What are the benefits to things that have balanced differential inputs?

I got to thinking more about this today after a discussion with some others.


Most car audio has RCA outs that are single ended outputs. Quite a few devices are having balanced differential input stages via RCA's.

Example:
https://jlaudio.zendesk.com/hc/en-us...alanced-Inputs


I can see the benefits if using high level inputs because those should be balanced (signal on both wires and nothing grounded), but other issues arise with high level as we all know. But rca's are usually grounded at the source which would defeat the whole balanced and/or differential input.


I also do not see a bunch of benefits for balanced RCA's (2 wire things) as then you lose some other noise rejection properties from the shield normally being grounded in "normal" rcas. Hence the 3 wire/connections (XLR, TRS, ETC) normally for balanced with a dedicated ground/shield and dedicated signal wires.


Anyone help me see why this is a "thing" without sources being balanced? What am I missing?






Note for those that do not know: balanced and differential are usually combined but are not the same thing. Many people use these words interchangeably.

Balanced means both wires have the same impedance to ground, aka not grounded anywhere and both carry a signal.

Differential is where the source signal has one wire with an inverted polarity signal compared to the other wire. Upon reaching the amp, the amp switches this signal back and this process eliminates most noise picked up on the path.

I'm curious too.

This is the best concise explanation I've seen:

https://youtu.be/QOagVDZLQnA

Covers single-ended, differential - kinda inherently covers why balanced and differential go hand in hand also.

Right, but doesn't the differential part also rely on the source being able to send out a differential signal? I always thought that both the source and the receiver had to be balanced to be balanced, and both had to be differential to be differential. If source is single ended, then that eliminates balanced right there so you'd lose all benefits of a balanced differential input.

Or am I missing something.....


Also I was given these documents to look over by another friend. I haven't gone completely through them yet, but so far it seems like what I said above is confirmed in these documents:

https://centralindianaaes.files.word...notes-v1-0.pdf


https://www.jensen-transformers.com/...ic-seminar.pdf

Quote:

---

Originally Posted by Jdunk54nl

Right, but doesn't the differential part also rely on the source being able to send out a differential signal? I always thought that both the source and the receiver had to be balanced to be balanced, and both had to be differential to be differential. If source is single ended, then that eliminates balanced right there so you'd lose all benefits of a balanced differential input.

Or am I missing something.....

---

You can mix differential and single-ended signal inputs and outputs, sometimes.

Differential inputs can accept differential signals and also single-ended signals.
Single-ended inputs can only accept single-ended signals.

One possible benefit of a differential input is it will be compatible with all consumer car audio gear which is typically single-ended, and also compatible with all professional and OEM gear like factory radios which are differential these days.

I'm cruising through the Whitlock paper from Jensen and it's full of great info. Lots to digest there.

So then one benefit would be compatibility then. That makes sense.

So im going to assume here that you are using the correct RCA for the type of rca input circuitry.

Single ended - coaxial

differential input - twitsted pair


On an amp that has single ended inputs, the shielding of the rca cable protects the middle pin(the actual signal) from getting noise. However if there is too much noise or simply a poor cable, noise will get in and you will get noise in your system.

On an amp that has differential inputs, whatever amount of noise gets "appended" to both conductors of the rca. There is no shield on the RCAs. And due to the way that differential input circuitry works, it looks at whatever is the same on both conductors. And since both conductors have the same noise, it gets removed.

Differential inputs are better

I only know of a single device that has *DIFFERENTIAL OUTPUTS*, and that would be a ford after you use for scan to convert the speaker wires into low level preouts.

Quote:

---

Originally Posted by test13371997

So im going to assume here that you are using the correct RCA for the type of rca input circuitry.

Single ended - coaxial

differential input - twitsted pair


On an amp that has single ended inputs, the shielding of the rca cable protects the middle pin(the actual signal) from getting noise. However if there is too much noise or simply a poor cable, noise will get in and you will get noise in your system.

