I'm gonna have to reorder some parts!
Just to be clear about the input. The RCA input on the UA-1 was grounded / shielded by the chassis so in order to avoid any loops I should isolate the RCA jack at the chassis and ground the other end of the input cable to my star point.
Thanks.
Just to be clear about the input. The RCA input on the UA-1 was grounded / shielded by the chassis so in order to avoid any loops I should isolate the RCA jack at the chassis and ground the other end of the input cable to my star point.
Thanks.
As this is an unbalanced system I suspect it will not make any difference. It was common practice in unbalanced gear of that era to connect zero volts to the chassis only at the input connector, Today, for safety reasons, you also need to connect the chassis to mains safety earth. If all your unbalanced gear is built like this then there is a potential for loops because there is on 0V path through the audio cables and another via the mains earths. This in unavoidable in unbalanced systems and any loops that cause hum can be cured by lifting the screens on one or more audio cables ( and NEVER by lifting mains safety earth).
Floating the RCA socket but connecting it to the chassis elsewhere makes no difference to the above.
Cheers
Ian
Lifting the screen at one end, when the screen is actually the Signal Return, forces the equipment to find another route for the Return Signal.
That is likely to be even worse than the RF pick-up.
That is likely to be even worse than the RF pick-up.
I still think the screen is a screen and the cable should have two conductors with a screen around them.
Each build will be different regards layout and grounding. The rules of star earth etc. apply as a guide then its sort out the hum and modify with other guidelines.
The safety of the equipment is paramount and being aware of what can happen under given circumstances. (Recognise danger when the equipment is put in a system)<<difference in potential between metal parts.
That is where silver can sometimes be useful the conduction on a buss bar voltage drop etc. But systems have been built without it for years..
Keeping input ground (signal) and chassis ground (shield) as a separate consideration.
The dark art bit is because layouts are different and voltages within the grounds will be different and the application of grounding to keep voltage drops under control. The thing to remember is there can only be current flow in a ground with a difference in potential. (No matter how small the difference is)
Regards
M. Gregg
Last edited:
many builders believe that Coaxial cable is better than twisted pair. (I don't)
Many builders might experiment with breaking a coax screen at one end. (I have)
Would any builders think about experimenting by breaking one end of one of the twisted pair? (I never have)
The problem tends to be that multiple paths for ground create circuits and current flow. If the shield is part of the signal path and you cut it then the signal has to return another way (possibly through the mains Earth).
If you reconnect the shield and disconnect the Earth the hum goes away but now you have no safety ground and the shield becomes the safety ground. If you pull apart the connection, you have a ground in one hand and the earth in the other a real danger.
People say that if two parallel grounds are connected at the same points then no difference in potential can exist, however you now have a shorted turn that will pick up magnetic fields etc. This might not be a problem with low gain but it is at high gain.
The mains earth is noisy it has all sorts of hash on it and transients..even mains filters are connected to grounds as one example.
So if you regard the chassis and safety Earth as one item, then have the circuit with minimum differences in potential and ground as a separate item you are half way their. You can now look at connecting the circuit ground to Earth potential at one point or one point for left and one for right that are taken to Earth. Now if that signal ground is shorted between right and left inside another component CD player etc. it will short the left and right channel together and you have a shorted turn ground. (pickup) So now people break one screen connection and the hum goes away etc.
Many types of commercial equipment are double insulated no Earth to worry about..
I like twisted pair or twisted with a screen and drain wire. It depends on how well the cable is driven by the circuit.
Regards
M. Gregg
Last edited:
H.OTT tells us that the drain wire + screen is not as good as screen alone.
He does tell us repeatedly to adopt a screen that has low resistance.
Adding a drain wire does not seem to be an acceptable way to lower the resistance.
But he does show a separate common wire connecting screen connected modules to lower the resistance between the modules.
I can't recall seeing a comparison with this extra wire to a drain wire, nor why one is good and yet the other is bad.
He does tell us repeatedly to adopt a screen that has low resistance.
