wow ok I'm going to have to read that a few more times but I think the part where it said the 2 signals do not have to be symmetric but must have the same impedance. Thank you so much for all your help!
Help with capsule choice
- Thread starter skyy7
- Start date
Help Support GroupDIY Audio Forum:
Two different concepts, signal symmetry (is pin 2 signal the same but opposite polarity as pin 3 signal), and impedance balance (is the impedance from pin 2 to the chassis the same as pin 3 to the chassis).
For noise rejection with a differential input the impedance balance is what needs to be maintained. The impedance at each pin is determined by the PNP emitter-follower circuits, so the balance will be maintained even when the signal is not symmetrical.
What will change with a frequency-dependent difference in only one leg is the slope of the combined output.
For my derivation of that see this earlier post.
I was wondering for a while now but forget to ask, output pair gets different impedance output from FET(source out has lower impedance than drain), does this reflect in anyway to the pair output, if you get what i mean. @Khron and others are welcome to chime in.
Well, it may not be a big problem, but you can trigger common mode hum issues under extreme conditions and notice the difference with other output configurations. I just did the following, unscientific test: I took a Behringer B-5 (Single-Ended, sort of impedance-balanced output), an MXL 991 (classic balanced Schoeps output) and a Behringer B-5 with my own circuit design that has a balanced Sziklai (CFP) output.
I connected the mics one by one to my Mackie ProFX12, Gain was set to max, , Output level to max and Headphone level to max. The microphone cable consisted of several shorter XLR cables, together measuring ~25 meters. The original B-5 picked up the most hum that was clearly audible through the headphones, and when I touched the XLR connectors in the cable, the hum level increased noticeably. The audible hum from the MXL 991 was audibly lower and was also less sensitive to touching the XLR connectors. And surprise, surprise: when the B-5 with Sziklai output was connected, there was no noticeable hum, even when XLR connectors were touched. Mind you: This mic is ~8 dB hotter than the original B-5. XLR connectors were not galvanically connected to the screen, and touching them must have coupled Common Mode hum into the cables.
Again, whether this is a real issue under normal circumstances, I don't know. You should judge for yourself. For me, the mere fact that, theoretically, you can generate common mode hum would be a reason to always choose a CFP output in a Schoeps circuit. Who cares about just a few cents extra on the BOM if all you want is undisputable quality? And finally, there are other reasons to choose the CFP output, such as lower distortion.
Jan
I connected the mics one by one to my Mackie ProFX12, Gain was set to max, , Output level to max and Headphone level to max. The microphone cable consisted of several shorter XLR cables, together measuring ~25 meters. The original B-5 picked up the most hum that was clearly audible through the headphones, and when I touched the XLR connectors in the cable, the hum level increased noticeably. The audible hum from the MXL 991 was audibly lower and was also less sensitive to touching the XLR connectors. And surprise, surprise: when the B-5 with Sziklai output was connected, there was no noticeable hum, even when XLR connectors were touched. Mind you: This mic is ~8 dB hotter than the original B-5. XLR connectors were not galvanically connected to the screen, and touching them must have coupled Common Mode hum into the cables.
Again, whether this is a real issue under normal circumstances, I don't know. You should judge for yourself. For me, the mere fact that, theoretically, you can generate common mode hum would be a reason to always choose a CFP output in a Schoeps circuit. Who cares about just a few cents extra on the BOM if all you want is undisputable quality? And finally, there are other reasons to choose the CFP output, such as lower distortion.
Jan
Matt Nolan
Well-known member
As per AES convention, the shells should indeed be connected to pin 1 (screen). But up to now, the vast majority of XLR cables did not have the shell connected to pin 1. Why I know that? Because I used to trade used pro audio gear and knowing that many XLR plugs/cables were not properly soldered, I often opened the XLR's to check the solder quality before I sold the cables with equipment. Some of the cables I used are from my son's rental service. He builds the cables himself and never connects pin 1 to the shell. He's made hundreds of cables, also for a studio builder. Never had any hum issues.
The reason for not connecting pin 1 to the shell is to prevent ground loops. Not relevant in context of mic cables, but XLR cables are often used universally for mic and equipment connections. Theoretically, a "floating" piece of metal close to conductors carrying audio signal could couple noise into the conductors, which I actually demonstrated, but should generally not bean issue, AFAIK. My best guess is that groundloop induced hum is probably a greater risk than by "floating" shells.
More on this here: Q. How should I wire XLR cables?
But I'm afraid this is getting very off-topic now, so let's leave it at that.
Jan
The reason for not connecting pin 1 to the shell is to prevent ground loops. Not relevant in context of mic cables, but XLR cables are often used universally for mic and equipment connections. Theoretically, a "floating" piece of metal close to conductors carrying audio signal could couple noise into the conductors, which I actually demonstrated, but should generally not bean issue, AFAIK. My best guess is that groundloop induced hum is probably a greater risk than by "floating" shells.
More on this here: Q. How should I wire XLR cables?
But I'm afraid this is getting very off-topic now, so let's leave it at that.
Jan
That is incorrect. Per standard AES54-1:
4.1.2 Connection of shields
The shields of all portable cables and tie lines shall be wired to the designated shield contact at each end, except
as provided under 4.3.2. There shall be no DC connection between the designated shield contact and any
conductive shell.
AES48 does call for pin 1 to connect to chassis in the terminating equipment. The reason as explained to me is that in large installations (think stadiums, conference centers, other really large commercial buildings) all exposed metal is required to be bonded to the building safety earth system. That would often result in patch panels being bonded to the building protective earth, so for long connections (e.g. one end of a stadium to the other) if cables were connected through the patch panels, the cable shells will be connected to building ground at various points not easily under control of the sound system designer.
By connecting the cable shields only to pin 1 and not to the connector shell you give up high frequency shielding effectiveness, but get the advantage that the low frequency connection points are under your control, namely at the equipment at each end of the full cable run, and you can decide based on system needs whether the shield to local gnd connection should be lifted (either using a switch in the equipment if provided, or an adapter connector).
OK, thanks; then I've been misinformed about that. Anyway, all of my XLR cables are AES54-1 compliant and do not have pin 1 connected to the shell.
Jan
Jan
Matt Nolan
Well-known member
When you chain cables it presents a problem. You have segments where the connectors are where the balanced wires inside are neither in a twisted pair nor have a screen around them. Hum and other interference will get in and not necessarily in equal common mode amounts.
You can lose a chassis connection at one end of a (chain of) XLR cables, but one end only. Not in the middle. In the microphone in a small studio situation, it is very unlikely to have a ground loop anyway.
That's true, and it was the whole point of my test: to create a situation where I could potentially demonstrate the difference in CMRR between various microphones. If I had used just a single, short XLR cable and a mixer with very CMRR of its own, I would probably not be able to demonstrate differences between different output stage topologies. So you could see this as a proof that you'd better not chain (many) XLR cables.
Jan
Jan
