My own view and practice: if the connection is inside the chassis and there is no significant RF creating circuitry inside the chassis (no switching power supply; linear supply sing soft recover diodes, etc), then a small gap between the end of the shield and the twisted signal wires inside the braid when correcting to the solder cups, terminals or PCB located within the chassis is not going to be an issue from the standpoint of RF.
Braid or foil shielding is not going to help with "magnetic" fields from AC power cables and transformer emanations. For that, you need to use twisted AC lines, and steel shields.
Braid or foil shielding is not going to help with "magnetic" fields from AC power cables and transformer emanations. For that, you need to use twisted AC lines, and steel shields.
You must use a two wire connection system.
Coaxial: core for hot and screen/shield for cold. Pin1 only goes to chassis.
Shielded twisted pair: one core for hot and one core for cold. The screen goes to pin1 and then to chassis.
Twisted pair: one core to hot, one core to cold. Pin1 to chassis only.
Andrew, Henry is asking specifically about how to solder the shield of his twisted pair shielded microphone cable to Pin 1 and the chassis tab of the XLR connector.
Henry, you mentioned that the tab and Pin 1 were only a bit apart and asked if you could just jam the shield in there and make a solder blob. I would prefer that you thread the twisted shield through the solder hole in the tab and route it to pin 1 with the shortest possible connection.
Try to leave as little of the signal pair exposed as possible. Don't kill yourself doing this. If it takes you an hour to solder two wires to the XLR when they're 8 mm long but a few minutes if they're 12 or 15 mm long, go for the 12-15 mm. It's just not that critical, especially not in a residential application.
If assembled according to the instructions, a Modulus-86 + Power-86 will not emit any RF energy, so there are no internal interference sources to worry about. Even with a switching power supply, I wouldn't worry overly about RF energy getting into the Modulus-86. The circuit layout is very compact and all the high-impedance nodes were kept short. In addition, the inputs have RFI filters on them. That is one of the many reasons I measure the exact same performance of the Modulus-86 when it's powered by a linear supply (Power-86 + transformer), a switching supply (Connex SMPS300RE), or a well-regulated lab power supply (HP 6643A).
Tom
Henry, you mentioned that the tab and Pin 1 were only a bit apart and asked if you could just jam the shield in there and make a solder blob. I would prefer that you thread the twisted shield through the solder hole in the tab and route it to pin 1 with the shortest possible connection.
Try to leave as little of the signal pair exposed as possible. Don't kill yourself doing this. If it takes you an hour to solder two wires to the XLR when they're 8 mm long but a few minutes if they're 12 or 15 mm long, go for the 12-15 mm. It's just not that critical, especially not in a residential application.
If assembled according to the instructions, a Modulus-86 + Power-86 will not emit any RF energy, so there are no internal interference sources to worry about. Even with a switching power supply, I wouldn't worry overly about RF energy getting into the Modulus-86. The circuit layout is very compact and all the high-impedance nodes were kept short. In addition, the inputs have RFI filters on them. That is one of the many reasons I measure the exact same performance of the Modulus-86 when it's powered by a linear supply (Power-86 + transformer), a switching supply (Connex SMPS300RE), or a well-regulated lab power supply (HP 6643A).
Tom
Thanks AndrewT. Your second example is what I plan on doing with a slight variation.
I think I mentioned that I have Neutrik NC3FD-LX-HE XLRs. They have a small tab, which is conductive to the body, right next to pin1 solder connection. The purpose of the tab is to provide the option of easily grounding pin1 to the chassis. It is used instead of a longer wire from pin1 to the chassis.
My question was about using the tab for chassis ground, as intended by Neutrik, but with a variation.
The space, between the tab and the outer diameter of pin1 solder connector, is less than the thickness of the twisted shield wires. I thought I would push the twisted shield wires in that space and join the tab, wire and pin1 connector as one solder joint. That eliminates the need of a second wire and makes the length of the connection less than 1/16" long. Any downside to doing that?
I think I mentioned that I have Neutrik NC3FD-LX-HE XLRs. They have a small tab, which is conductive to the body, right next to pin1 solder connection. The purpose of the tab is to provide the option of easily grounding pin1 to the chassis. It is used instead of a longer wire from pin1 to the chassis.
My question was about using the tab for chassis ground, as intended by Neutrik, but with a variation.
The space, between the tab and the outer diameter of pin1 solder connector, is less than the thickness of the twisted shield wires. I thought I would push the twisted shield wires in that space and join the tab, wire and pin1 connector as one solder joint. That eliminates the need of a second wire and makes the length of the connection less than 1/16" long. Any downside to doing that?
You're pretty mechanically inclined, so I suggest thinking about it in terms of mechanics. So mating area and such. Keep in mind that for a good connection, you need a nice cone of solder around the joint. In case of the solder cup, you need a nice curtain of solder draped over the conductor along its entire length within the cup. Also keep vibration in mind.
