Unbalanced output to balanced input

[yuka42]

I know this topic has been probably beaten to death but I still have a question about unbalanced outputs to balanced inputs (ie: old 4 track R2R tape machine (RCA) to balanced patchbay (TRS)). I understand that transformer isolation is the recommended method to connect such devices, something like the ART T-8 would work. I also understand going balanced output to unbalanced input is fine to connect tip to tip and shield to shield as it essentially just makes an unbalanced connection. My question is why does the RANE interconnection document recommend tip to tip, and shield on the RCA to the ring on the TRS? Just interested to know how the balanced input circuit reacts and why you wouldn't connect the shield to shield and leave the ring disconnected. See #18 on the attached document and the "last resort" method.

The tape machine output would most likely be connected (through the patchbay) to a balanced line in or mic input on an A&H mixer (transformerless) or CAPI VP28 Mic/line preamps (transformer). Would this change the preferred way of wiring? Thanks for sharing your knowledge everyone.

Ryan

 

[ruffrecords]

I know this topic has been probably beaten to death but I still have a question about unbalanced outputs to balanced inputs (ie: old 4 track R2R tape machine (RCA) to balanced patchbay (TRS)).

Connecting unbalanced outputs to balanced inputs in not a problem. Just connect cold (ring) and sleeve together at the TRS end.

I understand that transformer isolation is the recommended method to connect such devices, something like the ART T-8 would work.

Transformers are not necessary unless you have hum loop problems

I also understand going balanced output to unbalanced input is fine to connect tip to tip and shield to shield as it essentially just makes an unbalanced connection.

Connecting a balanced output to an unbalanced input is a potential problem. Connecting tip to shield is a bad idea as it potentially shorts out the hot output driver of the balanced output. In fact the same is true of connecting the ring to shield. It all depends on the output drive circuit. This is where transformers may be necessary to convert the balanced output to an isolated unbalanced one

My question is why does the RANE interconnection document recommend tip to tip, and shield on the RCA to the ring on the TRS? Just interested to know how the balanced input circuit reacts and why you wouldn't connect the shield to shield and leave the ring disconnected. See #18 on the attached document and the "last resort" method.

Not sure if you are talking unbalanced to balanced or vice versa here.

The tape machine output would most likely be connected (through the patchbay) to a balanced line in or mic input on an A&H mixer (transformerless) or CAPI VP28 Mic/line preamps (transformer). Would this change the preferred way of wiring? Thanks for sharing your knowledge everyone.

Ryan

Not at all. Unbalanced to balanced id not a problem. Just connect ring to sleeve at the balanced end.

Cheers

Ian

 

[scott2000]

My question is why does the RANE interconnection document recommend tip to tip, and shield on the RCA to the ring on the TRS? Just interested to know how the balanced input circuit reacts and why you wouldn't connect the shield to shield and leave the ring disconnected. See #18 on the attached document and the "last resort" method.

I know what you're asking and this has been beaten to death. Just can't recall how to explain it but Rane has another article that touches on it. Obviously there are many others.....
https://www.ranecommercial.com/legacy/note151.html

 

[Newmarket]

My question is why does the RANE interconnection document recommend tip to tip, and shield on the RCA to the ring on the TRS? Just interested to know how the balanced input circuit reacts and why you wouldn't connect the shield to shield and leave the ring disconnected. See #18 on the attached document and the "last resort" method.

The signal "seen" by the balanced input is properly the voltage differential between the T and R of the TRS signals. If you leave the R of the TRS disconnected then you are not defining that voltage. Depending on the detail of the balanced input it may sort of work eg by biasing / pull down resistors setting a voltage at R. But it isn't the correct way to do it.
And it won't work acceptably if you're simply going into an input transformer.
Rane #18 does OEO screen connection to minimise "Hum Loops" but it's a bit of a sticking plaster.

 

[yuka42]

Thanks for that you guys. That's a fantastic article Scott and is exactly what I was looking for, some explanation as to "why" and from what the article states, there isn't just one answer. It comes down to signal ground and chassis ground, design choices, and how these are implemented. What I’m learning is there doesn’t seem to be an answer, or rather many possible answers depending on the situation. Maybe that’s why this problem comes up so often.

