Hey there,
what do you think is the best way to terminate coaxial cables for internal wiring purposes? There are many ways:
a) Strip the outer insulation, twist the braided shield together and solder it to the ground connection point.
b) Same as a), but solder a thin, flexible wire to the shield, insulate with shrink tubing.
c) ...
A first attempt looks like below, but I realized later that I created a somewhat large current loop by going out of the back end with the green wire. Should have gone out of the front end directly to minimize loop area.
Other ideas? How do you usually solve this?
Regards,
Rundmaus
what do you think is the best way to terminate coaxial cables for internal wiring purposes? There are many ways:
a) Strip the outer insulation, twist the braided shield together and solder it to the ground connection point.
b) Same as a), but solder a thin, flexible wire to the shield, insulate with shrink tubing.
c) ...
A first attempt looks like below, but I realized later that I created a somewhat large current loop by going out of the back end with the green wire. Should have gone out of the front end directly to minimize loop area.
Other ideas? How do you usually solve this?
Regards,
Rundmaus
You may not want to ground the load end of the cable's shield, to avoid a ground loop.
If this is for audio circuits, it may be best to ground the shield's source end,
and to insulate the shield's load end.
That's a good point - but the question was simply about the practical design of the termination...
Regards,
Rundmaus
Regards,
Rundmaus
While it's not a big deal for a completely internal cable, that pig-tail is rather long. Shorter is better than longer. 2cm (1/2 inch) is good.
I think that a single-braid coax is just the same as any other 2-wire unbalanced connector, that is to say both the core wire and the shield must be connected at both ends. Current flows from the source to the receiver down the core wire, and there must be a return path to complete the circuit - this is the shield for plain coax. You do not want the return path to have to go out the PS of the "receiver" end equipment, out the mains cord to the wall socket, back up the mains cord of the "source" equipment, and then through its power supply in order to complete the return path. That is the only other available path if there are no other interconnections between the equipment. Now THAT would be a nice loop!
The construction of coax is to maintain a certain impedance at high frequencies (e.g. 50 or 75 Ohms) not to provide shielding. If you want a shielded unbalanced interconnect for audio signals you should use something like microphone cable (two twisted leads plus a shield). Some options here:
Microphone Cable & Signal Cable in the Wire & Cables Department at Parts Express | 1610
In that case, you can connect the shield at one end only and use the two twisted wires to carry signal and return.
At one time I thought "that shielded coax will be awesome for interconnects" and I bought hundreds of feet of bulk coax wire a crimper tool, and a gross of RCA crimp-on ends. I used this for long runs between a miniDSP and a powered loudspeaker, 4 channels to each speaker. Unless I used rubber bands to keep all the coax lines very close together I would pick up hum, in fact I could pick up the loose cables (not held together with rubber bands) and separate and bunch together the cables and this would very audibly increase and decrease the hum.
Last edited:
(two twisted leads plus a shield) you describe whats become known as pseudo-balanced. Some don't believe it's a good thing saying the unconnected screen at one end acts like an antenna for noise. I found it worked fine even with 15 foot cables but i don't have that much radio frequency interference here.
Whether you connect the shield at both ends depends on exactly how you are using the cable in the circuit - there is no recipe to follow. If you simply need an electrostatic shield then connect it at one end - in this case you must have arranged some other path for the signal return current. If the shield is part of the return path then you need to connect it at both ends.
The general rule is to keep the pigtail short. I just use the shield itself. Audio is probably not too bothered, but RF likes loops and long connections so it can sneak in where it is not wanted.
The general rule is to keep the pigtail short. I just use the shield itself. Audio is probably not too bothered, but RF likes loops and long connections so it can sneak in where it is not wanted.
No. The construction does both. Both are necessary.CharlieLaub said:
This is OK if the source is floating. In most cases inside equipment (which is what this thread is about) the source is not floating so a pseudo-balanced connection is not appropriate.
You probably had a ground loop formed by the two stereo channel shields. Solution is to keep them close together, maybe even loosely twisted around each other.
I followed DF96s advice and used the shield directly instead of a pigtail, so loop area as well as unshielded length are minimized. A piece of shrink wrap reduces mechanical strain on the fragile inner conductor connection.
Regards,
Rundmaus
Regards,
Rundmaus
It is a ground connection to a shield, I can't imagine an extra inch of wire on the ground side will matter.
It is also the return path for the signal...