On an amp that has differential inputs, whatever amount of noise gets "appended" to both conductors of the rca. There is no shield on the RCAs. And due to the way that differential input circuitry works, it looks at whatever is the same on both conductors. And since both conductors have the same noise, it gets removed.

Differential inputs are better

I only know of a single device that has *DIFFERENTIAL OUTPUTS*, and that would be a ford after you use for scan to convert the speaker wires into low level preouts.

---

Right. But does all of that differential input stuff still work with single ended outputs, which is the case for 99% of the audio sources I could find without getting into the pro level gear.

If the source you are using is a single ended single, that means there is not equal impedance to ground that is a must to be a balanced signal. It also isn't going to be differential, like you stated above 99% of the time.

So we are feeding these amps signals, from my understanding (after reading the Whitlock papers above) there is 0db of noise rejection.

The top picture is basically what would happened with a rca output to an amps rca input. Which is 0db noise rejection. We have tied the shield to the signal and ground on both ends. I don't think twisted pair would do anything if you can't separate the shield and lo signal side of the signal at the input end. I could be wrong about that though.

Page 29 from here: https://www.jensen-transformers.com/...ic-seminar.pdf

Attachment 15778

I also want to re-iterate, I am just going based on what I can read on these things from respected sources. My knowledge is very limited here, which is why I posed the question.

I guess I am unsure of how many sources actually do not tie their shield or outer shell of the rca plug to ground. This could be an important thing as that could make them balanced outputs and would 100% fix my understanding.

Quote:

---

Originally Posted by Jdunk54nl

Right. But does all of that differential input stuff still work with single ended outputs, which is the case for 99% of the audio sources I could find without getting into the pro level gear.

If the source you are using is a single ended single, that means there is not equal impedance to ground that is a must to be a balanced signal. It also isn't going to be differential, like you stated above 99% of the time.

So we are feeding these amps signals, from my understanding (after reading the Whitlock papers above) there is 0db of noise rejection.

The top picture is basically what would happened with a rca output to an amps rca input. Which is 0db noise rejection. We have tied the shield to the signal and ground on both ends. I don't think twisted pair would do anything if you can't separate the shield and lo signal side of the signal at the input end. I could be wrong about that though.

Page 29 from here: https://www.jensen-transformers.com/...ic-seminar.pdf

Attachment 15778

---

Yes, differential input still works with single ended outputs. Mess with desmos.com https://imgur.com/a/pTpNdr2. a and b is the "ideal" or exactly what is coming out of the HU. Signal = ground. Since its a differential input, you use unshielded twisted pairs. That means that noise get radiated onto both conductors. So c and d is a and b plus the imaginary noise I came up with. When you do a-b and c-d, you get the same thing. You mention "balanced", but there are no balanced cables with RCAs in car audio. For a cable to be balanced, it needs to have 3 conductors. Which rcas dont have. Those pictures are talking about XLR adaptors and balanced stuff, which isn't a thing in car audio. I dont have anything to say about those pics

Quote:

---

Originally Posted by test13371997

Yes, differential input still works with single ended outputs. Mess with desmos.com https://imgur.com/a/pTpNdr2. a and b is the "ideal" or exactly what is coming out of the HU. Signal = ground. Since its a differential input, you use unshielded twisted pairs. That means that noise get radiated onto both conductors. So c and d is a and b plus the imaginary noise I came up with. When you do a-b and c-d, you get the same thing. You mention "balanced", but there are no balanced cables with RCAs in car audio. For a cable to be balanced, it needs to have 3 conductors. Which rcas dont have. Those pictures are talking about XLR adaptors and balanced stuff, which isn't a thing in car audio. I dont have anything to say about those pics

---

Balanced just means the signal wires have equal impedance to ground (Pin 2 and Pin 3 of a XLR). You can have balanced without 3 conductors as the third conductor (pin 1) of an XLR/TRS/etc. plug is just a ground, this isn't needed to be balanced signal as it has nothing to do with the actual signal.