That makes sense about the balanced input “seeing” voltage difference between T and R and why it would need a connection on the return ring.

Thanks for the help everyone. I think I’ve got a plan. I’ll wire it as suggested by Ian connecting the ring and shield together at the TRS end. For that matter, could I just use an unbalanced patch cable in the patchbay for the tape to mixer and achieve the same result? It would tie ring to ground, correct?

 

[ruffrecords]

Thanks for the help everyone. I think I’ve got a plan. I’ll wire it as suggested by Ian connecting the ring and shield together at the TRS end. For that matter, could I just use an unbalanced patch cable in the patchbay for the tape to mixer and achieve the same result? It would tie ring to ground, correct?

Yes, a regular TS plug should work fine.

One last thing. A balanced signal exists only between the hot and the cold pins. The third pin is a screen and is not a signal conductor. One problem is that "modern" transformerless outputs tend to reference their outputs to the third pin which confuses newcomers who are falsely led to believe it is also a signal conductor and also that the hot and cold outputs are equal and opposite phase signals referenced to ground - they do not have to be. The "balance" in a balanced connection refers to impedance not voltages. The confusion is made worse by op amp based differential inputs used as balanced receivers because their inputs are also referenced to their local ground which is connected to the third pin.

Cheers

Ian

 

[yuka42]

Yeah I guess that’s what was confusing me with a balanced circuit. Reading more about differential amplifiers and common mode rejection is clearing up some confusion for me.

Here’s a great resource that you all probably know about but I just discovered.

Differential inputs
https://sound-au.com/articles/balanced-4.htm

CMRR measurements with inputs and outputs.
https://sound-au.com/articles/balanced-io.htm

 

[ruffrecords]

Yes, Rod Elliot is one smart cookie.

Cheers

Ian

 

[Newmarket]

Yes, a regular TS plug should work fine.

One last thing. A balanced signal exists only between the hot and the cold pins. The third pin is a screen and is not a signal conductor. One problem is that "modern" transformerless outputs tend to reference their outputs to the third pin which confuses newcomers who are falsely led to believe it is also a signal conductor and also that the hot and cold outputs are equal and opposite phase signals referenced to ground - they do not have to be. The "balance" in a balanced connection refers to impedance not voltages. The confusion is made worse by op amp based differential inputs used as balanced receivers because their inputs are also referenced to their local ground which is connected to the third pin.

Cheers

Ian

Indeed. All goes back to "Pin 1 Problem". If we could all start again it would be a lot easier. But so much kit out there (including where I am at home now) that doesn't comply properly with AES guidance. And then unbalanced kit...
Now throw in P48

 

[ruffrecords]

Indeed. All goes back to "Pin 1 Problem". If we could all start again it would be a lot easier. But so much kit out there (including where I am at home now) that doesn't comply properly with AES guidance. And then unbalanced kit...
Now throw in P48

Things were so much easier in the old days with floating transformers at both ends. These new fangled "electronic balancing" circuits are what caused the problem.

Cheers

Ian

 

[Newmarket]

Things were so much easier in the old days with floating transformers at both ends. These new fangled "electronic balancing" circuits are what caused the problem.

Cheers

Ian

I'll read that with the, I assume intended. "pinch of salt"
But yes. In terms of Isolation/CMRR transformers are hard / impossible to beat.
Although I have to admit that the Thor guy has a valid contribution wrt shipping the reference voltage separately from the screen. Don't tell him
But transformers are always (even the finest) are always compromised in terms of THD etc. and that is point for some I know.
Obvs the economics for mass market products generally excludes transformers.

 

[ruffrecords]

I'll read that with the, I assume intended. "pinch of salt"
But yes. In terms of Isolation/CMRR transformers are hard / impossible to beat.
Although I have to admit that the Thor guy has a valid contribution wrt shipping the reference voltage separately from the screen. Don't tell him
But transformers are always (even the finest) are always compromised in terms of THD etc. and that is point for some I know.
Obvs the economics for mass market products generally excludes transformers.