Adding a drain wire does not seem to be an acceptable way to lower the resistance.
But he does show a separate common wire connecting screen connected modules to lower the resistance between the modules.
I can't recall seeing a comparison with this extra wire to a drain wire, nor why one is good and yet the other is bad.
I rebuilt one pair of UA-1's using the SDS labs power supply boards. I didn't replace the old electrolytic cans on the amp and covered the mounting hols up with perfed circuit boards. The underside was a convient place to collect my star ground wires and then send it back to the SDS board. All this was electcrictly isolated from the chassis. I had no trouble using the chassis as a shield for the RCA without causing some sort of loop somewhere. The amps work wonderfully and are dead silent now.
I'd like to mention that using a Earth ground is something that I have no problem doing but just want to make sure that using the chassis as a shield is still okay. Or should I isolate the RCA from the chassis and just have the signal ground covering sttached at the star ground only.
Thanks a lot.
I'd like to mention that using a Earth ground is something that I have no problem doing but just want to make sure that using the chassis as a shield is still okay. Or should I isolate the RCA from the chassis and just have the signal ground covering sttached at the star ground only.
Thanks a lot.
The problem,
with a screen connected at one end is that one end of the screen is effectively closer to the Earth than the other end. This is also the case with a drain wire. However you can reduce this effect by using silver plated screen and/or drain wire. (Part of the myth or fact of directional construction) not going there..LOL
However if this was the case then which equipment has the best Earth/shield to connect the grounded end of the cable to..LOL
The problem is also that theoretically there should be no voltage drop across the screen because its not connected to anything..of course that isn't the whole story because you have a capacitive effect and a pick up effect RFI etc.
You can reduce the effect of a drain wire by using silver plated cable etc.
And or wrapping it around the shield not just along one side. about 20mm twist.
The problem of using an extra wire to link all the equipment together is an attempt to reduce differences in potential across cables/screens etc. ie mask over the problem. It can work quite well, but its still better to not have the problem in the first place..it doesn't remove shorted turns..
The thing to remember is the safety Earth is a conductor and must be able to carry enough fault current to operate the fuse, what you don't want is grounding cables acting as a safety Earth or as the only safety Earth. If its in parallel with the earth under fault it will carry current.
I prefer to have the same metal etc as the ground and input/output.
Ie not some aluminium screen etc. (Its just a personal view).
Regards
M. Gregg
with a screen connected at one end is that one end of the screen is effectively closer to the Earth than the other end. This is also the case with a drain wire. However you can reduce this effect by using silver plated screen and/or drain wire. (Part of the myth or fact of directional construction) not going there..LOL
However if this was the case then which equipment has the best Earth/shield to connect the grounded end of the cable to..LOL
The problem is also that theoretically there should be no voltage drop across the screen because its not connected to anything..of course that isn't the whole story because you have a capacitive effect and a pick up effect RFI etc.
You can reduce the effect of a drain wire by using silver plated cable etc.
And or wrapping it around the shield not just along one side. about 20mm twist.
The problem of using an extra wire to link all the equipment together is an attempt to reduce differences in potential across cables/screens etc. ie mask over the problem. It can work quite well, but its still better to not have the problem in the first place..it doesn't remove shorted turns..
The thing to remember is the safety Earth is a conductor and must be able to carry enough fault current to operate the fuse, what you don't want is grounding cables acting as a safety Earth or as the only safety Earth. If its in parallel with the earth under fault it will carry current.
I prefer to have the same metal etc as the ground and input/output.
Ie not some aluminium screen etc. (Its just a personal view).
Regards
M. Gregg
I used a parallel earthing conductor (PEC) to good effect on my Midlife Crisis mono amps. They consist of 9.5 gauge copper litz wire with copper geophone clips on the ends; clipping them onto the RCA grounds at each end of the input RCAs from the preamp to the amps reduced hum and buzz from barely audible at 12" from the speaker to silent.
No idea why he connects different modules together, but low-resistance screen does appear to have a significant effect.
