Balanced inputs vs rca

[sivadselim]

What about THIS thang?

[Rivendell61]

Craig,

I suspect some of the 'pro' oriented places have slightly better prices (Redco.com, Haveinc.com, etc) but Blue Jeans prices seem good to me too--and they have a much easier web site to navigate (for me anyway). Also the customer service is excellent. I ordered several sets of interconnects last summer. Ordered each late in afternoon--always had an email within two hours that cables were assembled and mailed Priority from Seattle--and had them in two days in Virginia. Really sold me.

Mark

[Marvel]

Btw, there are only two ways to connect between an unbalanced and

balanced source. The first is the fancy cable trick which we both

suggested to try first because its cheaper and easier. And second (the

"correct way") is to use a transformer (which we both mentioned too).

In both methods the ground signal is the same, but with the transformer

the phase is being reversed on pin 1.

Three ways, the third being an active circuit to balance the output of the preamp.

[Coytee]

The following is from the standpoint of understanding the concept as opposed to thinking it really applies to me.

I might be able to make a connector converter that is RCA on one side (to connect to my Peach) and balanced on the other side to connect to amp (I don't own this amp, I was just looking at it) and that scenario 'might' work sufficiently.

I could also take the RCA out of the Peach, connect it to a transformer that is specifically made for this purpose. The output of the transformer is a balanced connector that I could then take to the same amp.

There are no downsides of using balanced connectors, other than perhaps the challange of wiring them right and the cost of acquisition. There are some gains in balanced connectors, those gains MIGHT be perceptable and if the system is good enough as is, then those gains might not be perceptable?

If the distance between the Peach & the amp is long enough, then the balanced connectors will allow you to make that connection with zero, or minimal sound degradation, especially when compared to the sound the RCA connector might have under similar circumstances.

Most preamp manufacturers don't put balanced outputs on their amps for cost considerations. (I say that because I'm inferring that balanced is always better than RCA, but cost more)

[Marvel]

Hi Doc,

Actually the balanced noise rejection stuff is without reference to

signal. Don't need a signal on both wires. Signal symmetry will give

added benefits re headroom and crosstalk--but not noise rejection.

The 'Balanced' connection and noise rejection is a function of the two signal conductors having equal impedance.

Mark,

If pin 1 (ground) is tied to either pin 2 or 3, is is no longer

balanced, because it is the 180 degree phase difference that makes it

balanced (along with impedance) and offer noise rejection. If

they aren't out of phase, there is no benefit. With one of the signal

lines tied to ground it simply becomes an unbalanced. circuit.

Bruce

btw, that's how Cat5 network cable works. Each pair feeds a + and -

input. If it didn't, the fast rise time of the signal (square waves)

would shunt to ground due to the capacitance of the cable. Wouldn't get

very far through the cable before you had no signal at all. That is the

other benefit of balanced lines in audio if going long distances. High

impedance audio cables tend to lose the high end over long cable runs.

[JSharp]

Here are a couple of notes that are applicable to this conversation I think.

About balanced and unbalanced connections - http://www.rane.com/note110.html

About grounding and shielding - http://www.rane.com/note151.html

The common instrument balanced to unbalanced transformers are not

useful for a preamp-amp connection because in addition to performing

this conversion, they also perform an impedance transformation. They

are typically used to adapt something like a high impedance 50K ohm

microphone to a low impedance 300 balanced panel input. Unbalanced line

level inputs and outputs are commonly 600 ohm impedance so while we

theoretically need some impedance matching, it's not of the same

magnitude as the microphone case.

Balanced inputs and outputs don't have any practical advantages with

short cable runs between low impedance circuitry. The advantage appears

as the cable runs become longer and any noise sources become of greater

magnitude relative to the signals. In the case of balanced line that

noise will be a common mode signal and not seen at the differential

inputs of the devices.

I have a pro style eq with balanced inputs and outputs with the

connections tied single ended on both sides connected in the tape

monitor loop of my preamp. I can hear no difference in the noise floor

with the eq switched in or switched out using 3' cables between the

components. It's likely though that if I increased those cable lengths

that would no longer be true...

[sfogg]

Mark,

"Actually the balanced noise rejection stuff is without reference to signal. Don't need a signal on both wires."

Exactly correct. If the impedance is matched/balanced in both signal legs you can have signal on only one leg and still get the CMNR benefits of the balanced connection. The impedance needs to be matched/balanced so noise couples equally into both legs. Note: the noise will be IN PHASE in both legs.

For those that doubt it consider what happens on the balanced input on the device down the chain.

Basically the device with the balanced input subtracts the signal on one leg from the signal on the other leg.

If noise coupled into both legs equally (which would make it in phase) when you subtract one leg from the other you end up canceling out the noise because you subtract the signal on one leg from the signal on the other leg and since they were both in phase you end up with nothing.

It makes no difference if signal is being carried on both legs WRTs CMNR.

When you pass signal on both legs with one having inverted phase when one leg is subtracted from the other the signal level increases. Minus a minus (subtracting something in inverted phase) equals a plus.

