Installing BNC connectors on digital gear...

[klettermann]

to replace crappy RCA connectors - is there a real downside? ChatGBT seems to think so. The reason is that supposedly the transmitter/reciever design for RCA/spdif is optimized around that connector and a BNC connector would disrupt that. Yet a 75ohm connector seems like the whole chain would be less disrupted. And then there are 75ohm rca-to-bnc cables. That seems like compromising a 75ohm cable just to make it work with an RCA socket. And then there are also special 75ohm RCA conectors! I'm a hobbyist, not an engineer. To be clear, I have little expectation of some sonic breakthrough. I'm want to minimze bad stuff, noise, etc to the extent that I can. Connemts and suggestions very welcome! Thanks and cheers,

 

[wwenze]

Does it work? If yes it works, why bother. If no it doesn't work, manufacturers already use BNC in places where RCA doesn't work.

Since you mentioned you're not an engineer, here's a nice guide about how to interpret text on technical topics. We are moving towards bottom right.

 

[tineared]

It sounds like a solution in search of a problem
What is so crappy about RCA connectors?

 

[NTK]

There is no problem at all interchanging RCA with BNC connectors for digital audio signals.
From Wikipedia (S/PDIF):

 

[klettermann]

Thank you, but I was hoping to fall farther down this rabbit hole. Though I'm not an RF engineer or RF circuit designer I'm not the average Best Buy customer either. So here goes.

We all know that commercial/pro gear uses properly BNC terminated 75ohm cables. Why? Because spdif was designed that way and mismatch causes reflections, leading to power loss, signal distortion (ghosting, errors), standing waves, etc. Consumer stuff uses RCAs and, with very few exceptions, they are NOT 75ohms. Why? Cables are generally short enough that depite their shortcomings, the audible effects are generally negligible to none and the connectors cost a few cents, not a few $$. OTOH, plenty of high end digital stuff DOES use BNCs., same as commercial gear. Why? Fewer cost constraints and spdif was actually designed for 75ohm transmission in the first place. So, Heathkit-hobbyist that I am, I replace spdif RCA sockets with 75ohm BNC's if it's practical to do so.

Now, back to the underlying question. In exploring this topic with ChatGBT, it suggested that it's common in consumer product spdif I/O circuitry, i.e., what's behind the RCA socket, to be DEoptimized. Says Google: "most standard RCA plugs and sockets have an actual impedance of around 20-30 ohms due to their physical design. However, the gear is designed to work with this non-ideal termination, and for typical short runs..." ChatgGBT's implication was that replacing an RCA socket with a BNC would actually make matters worse because such spdif design. I'm trying to optimize my setup just on principle, whether audible or not, and I like messing with stuff. Happy Holidays!

 

[-Matt-]

Not sure why you would want to, but these things exist...

 

[voodooless]

is there a real downside?

- It's work
- It costs money
- It will lower resale value
- It may void warantly
- You need new cables
- It has no audible benefit
- You may damage your equipment in the process

 

[RandomEar]

It's a digital signal. As long as you don't hear dropouts or pops/clicks during playback, it's working perfectly. There's nothing to optimize and a lot to make worse.

If you worry about ground loops or such, just switch to Toslink.

 

[MCH]

With the nowadays always small and cramped devices, you will have to be lucky to have a big enough space in the right spot in the back of the device to securely mount a new panel mount connector (?).
I also like the locking mechanism of those connectors, it is very satisfying to plug and unplug them (I like LEMOs too) but I doubt it is worth it unless you are experienced with these mods or have a way to make a new back pannel altogether.

 

[JeremyFife]

Stop using AI.

It has its place but the training material for hifi is loaded with subjective opinion and is often just wrong.

If you must, then dig into the early responses, demand sources and query the impacts in terms of engineering and human audibility. It usually retracts its initial answers

 

[sergeauckland]

to replace crappy RCA connectors - is there a real downside? ChatGBT seems to think so. The reason is that supposedly the transmitter/reciever design for RCA/spdif is optimized around that connector and a BNC connector would disrupt that. Yet a 75ohm connector seems like the whole chain would be less disrupted. And then there are 75ohm rca-to-bnc cables. That seems like compromising a 75ohm cable just to make it work with an RCA socket. And then there are also special 75ohm RCA conectors! I'm a hobbyist, not an engineer. To be clear, I have little expectation of some sonic breakthrough. I'm want to minimze bad stuff, noise, etc to the extent that I can. Connemts and suggestions very welcome! Thanks and cheers,

Yet another load of rubbish from Chat GPT!!! Why does anyone ask it anything factual?

Yes, there are downsides to changing, all of which Voodooless has very eloquently expressed above. However, none of those are to do with 'optimisation' for the RCA connector. S-PDIF is incredibly rugged, the change from an RCA connector with its uncertain impedance to a correctly 75 ohms BNC will make 10% of diddly-squat difference.

If it works, it works, you can't get better than it works.

S.

