Balanced connection for domestic audio: does it make sense?

[levimax]

This is also questionable and true only under certain conditions.

Transformer:
- CMR at low to middle frequencies is very good
- CMR at high frequencies is not good
- CMR at RF is poor
- distortion at low frequencies is poor
- frequency response slightly above audio band is prone to peaking and R-C is to be used to flatten it
- usable CMV at low frequencies is higher than with active circuit
- top DAC and PRE specs are always degraded by the transformer

While transformers have some issues what is better performing overall than a transformer? A "THAT" chip works well but it is not perfect either and is fairly expensive and not used on the vast majority of consumer balanced equipment.

For the points you laid out above (CMR, distortion, etc.) how does the "average" balanced solution used in consumer products compare to a good transformer or a THAT chip?

 

[chris719]

A transformer is probably better in harsh environments. I don't think it has a place in home audio, though.

 

[Harmonie]

The second reason is suppression (cancellation) of even harmonics. You get a bit better THD, especially with older circuits. This is the fact, not an opinion. Of course this does not work in resistor only pseudo-balanced circuit.

Please explain what a pseudo-balanced circuit is.
That would help me responding to this post:
https://www.audiosciencereview.com/...ic-audio-does-it-make-sense.20978/post-701220

 

[chris719]

Balanced impedances on hot and cold, but not necessarily a differential signal.

If LF common-mode rejection is all you care about, then pseudo-balanced is fine.

 

[restorer-john]

A transformer is probably better in harsh environments. I don't think it has a place in home audio, though.

Most of the absolute TOTL, no expense spared CD players and standalone D/A converters of days gone by used transformers for the 600R balanced outputs.

 

[pma]

For the points you laid out above (CMR, distortion, etc.) how does the "average" balanced solution used in consumer products compare to a good transformer

 

[chris719]

Most of the absolute TOTL, no expense spared CD players and standalone D/A converters of days gone by used transformers for the 600R balanced outputs.

While I have no doubt that is true, it's also been pointless in a CD player since the NE5532 (or other decent op-amps) existed. The Japanese brands have done some weird stuff. That doesn't make it optimal. There is just no real benefit in a domestic environment.

 

[levimax]

While I have no doubt that is true, it's also been pointless in a CD player since the NE5532 (or other decent op-amps) existed. The Japanese brands have done some weird stuff. That doesn't make it optimal. There is just no real benefit in a domestic environment.

Unless you are talking about dedicated IC's like a THAT chip discrete solid state balanced transmitters and receivers have orders of magnitude lower CMR than a transformer.

 

[chris719]

Unless you are talking about dedicated IC's like a THAT chip discrete solid state balanced transmitters and receivers have orders of magnitude lower CMR than a transformer.

Did I say otherwise? Yes, a level of CMRR that is generally unnecessary and not worth the other tradeoffs at home.

 

[levimax]

View attachment 117705

View attachment 117708

Sorry for being dense but not sure what I am looking at. Are these for a transformer, a THAT chip, or a discrete balanced receiver? Thanks

 

[Lambda]

not sure what I am looking

A schematic

Are these for a transformer

the thing in the red circle is.

a THAT chip

why dose a manufacturer matters?

 

[chris719]

Some of the THAT line receiver chips implement Whitlock’s bootstrap circuit and feature common mode rejection that is hard to beat with discrete op-amps.

Again, not important in a home environment. A standard balanced input is enough.

 

[levimax]

A schematic


the thing in the red circle is.


why dose a manufacturer matters?

I think we miscommunicated. I am not talking about Restorer John's schematic post I am talking about the two graphs posted by @pma.

 

[levimax]

Some of the THAT line receiver chips implement Whitlock’s bootstrap circuit and feature common mode rejection that is hard to beat with discrete op-amps.

Again, not important in a home environment. A standard balanced input is enough.

This kind of loops back to the OP question if balanced makes sense or not. If you are going to the trouble and expense of dealing with balanced connections is a "standard" balanced input with ~50 dB of CMRR "enough" when you can get ~90 dB CMMR with a transformer or dedicated IC?

 

[Lambda]

I think we miscommunicated. I am not talking about Restorer John's schematic post I am talking about the two graphs posted by @pma.

Sorry. my bad.

 

[pjug]

View attachment 117708

That's strange, the graph I see for JT-11-BMCF on the Jensen site looks similar but THD increases with signal level. To me this looks like good performance and an acceptable solution if balanced cables are not possible. [edit: I commented only because I don't understand the JT-11-HFMPC plot. Is it a dBu labeling error or real?]

 

[restorer-john]

While I have no doubt that is true, it's also been pointless in a CD player since the NE5532 (or other decent op-amps) existed. The Japanese brands have done some weird stuff. That doesn't make it optimal. There is just no real benefit in a domestic environment.

The NE-5532/4 existed long before and was used in those machines. Many of the TOTL machines of course, found use in professional environments like radio stations where 600R balanced lines were common.

 

[JSmith]

So you would spend >100$ extra on an extra devices that introduces an additional ~0.01% distortion and changes the frequency response.

Mate was just an example of a ground isolator... which gets rid of ground loops.

If balanced is the panacea, then why even make this?

https://www.jensen-transformers.com/product/pi-2xx/

Would it really change the frequency response much? There is also the point of any “degradation” in sound quality will be irrelevant compared to the annoying hum.

I think we're on a similar page, I'm not trying to argue with you, just providing counter points that explore the topic further and give food for thought to others.



JSmith

 

[chris719]

This kind of loops back to the OP question if balanced makes sense or not. If you are going to the trouble and expense of dealing with balanced connections is a "standard" balanced input with ~50 dB of CMRR "enough" when you can get ~90 dB CMMR with a transformer or dedicated IC?

There’s a world of difference between 0 and 60 dB of rejection, not that you can’t do better at low frequencies anyway. The other solutions come with compromises. A differential input is nearly free or required with a good ADC anyway. Who wants to use a big, expensive transformer or limit yourself to one IC from a niche manufacturer that also doesn’t perform as well as other options?

 

[chris719]

The NE-5532/4 existed long before and was used in those machines. Many of the TOTL machines of course, found use in professional environments like radio stations where 600R balanced lines were common.

Well, that may be true but this thread is about domestic audio.

I know you really like these old machines, but the world has moved on since the early 80s.

Transformers have a place in hostile environments and systems where you can’t fix or control interchassis currents and common mode voltages. They don’t really have a place in DACs or players anymore and the rest of the world seems to agree. They are rare even in pro interfaces now.