Shield's terminated at one end only can act as an antenna and pick up RF noise, using a cap to terminate the second end is a better option.
Shield's terminated at one end only can act as an antenna and pick up RF noise, using a cap to terminate the second end is a better option.
Wha???
As rayma said in post #2, we generally run only one end shield grounded to avoid ground loops. If my circuit is picking up RF on the shield of a cable inside the unit and an inch of ground wire makes a difference, we have much more serious problems in the amp.
As rayma said in post #2, we generally run only one end shield grounded to avoid ground loops. If my circuit is picking up RF on the shield of a cable inside the unit and an inch of ground wire makes a difference, we have much more serious problems in the amp.
As I said only terminating one end of a shield makes it an antenna, likely to pick up RF noise.
Internally if you are using co-ax then you need to use it correctly, and that means using the shield as the return current path.
Terminating one end of the cable within a component is a bit silly (as it is when co-ax is used to connect components such as a pre-amp to an amp, where does the return current flow, it is the differing return current paths of audio signals (due to frequency) that often are the cause of hum (ground loops). Using a good co-ax with a low screen resistance is the best option.
Internally if you are using co-ax then you need to use it correctly, and that means using the shield as the return current path.
Terminating one end of the cable within a component is a bit silly (as it is when co-ax is used to connect components such as a pre-amp to an amp, where does the return current flow, it is the differing return current paths of audio signals (due to frequency) that often are the cause of hum (ground loops). Using a good co-ax with a low screen resistance is the best option.
Roundmouse, is that a pot meter? What are you using it for that you only have two wires connecting each channel?
Vacuphile,
it is a standard two-channel volume pot. The two solder lugs where the shields are connected go to the ccw ends of the pot, the inner conductors are connected to the wipers, they lead to the amplifier input section.
The 2nd set of connections from the input selector is still missing, that's what the 2nd insulating post is for. Was late yesterday, I'll finish it today!
Regards,
Rundmaus
it is a standard two-channel volume pot. The two solder lugs where the shields are connected go to the ccw ends of the pot, the inner conductors are connected to the wipers, they lead to the amplifier input section.
The 2nd set of connections from the input selector is still missing, that's what the 2nd insulating post is for. Was late yesterday, I'll finish it today!
Regards,
Rundmaus
MArce: inside a unit, I might have to run a signal from the rear of the board to a front panel control or input jack. The jack is grounded at the panel, and the board is just getting the signal lead. there is no return path through the shielded lead then. We ground the shield at one end so there are not TWO ground paths. The OP was not asking about interconnecting cables for outside the unit. I don't see my ground wires becoming antennas in my gear.
If I cut the ground on a interconnecting cable, then yes, we'd pick things up, but that is a separate topic for another thread.
If I cut the ground on a interconnecting cable, then yes, we'd pick things up, but that is a separate topic for another thread.
You'd be surprised... depends how good the shielding is.
Where does the input return current flow then! Signal and return should always be in intimate contact, co-ax, twisted pair, broadside coupled even figure of 8.
Where does the input return current flow then! Signal and return should always be in intimate contact, co-ax, twisted pair, broadside coupled even figure of 8.
The jack (or pot) should not be grounded at the panel. It should get its ground from the coax shield. One exception: an old-fashioned pot in a metal housing can have the housing grounded at the panel, but the cold end of the track should still get its ground from the cable. Audio return currents should always be kept well away from the chassis.Enzo said:
It's like DF96 and marce said, the RCA jacks are isolated from the panel and I intend to use the coax shield as the signal return, as shown here:
http://www.diyaudio.com/forums/analog-line-level/298666-sanity-check-input-grounding-scheme.html
Regards,
Rundmaus
http://www.diyaudio.com/forums/analog-line-level/298666-sanity-check-input-grounding-scheme.html
Regards,
Rundmaus
Finished the wiring to the volume pot, next step is the source selector and the input RCA lines.
I followed the advice to connect the shield directly without pigtails to make loop area as small as possible. Needed to add some mounting hardware due to the rigidity of the RG-179 insulation though. The cable is too stiff to keep all the mechanical load on the solder connection of the shield.
Regards,
Rundmaus
I followed the advice to connect the shield directly without pigtails to make loop area as small as possible. Needed to add some mounting hardware due to the rigidity of the RG-179 insulation though. The cable is too stiff to keep all the mechanical load on the solder connection of the shield.
Regards,
Rundmaus
Last edited:
- Status
- Not open for further replies.