I have known for a long time that people still confuse component cost with cost of ownership. Yes, the component cost of transformers is higher than the equivalent electronic balancing circuit, but if your business depends on being able to reliably couple together a complex and diverse set of audio components from different manufacturers in almost any combination you can think of (i.e. a typical professional recording studio) you cannot afford to waste an hour of the studio and the client's time hunting down and fixing a hum problem. The cost in lost future business could be thousands. On that basis alone, transformers are worth every penny.

Cheers

Ian

 

[Newmarket]

I have known for a long time that people still confuse component cost with cost of ownership. Yes, the component cost of transformers is higher than the equivalent electronic balancing circuit, but if your business depends on being able to reliably couple together a complex and diverse set of audio components from different manufacturers in almost any combination you can think of (i.e. a typical professional recording studio) you cannot afford to waste an hour of the studio and the client's time hunting down and fixing a hum problem. The cost in lost future business could be thousands. On that basis alone, transformers are worth every penny.

Cheers

Ian

Sure. Transformers are a valid technical solution in many cases. Although I wasn't thinking exclusively about professional / commercial studio scenarios. Rather "prosumer"/hobbyist use. After all, the example case of a four track R2R with unbalanced outputs is a fairly uncommon requirement.

 

[ruffrecords]

Sure. Transformers are a valid technical solution in many cases. Although I wasn't thinking exclusively about professional / commercial studio scenarios. Rather "prosumer"/hobbyist use. After all, the example case of a four track R2R with unbalanced outputs is a fairly uncommon requirement.

I agree 100%. For prosumer/hobbyist use the only additional cost is your own time. However, given the sheer number of "How do I fix this hum problem" posts on audio forums, an awful lot of it is spent on this problem.

Cheers

Ian

 

[JohnRoberts]

I have known for a long time that people still confuse component cost with cost of ownership. Yes, the component cost of transformers is higher than the equivalent electronic balancing circuit, but if your business depends on being able to reliably couple together a complex and diverse set of audio components from different manufacturers in almost any combination you can think of (i.e. a typical professional recording studio) you cannot afford to waste an hour of the studio and the client's time hunting down and fixing a hum problem. The cost in lost future business could be thousands. On that basis alone, transformers are worth every penny.

Cheers

Ian

There are different time constraints associated with recording studio and live performance troubleshooting. The ASSumption is that a recording studio has the luxury of time to troubleshoot hum problems and provide a clean audio path. Live sound is more like emergency room (meatball) surgery where the show must go on right now. While at Peavey I included some transformer I/O on premium live sound SKUs (like higher end GEQs).

Coincidentally designing products for the sharp pencil (cheap) fixed install (background music) industry they also had a strong preference for transformer I/O because of a different time consideration. Labor to debug new system installs was a huge cost factor for them and they were willing to accept relaxed audio path performance to enjoy the more wiring mistake tolerant transformer I/Os. Another consideration was that transformer I/O could tolerate one leg of an audio circuit shorted to ground and still keep working, avoiding a profit killing service call after the installers had finished and long ago gone home.

JR

 

[Cqwet Dbdfte]

Low level signal transformers, other than 1:1, have the added benefit of noiseless amplification. Distortion added is not much, transducer non-linearities swamp most.
Differential signalling is for avoiding ground loops and signal integrity.

 

[JohnRoberts]

Low level signal transformers, other than 1:1, have the added benefit of noiseless amplification.

Transformers give us the ability to match low impedance sources with higher impedance amplification devices. Modern lower noise active devices offered an alternative.

JR

Distortion added is not much, transducer non-linearities swamp most.
Differential signalling is for avoiding ground loops and signal integrity.

 

[Newmarket]

Low level signal transformers, other than 1:1, have the added benefit of noiseless amplification. Distortion added is not much, transducer non-linearities swamp most.
Differential signalling is for avoiding ground loops and signal integrity.

It's not really noiseless though. In the sense that impedance is increased which is a factor in the noise level of an amplifier stage.

 

[ruffrecords]

It's not really noiseless though. In the sense that impedance is increased which is a factor in the noise level of an amplifier stage.

I think what he means is almost noiseless. Transformers do have a usefully low noise figure, and in many tube based mic pre they are the prime determinant in the overall noise performance. There is an excellent example in one of the Jensen papers.

Cheers

Ian