 

[klettermann]

- It's work
- It costs money
- It will lower resale value
- It may void warantly
- You need new cables
- It has no audible benefit
- You may damage your equipment in the process

Wow, so much scepticism! @voodooless Some of up. Of course! However....
1. I greatly enjoy this kind of work.
2. I spend more money on a beer.
3. Of no concern to me.
4. Also of no concern.
5. I've been at this for a while. My cable collection is substantial.
6. Now THIS may be true.
7. Possible but unlikely. But I'm experienced and have a good sense of the risk. No pushing the envelope.

Moreover, female surface mount BNC connectors come in about a zillion types and pinouts. You guys are convincing me to go for it. On top of that, I got some new stuff that needs to be accommodated. If I don't I'm going to have to use a toslink cable and I can't abide that. I'll let you know how it goes.

 

[mcdn]

The type of physical connector is unrelated to the terminating impedance, which is set by a simple 75ohm resistor between the signal and ground pins. A 75ohm cable on the other hand has a characteristic impedance of 75ohms at frequencies of interest to the application, this eliminates reflections : https://en.wikipedia.org/wiki/Impedance_matching

The RCA input in your system should already have a load resistor, so for fun you can indeed use a 75ohm cable but it can be RCA terminated: https://benchmarkmedia.com/products/benchmark-rca-to-rca-coaxial-cable

 

[somebodyelse]

The type of physical connector is unrelated to the terminating impedance, which is set by a simple 75ohm resistor between the signal and ground pins. A 75ohm cable on the other hand has a characteristic impedance of 75ohms at frequencies of interest to the application, this eliminates reflections : https://en.wikipedia.org/wiki/Impedance_matching

Connectors have a characteristic impedance too, and if it's mismatched with that of the cable or the terminating impedance you can also get reflections. BNC connectors come in 50R and 75R varieties, and if you use the wrong one in an application where it matters you'll break things - mixing things up between video and 10Base2 networking for example. RCA doesn't have a standardised characteristic impedance, but its dimensions impose some limits on what impedance they can be made in - and 75R is outside that range. Luckily for short cables in this application it doesn't matter much.

 

[xanalog]

Unlikely that you will be operating at 24bit/192kHz data rates on the S/PDIF coax… however, due to biphase mark code (BMC) encoding, clock recovery/extraction is ensured by doubling transitions, the actual symbol (or line) rate becomes 18.432 MHz.
Consumer-grade hardware running at 20bit/48kHz (Z=75R) will have a symbol rate of 6.144MHz.
These are the rates at which the signal transitions on the S/PDIF cable occur.
Reclocking (for jitter) and correction (for errors) are well understood, built into the standards and are integral part of the receiver circuitry, for S/PDIF.
Tweak if you will, but if you don't hear any problems currently; you certainly will not see benefits, except -possibly- for sleeping better at night.

 

[mcdn]

Connectors have a characteristic impedance too, and if it's mismatched with that of the cable or the terminating impedance you can also get reflections. BNC connectors come in 50R and 75R varieties, and if you use the wrong one in an application where it matters you'll break things - mixing things up between video and 10Base2 networking for example. RCA doesn't have a standardised characteristic impedance, but its dimensions impose some limits on what impedance they can be made in - and 75R is outside that range. Luckily for short cables in this application it doesn't matter much.

Fair point, but if that's a concern you should also worry about the traces on the PCB - and some do, but not for the <10MHz range we're talking about.

 

[klettermann]

Unlikely that you will be operating at 24bit/192kHz data rates on the S/PDIF coax… however, due to biphase mark code (BMC) encoding, clock recovery/extraction is ensured by doubling transitions, the actual symbol (or line) rate becomes 18.432 MHz.
Consumer-grade hardware running at 20bit/48kHz (Z=75R) will have a symbol rate of 6.144MHz.
These are the rates at which the signal transitions on the S/PDIF cable occur.
Reclocking (for jitter) and correction (for errors) are well understood, built into the standards and are integral part of the receiver circuitry, for S/PDIF.
Tweak if you will, but if you don't hear any problems currently; you certainly will not see benefits, except -possibly- for sleeping better at night.

Hahaha, yes, exactly! Even in the absence of anything else I'll sleep better. And my sleep needs all the help it can get.

 

[Mnyb]

Yes it should have been BNC from the beginning some old transports have this , but in practice it does not matter for a video signal on the other hand ….

Afaik digital audio in short cables does not stress the interface it’s basically reused analog video hardware the 75 ohm standard and it’s ment for much higher frequencies, then you have problems.

 

[mcdn]

This is an interesting bit of commentary on RCA "75 ohm" plugs: https://www.bluejeanscable.com/articles/75ohmrca.htm

 

[xanalog]

Hahaha, yes, exactly! Even in the absence of anything else I'll sleep better. And my sleep needs all the help it can get.

...but if you are going to do this BNC swaparoo to impress friends/relatives, you might as well go with SMAs.
The latest fad, coming from the other side of the Pacific, is SMA connectors on everything RF and/or digital.

Dirt cheap to make; dirt cheap to buy; saves on materials; saves on shipping; and possibly saves us from another existential threat!

meh to BNCs... your friends/relatives surely will say 'wtf, yo?'