.

Cheers

Ian

Probably because they’re $25+ per piece.

 

[CJ]

can you post a pic of the transformers?

 

[ruffrecords]

Probably because they’re $25+ per piece.

Ouch!

Cheers