Shawn

[Rivendell61]

Thanks Shawn,

I said I'd post some links last night--but forgot.... Not sure if anyone is interested but here are two which address these questions:
http://www.aes.org/sections/pnw/pnwrecaps/2005/whitlock/whitlock_pnw05.pdf
That is a PowerPoint article by Whitlock: Real-World Balanced Interfaces and Other-World Myths. His primary paper on it was in 1995--this is a sort of pop/sales version. Look at pages 38-40 (each page is like a presentation slide) headed "Myth: Balanced=Signal Symmetry. There are good quotes from Henry Ott (author of "Noise Reduction Techniques in Electronic Systems") and the IEC Standard. And page 82 shows a good diagram of the wiring solution I was suggesting for unbalanced to balanced. And Whitlock's surmise of 30dB rejection for a 'typical' configuration.

Also.... http://www.svconline.com/mag/avinstall_mixandmatch_interfaces/
A more concise article on connecting unbalanced and balanced gear. A good general read on the issues. Its the place where I grabbed that quote about the cable solution being "The correct way". See the paragraph about half way down, starting with "Figure 1A". Whitlock sells transformers so I thought it significant. He also says some interesting stuff about transformers in the following sections.

Hope those are of interest.

Mark

[DrWho]

Mark, "Actually the balanced noise rejection stuff is without reference to signal. Don't need a signal on both wires." Exactly correct. If the impedance is matched/balanced in both signal legs you can have signal on only one leg and still get the CMNR benefits of the balanced connection. The impedance needs to be matched/balanced so noise couples equally into both legs. Note: the noise will be IN PHASE in both legs. For those that doubt it consider what happens on the balanced input on the device down the chain. Basically the device with the balanced input subtracts the signal on one leg from the signal on the other leg. If noise coupled into both legs equally (which would make it in phase) when you subtract one leg from the other you end up canceling out the noise because you subtract the signal on one leg from the signal on the other leg and since they were both in phase you end up with nothing. It makes no difference if signal is being carried on both legs WRTs CMNR. When you pass signal on both legs with one having inverted phase when one leg is subtracted from the other the signal level increases. Minus a minus (subtracting something in inverted phase) equals a plus. Shawn

I follow this all the way except for one important point...that summed signal 'without the noise' is still in reference to the ground cable which will also have picked up noise along the signal path. If you claim this ground is forced to '0V', then you the signal on pin1 is also forced to 0V. If the ground isn't "shorted" then you still get noise from the ground reference. Man that's hard to say...I hope it made sense....

[sfogg]

I was talking about that with regards to a balanced connection on both

ends. Not a connection with an RCA on the output end and balanced input

on the other end. In a situation like that if you ground one of the

signal leads (or leave it unconnected at the RCA side) then obviously

the connection isn't balanced/impedance matched between both signal

legs.

An actual balanced connection (XLR on both ends) does not need to have

the signal inverted in phase on one leg to gain the CMNR features of a

XLR connection. Think about it... do you loose the CMNR between tracks

on a CD when there is no signal on either leg?

Shawn

[DrWho]

What does CMNR stand for?

And are you talking about a CD player running balanced outputs?

[sfogg]

Common mode noise rejection/reduction.

I'm talking about the interconnection between any two devices that use an XLR connection between them.

Source to pre-amp... or pre-amp to power amp...etc...etc...

Shawn

[DrWho]

sounds like I've got some more reading to do []

[sfogg]

Don't feel bad..... it is an *extremely* common mistake to think that

without the phase inverted signal on one of the leads that the CMNR

benefits are lost.

That isn't accurate though. As long as the connection is balanced,

meaning the impedance on each signal leg is the same, you get the CMNR

even with signal on one leg along as long as the receiving end does the

subtracting of the second leg from the first.

Shawn

[DrWho]

Ok, it seems to me that a balanced connection is no matter what

implementing the phase reversal...the articles I was reading were just

pointing out that you don't have to reverse the phase of the actual

audio signal or even send it down both lines, but I always thought that

was intuitive...the "same impedance" for both lines just guarantees

that the noise pickup is identical (which means perfect cancellation).

I think some of the charts were misleading because they don't show the

ground wire going between sources on the balanced connections

lol, writing at the same time []

[Rivendell61]

I believe the abbreviation is more frequently given as CMRR (Common Mode Rejection Ratio).

And....in case it is not clear from Shawn's mentioning any connection using XLRs having CMRR....also the specified RCA/XLR connection has it (which is the whole point)--just less, i.e., maybe 30bdB rejection--but more than zero (adapter).

Mark

[DrWho]

And....in case it is not clear

from Shawn's mentioning any connection using XLRs

having CMRR....also the specified RCA/XLR connection has it (which is

the whole point)--just less, i.e., maybe 30bdB rejection--but more than

zero (adapter).

Ya, I still don't buy that though...Imma go give some profs a visit and see what they say []

[sivadselim]

So, would cables like THESE be useful or not?

[Duke Spinner]

And....in case it is not clear

from Shawn's mentioning any connection using XLRs

having CMRR....also the specified RCA/XLR connection has it (which is

the whole point)--just less, i.e., maybe 30bdB rejection--but more than

zero (adapter).

Ya, I still don't buy that though...Imma go give some profs a visit and see what they say []

AAAAHHH ... to be young and curious.........

instead of old and jaded .....

[JSharp]

So, would cables like

THESE

be useful or not?

Useful, but much modern equipment uses jacks that allow 1/4" 2

conductor or 3 conductor plugs to be used. All the discussion in the

thread to be considered when you start using 2 conductor plugs and

cables and connecting balanced and unbalanced lines...