I am not 100% sure what your picture is doing or trying to show me.

There is also this from that paper I have linked above, which basically says differential DOES NOT help with noise rejection, only balanced does, which requires both the source and the receiver to be balanced in order to have the required equal impedance to grounds on the signal wires.

This is page 22 so you can read the whole excerpt as it continues past this picture.

Attachment 15785





And this on the bottom of page 16.



Attachment 15784

Did you see page 29? Some interesting claims about noise rejection when connecting single-ended outputs to differential inputs, even if the output is not balanced.

Right, that seems like specific to using a 2 conductor wire and shield (for 3 pin connector on the receiver end). Does the wording mean it gives 20-30db by being balanced, and an additional 30db by being balanced and differential for 50-60db noise rejection?
https://www.caraudiojunkies.com/atta...8&d=1621611272

Quote:

---

Originally Posted by Jdunk54nl

Right, that seems like specific to using a 2 conductor wire and shield (for 3 pin connector on the receiver end). Does the wording mean it gives 20-30db by being balanced, and an additional 30db by being balanced and differential for 50-60db noise rejection?
https://www.caraudiojunkies.com/atta...8&d=1621611272

---

That is impressive.
Does that really mean "3 connector" though?
Or is that the two RCA - plus the shield simply grounded, like some RCAs offer with the center conductor?
Not that that isn't "3 connections"...

Sent from my LM-G820 using Tapatalk

Quote:

---

Originally Posted by geolemon

That is impressive.
Does that really mean "3 connector" though?
Or is that the two RCA - plus the shield simply grounded, like some RCAs offer with the center conductor?
Not that that isn't "3 connections"...

Sent from my LM-G820 using Tapatalk

---

It seems, IF my understanding is correct, twisted pair conductor needed with a shield and a way to separate the lo signal and ground noise at the receiving end (one to pin of RCA to pine 2 of XLR, sleeve of RCA to pin 3, and shield of cable to sleeve or RCA and pin 1 of xlr). In my mind that means you must have three separate connection points (Pin 1, 2, and 3) on the receiving end.

On XLR, yeah - 3 pins. it's kind of the same - but hard to compare because those are balanced signals.

If you ignore that though, yeah - "shield" is just grounded, conductor A is the signal, and conductor B is the signal inverted.

We unfortunately don't usually have that with single-ended (unbalanced) sources and amps.

I've been watching some demos on Star quad cable but so far all have been that "balanced" scenario demonstrating noise rejecting, and it makes sense in that context...
The two white conductors are the signal, the two blue are the inverted signal - you get a magnetically null sum, and noise is easy to reject.

I don't think it's possible to take a single ended source and send it down one strand, and send the inverse down the other strand even to get passive noise rejection (without a balanced receiver inverting and subtracting the noise), but I'm going to think on that a bit.

Basically trying to figure out of it is even an advantage to use star quad cable at all [EDIT: since I'm in the planning of making my own RCA cables with single-ended everything].

Quote:

---

Originally Posted by geolemon

Basically trying to figure out of it is even an advantage to use star quad cable at all [EDIT: since I'm in the planning of making my own RCA cables with single-ended everything].

---

Isn't the advantage of starquad geometry the cancellation of magnetic interference? I'm pretty sure it's the same as a normal twisted pair but better magnetic interference rejection.

Quote:

---

Originally Posted by Justin Zazzi

Isn't the advantage of starquad geometry the cancellation of magnetic interference? I'm pretty sure it's the same as a normal twisted pair but better magnetic interference rejection.

---

Yes exactly - that makes it downright great in a balanced application, like XLR for mics and instruments - two conductors carry the positive signal, two conductors carry the inverted signal. The shield provides passive blocking, and the balanced differential input sums away any induced noise that did make it through the shield.

But in all the car stereo gear I've had, everything has been single-ended, not differential.