I'd be curious for your evaluations of these scenarios:
1) Wire staple from Pin 1 to the chassis tab. Screen soldered to chassis tab.
2) Wire staple from Pin 1 to the chassis tab. Screen soldered to Pin 1.
3) Screen jammed between Pin 1 and the chassis tab and soldered there.
What you're after is something that's mechanically solid. In addition, you want the shortest connection of the shield as possible. That means the distance from the incoming cable shield, through the two mating connectors, to the cable shield on the inside of the the enclosure should be minimized. Recall the external cable shield is connected at Pin 1 of the mating connector and does not connect to the shell of the connector.
With these constraints in mind, I'm curious for your thoughts on options 1-3 above. I'm pretty sure that with your mechanical savvy you can find an optimum. I suggest drawing it out on a piece of paper if that helps.
My goal here is to nudge you along a little bit so you can think more clearly about the tradeoffs and progress a little faster on your project.
Tom
Last edited:
While I am anxious to finish this build, I'm enjoying learning a bit about electronics.
The AC entry ground terminal will be connected to a, conductive, chassis panel. The XLR body will be grounded to earth by its contact with the same panel and its mounting screws.
Between what I've read here and the response from the Neutrik tech, it seems that, alone, should provide a good RFI/EMI shield around the XLR. That assumes the preamp is grounded correctly.
Bottom lines is it seems connecting pin1, to the chassis, with a second wire or whatever, will only provide a second ground to earth for the shield. This makes me think that two ground connections for the shield could lead to a ground loop. Am I thinking correctly?
The AC entry ground terminal will be connected to a, conductive, chassis panel. The XLR body will be grounded to earth by its contact with the same panel and its mounting screws.
Between what I've read here and the response from the Neutrik tech, it seems that, alone, should provide a good RFI/EMI shield around the XLR. That assumes the preamp is grounded correctly.
Bottom lines is it seems connecting pin1, to the chassis, with a second wire or whatever, will only provide a second ground to earth for the shield. This makes me think that two ground connections for the shield could lead to a ground loop. Am I thinking correctly?
Last edited:
You're pretty close. The ground loop is inevitable in some form. Making the low impedance connection discussed here is preferable as doing so minimizes the voltage across it. Combined with the 60dB minimum CMRR of the Mod (the commonly mentioned 90dB figure for the THAT 1200 is the typical value) the result's best in class for any commercially available amplifier I'm aware of. There will still be some small error residual but in nearly all cases it'll be 20dB or more below the noise floor; quite comfortably under the ~0 dBGF threshold where things are no longer worth worrying about.
More on the topic over here as well as a few other places.
More on the topic over here as well as a few other places.
Last edited:
Regarding soldering; the Neutrik tech didn't think there would be any long term problems using simple 63/37 electronic solder on the gold plated XLR solder connections. Is that concern, I just read about, another example of "audiophile sales generated press"? Sorry to be blunt but.. time is passing.
I appreciate your response but its technically over my head. Are you saying a second wire from pin1 to its chassis enclosure is not the best scheme?
twest820 I appreciate your replies. Tom gave me some options and I was just trying to expand or describe my thoughts on what might be the best path.
My experience as an aerospace M.E., for over 40 years, has often led me down complex paths to find solutions. After getting there, it has often become apparent that simple worked.
My experience as an aerospace M.E., for over 40 years, has often led me down complex paths to find solutions. After getting there, it has often become apparent that simple worked.
The worst case CMRR of the THAT1200 is 70 dB at 60 Hz (and at DC). This assumes matched source impedances (the output impedance of the non-inverted output is the same as that of the inverted output). There are no minimum specs for any other frequency or output configuration.
I have seen samples that were around 80 dB CMRR and others that were 105 dB CMRR. I ended up with 90 dB in the spec table as a typical value.
Connecting pin 1 (shield) to the chassis routes any RF energy injected on the shield to the chassis. This is desirable. The recommended way to connect Pin 1 to the chassis is through the shortest connection possible as this minimizes the inductance between Pin 1 and the chassis, thus, providing the best RF path to the chassis. A small wire staple bent from a component lead will work fine here.
Tom
Yeah. You do seem to overthink things a bit.
If it results in learning, cool! Just beware that it has the power to lead you astray as well.No worries. All is well.
Tom
Dude. Seriously. You're way overthinking it. Take a a piece of component lead clipping or 20 AWG solid core wire. Bend it into an L shape so that one end of the L will go through the hole in the tab and the other into the solder cup. Solder both ends. Then solder the shield to pin 1. Done.
Tom
Tom
No worries. I do agree that the THAT1200 and, thus, Modulus-86 offer stellar CMRR. You can compare it with a decent discrete solution here.
Tom
That's not necessarily a bad thing. I must admit that I was secretly wondering if your way of overthinking was really just your way of procrastinating.
If I'm right, you may get something out of the book I'm currently reading, "The Procrastination Equation" by Piers Steel. I'm trying to become a bit more efficient in my use of time.
Tom