Hum Rejection in Unbalanced Audio Cables -- Blue Jeans Cable
You would have to read "Electromagnetic Compatibility Engineering" and the info on his site, as AndrewT's comment barely scratches the surface. (Ott is a recognised expert in this field, so taking his solutions into consideration is recommended).
home page
One trick that is used often is to have spare Ground connection points on PCBs and cases to allow the connection of thick cable with very low resistance between two pieces of equipment (be they boards in a case or different units interconnected), this will provide a low impedance path between the two unbalanced grounds (at different potential because there is a high impedance connection between them) lowering the difference in potential between them and thus removing the noise caused by a ground loop.
home page
One trick that is used often is to have spare Ground connection points on PCBs and cases to allow the connection of thick cable with very low resistance between two pieces of equipment (be they boards in a case or different units interconnected), this will provide a low impedance path between the two unbalanced grounds (at different potential because there is a high impedance connection between them) lowering the difference in potential between them and thus removing the noise caused by a ground loop.
If you have a ground loop in an unbalanced system you have to lift a ground to fix it. There are only three options:
1. Lift the mains safety ground. This is very dangerous and should NEVER be done
2. Lift the screen of the signal cable. RFI is not likely to be a problem. An unbalanced cable gives no screening against the H-field of RFI and the screen, as an equipotential, stops most of the E-field.
3. Insert a transformer to galvanically isolate the two ends of the signal circuit.
Cheers
Ian
Sorry, but you can engineer a solution. I would suggest a read of Henry Ott, Tony Waldran and many others.
Lifting the screen at one end leaves you open to rf ingress, look at the many solutions to do proper EMC screening, often ignored in DIY design.
Lifting the screen at one end leaves you open to rf ingress, look at the many solutions to do proper EMC screening, often ignored in DIY design.
Of course, you can 'engineer' a solution to a hum loop, but it will involve breaking the loop. You are not being much help if you say 'look at the many solutions to do proper EMC screening'.
Cheers
ian
And why does promoting and providing sources of information such as Henry Ott not help. It is an EMC problem anyway so learning the mechanisms and how to solve them is paramount. And you don't have to break the loop just remove the conditions that cause it, breaking the loop is one solution, not creating the situation where you cause a loop is another. A loop is due to high impedance path between two 'ground' points causing a current to flow between these points. One of the usual causes is unbalanced interconnects with a high impedance braided or foil screen, where both bits of equipment are connected to PE this means there are two paths for current to flow (the screen and via PE), so quite often the lower frequencies especially the mains frequency will follow the PE path the higher frequencies the screen (there is no hard divide, both paths will be used by all frequencies but lower frequencies will prefer the path of least resistance the higher the path of lowest impedance). Breaking the loop is one method, but you still have to worry about rf especially these days and so shielding should be continuous and connected to the case at cable entry by one of the many accepted methods.
Read the last paragraph in #33 above, that is how it is often done in real world situations in many many other fields of electronics, this at least allows you to isolate and investigate the problem more.
The whole issue is EMC based problems and with todays increasing RF etc the whole situation has to be solved....
Read the last paragraph in #33 above, that is how it is often done in real world situations in many many other fields of electronics, this at least allows you to isolate and investigate the problem more.
The whole issue is EMC based problems and with todays increasing RF etc the whole situation has to be solved....
Some interesting info on ground loops can be found in Jneutrons gallery:
system_ground_layout_options - My Photo Gallery
system_ground_layout_options - My Photo Gallery
If someone asks about ohms law, it is not helpful to say it is an electronics problem so first you need to go to university and get a degree in electronics. Ott et al have written a huge amount on EMC; rather than ask the OP to read it all, it would be a lot more helpful to point him to some specific articles that will actually help him solve his problem such as the link in your most recent post. In fact if you read this article by Ott:
http://www.hottconsultants.com/pdf_files/Audio%20Interconnections.pdf
He say pretty much what I did.
Cheers
Ian
Last edited:
- Status
- Not open for further replies.