I just read a thread on a high-end home DIY forum criticizing DIY RCA cable builders who use star quad for unbalanced RCA because:

1) the shield in coax is a 75 ohm rating, which apparently has a stabilizing effect (I want to find more on this). On star quad, it's not rated. I would have assumed "lower is better", so I'm kind of curious here - but bottom line is it was touted as an advantage for coax.

2) The shield in coax is braided. It's also braided in most star quad that I've seen, but did see someone complaining about how tough it is to unbraid coax, and how much easier it was to unwind the simply-spiralled shield on their quad cable brand of choice. It does make sense to me that the spiral shield could create inductance where the braid would inherently avoid it.

3) They measured nearly 4x the capacitance using two strands of quad for positive pin, two strands plus shield for outer. They improved it by using only one of the two conductors "but still measured more capacitance". It was still only 180pF, so I'm not sure this could amount to any frequency limiter.

I saw another DIY thread where they suggested using all 4 conductors to center pin - I would believe that to increase capacitance further but...
I found an RC filter calculator that allows pF and will calculate frequency with R and C supplied and so I figured "let's see what 0.1 ohm, 75 ohm, and 150 ohm give us with a 180pF capacitor?"

[Drum roll here]
0.1 ohm and 180pF: 8846426426043Hz
75 ohm and 180pF: 11795234 Hz
150 ohm and 180pF: 5897617 Hz

Before you say "that's way above hearing range" (it is) - this is a high pass filter, so in theory that means in all three cases there would be a sloping 6dB/octave attenuation starting way up there - but I couldn't say that a filter whose elbow was THAT far above the used frequency range would/could still provide a 6dB/octave slope. That's a tiny amount of capacitance, so there's only so much available energy to do anything. I expect it's trivial either way - every wire has some capacitance, and of course that doesn't mean "slope".

I have a feeling BOTH arguments are true:
1) "Technically, coax is the superior choice for unbalanced RCA applications."
2) "The differences are so tiny between coax and quad that there's no possible audible advantage either way."

Sent from my LM-G820 using Tapatalk

So thinking about the picture below more (the correct one).
Wouldn't this ultimately destroy any balance properties? You have the shield of the cable tied to ground on the receiver end and tied together on the RCA sleeve, so if you measured pin 3's impedance to ground, it should be way less than pin 2's, so no more balance by definition.....

I am still so confused on this stuff.....The more I think I understand, the more I find errors in my thinking and potentially of others thinking....


Page 29 from here: https://www.jensen-transformers.com/...ic-seminar.pdf

Attachment 15778

Just thinking out loud here, and trying to do that scientifically.
Here are the two parts that actually differ, to help think through it:
Attachment 15806
Attachment 15807
The balanced receiver end, is same for both:
Attachment 15808

In the coax cable, the signal and shield are separate at the source, and therefore from the perspective of the balanced receiver, "lo" and "sh" have to be soldered. Also, on the source end "shield" is often at the ground potential. And on the receiver end, pin 1 is grounded - therefore pin 2 is also.

On the shielded twisted pair (or quad), they show the shield and one conductor soldered at the source. Then at the receiver, you have 3 distinct pins [with only pin 3 grounded, in the receiver].
Overall, it's true pin 2 is also grounded (because they are connected, on the source side), however I think it has to factor in the length of the cable and the assumption that the noise is entering somewhere during that cable length.

That could result in the noise being on pin 2 but not the double-grounded shield, and therefore the balanced circuit able to remove it - but since we're not dealing with a balanced differential driver or inverted signal on pin 2, I'm not yet understanding the removal or rejection mechanism fully.

And actually - maybe I'm thinking of that wrong anyway. There's only one op amp in the receiver and I think actual balanced receivers require 3, with two of them doing that noise removal by inverting and combining...
But regardless I'm not seeing how the noise is attenuated. Has to do with pin 2 not being grounded...

I am glad I am not the only one struggling with this.

I did start a post on ASR as well, some more people in the "industry" live over there so thought it might help:

https://www.audiosciencereview.com/f...rential.23746/

I was in electrical engineering for three years but never finished...

Had a too-good-to-waste opportunity with some of the best data warehousing and business intelligence pioneers in the late '90s and that changed my whooole professional and education direction.

So - back to this stuff just being a hobby!

25 years later... My EE teachers would be disappointed with me. I still feel like I should know this. [emoji38]

Sent from my LM-G820 using Tapatalk

Quote:

---

Originally Posted by geolemon

I was in electrical engineering for three years but never finished...

Had a too-good-to-waste opportunity with some of the best data warehousing and business intelligence pioneers in the late '90s and that changed my whooole professional and education direction.

So - back to this stuff just being a hobby!

25 years later... My EE teachers would be disappointed with me. I still feel like I should know this. [emoji38]

Sent from my LM-G820 using Tapatalk

---

Seriously thinking about going to school for EE. Thoughts?

Sent from my XT1710-02 using Tapatalk

What career do you want from an EE degree?

I think it's a great career, with a few caveats:
1) I found it more difficult than I anticipated to get started - I was shocked that the majority of the class already knew quite a bit... Either their fathers were electoral engineers, or they'd been raised building electronics like Heathkits. It was impressive - but the teacher taught to their level. I thrive with my feet in the fire (it's why I was recruited by the BI IT consultant team) but it was hard.

2) There's a range of careers, and what you land in can depend on your school. There are electronics repair people - those jobs don't pay too well but lots of people find that fun. I wish I even qualified to do that - I have several old amps I'd fix!
Actual electrical engineers can work on anything from consumer product development projects, to aerospace. A couple decades back I helped co-found an SMT contract manufacturing business, and slid into the role of manufacturing engineer, I programmed and set up and ran the solder printer, the pick and place machines, and the oven (ours had 7 zones, all programmable) and we targeted the aerospace and medical contractors - big money and difficult standards. We had to automatically (that one was mine) manually and X-ray inspect all the boards to guarantee if they were dosing anesthesia or up in a satellite, they'd work. Very different than consumer electronics.
Other electrical engineers work on large scale power distribution - buildings, factories, power companies - I feel like the sky is the limit.

Sent from my LM-G820 using Tapatalk

Quote:

---

Originally Posted by Jscoyne2

Seriously thinking about going to school for EE. Thoughts?

Sent from my XT1710-02 using Tapatalk

---

Good job opportunities, I would do it. Just be prepared for how difficult it is. You will learn theory, start simulating circuits in software, and test on the breadboard in the lab. That is just the start. It is a very unique way of thinking. The best EEs have a near instant grasp when learning new things coupled with the creative mind of an artist. The best of both worlds.

Look around you at everything electronic and the career options are many. I know an EE that works in product testing. They hand him a prototype and he does every stupid thing the end user might do to try to break it. Also knew a guy in VLSI circuit design.

In college, we worked a bit in the lab on implanted sensors in the human body. And there is also the patent law side of it if you want.

Im just Physics with some EE. Ive worked with LVDS, digital, and by presenting the differential signal at each side, the noise is subtracted out. That is the obvious theory. Beyond that, youd need someone who works specifically in that area to explain it. EE is specialized like everything else and is a pretty broad field in and of itself.

there aren't any balanced line output sources anyway, so the start of the signal chain will be unbalanced (signal and shield on an RCA cable) and would need to have a conversion to become "signal, inverted signal and ground reference". that is where the 3 signals would come from, so that the reversion process of combining the "signal (+) and the "inverted signal" (-) would be combined after the "-" signal is inverted again to add constructively to the "+" signal, and the "noise " that is present in both signals, would have one half of it inverted, and then that would combine destructively with the + signal and be cancelled out, reducing the noise induced over the length of the signal wire. but we do not have that in car audio, so most source outputs are just signal and shield. that means if the shield is used improperly or does not shield enough, the noise and the signal are both amplified.
kenwood had a cd player (KDC-X911) that had fully balanced outputs using RCA outs, where one RCA was the + and one RCA was the -. worked AMAZINGLY well, sadly, never caught on.