Hi all
I'm a total newbie here, so hello to everyone, and apologies in advance for this question - I'm sure it's a golden-oldie.
I have looked at some of the other threads on this forum about it, plus I've read the Bill Whitlock article and other online info - and I am happy to raise my knowledge greatly about the physics involved in ground loop and its solutions - but I'd still like to know what is the most common solution people use.
My setup:
The sole music source I use is a DAC - an Edirol UA-25 - USB into a PC.
This is the only input into a Sansui AU-2900 I've just bought.
My previous amp - an 80s Technics - was 2-pin plug, class-II - didn't suffer ground loop hum when connected to the PC/UA-25. But the AU-2900 is 3-pin Class-I - I'm assuming that's the difference.
* The AU-2900 doesn't have a hum when the RCA's from the UA-25 are out - and there are no other inputs - proving it's a ground loop between amp and DAC. I haven't got another device like a CD player so I can't check the AU-2900 with another unit to test that.
* It doesn't improve the hum when amp and PC are on the same power-board.
* Due to the layout of the room, the PC and the amp need a 6 metre audio lead. The lead is reasonable quality - fairly thick braided cable with good RCA connectors.
* The UA-25 has got a USB lead to the PC with bead chokes at both ends (in case that's relevant).
I have read about using isolators, and I can see there's another solution involving amongst other things inserting a resistor (from memory - I can soon find the full details for this). It seems that the isolators can have a cost in loss of signal - or do you need a really expensive one?
There'd be no point in having something which degraded the signal - I might as well just go back to the Technics or buy another amp with a 2-pin plug.
Or am I missing something here, with the AU-2900? The amp otherwise sounds really great, apart from this hum. It seems quiet without any inputs. Although it's 40 years old, I'm not hearing the tell-tale signs of failing caps.
Thanks in advance for reading this and considering this problem,
John
I'm a total newbie here, so hello to everyone, and apologies in advance for this question - I'm sure it's a golden-oldie.
I have looked at some of the other threads on this forum about it, plus I've read the Bill Whitlock article and other online info - and I am happy to raise my knowledge greatly about the physics involved in ground loop and its solutions - but I'd still like to know what is the most common solution people use.
My setup:
The sole music source I use is a DAC - an Edirol UA-25 - USB into a PC.
This is the only input into a Sansui AU-2900 I've just bought.
My previous amp - an 80s Technics - was 2-pin plug, class-II - didn't suffer ground loop hum when connected to the PC/UA-25. But the AU-2900 is 3-pin Class-I - I'm assuming that's the difference.
* The AU-2900 doesn't have a hum when the RCA's from the UA-25 are out - and there are no other inputs - proving it's a ground loop between amp and DAC. I haven't got another device like a CD player so I can't check the AU-2900 with another unit to test that.
* It doesn't improve the hum when amp and PC are on the same power-board.
* Due to the layout of the room, the PC and the amp need a 6 metre audio lead. The lead is reasonable quality - fairly thick braided cable with good RCA connectors.
* The UA-25 has got a USB lead to the PC with bead chokes at both ends (in case that's relevant).
I have read about using isolators, and I can see there's another solution involving amongst other things inserting a resistor (from memory - I can soon find the full details for this). It seems that the isolators can have a cost in loss of signal - or do you need a really expensive one?
There'd be no point in having something which degraded the signal - I might as well just go back to the Technics or buy another amp with a 2-pin plug.
Or am I missing something here, with the AU-2900? The amp otherwise sounds really great, apart from this hum. It seems quiet without any inputs. Although it's 40 years old, I'm not hearing the tell-tale signs of failing caps.
Thanks in advance for reading this and considering this problem,
John
This is a fairly straightforward explanation. I have disconnected the audio cable screen at one end to eliminate hum. It makes no difference to SQ imo.
Q. What is a Ground (Earth) Loop? |
Q. What is a Ground (Earth) Loop? |
ok rather than starting yet another thread, im in the process of building a combined psu/amp/minidsp in a single case. ive not got to the internal wiring yet, but want it to be as good as possible off the bat, to reduce troubleshooting later.
i have purchased some twisted 2 core shielded wire for internal signal routing. my idea was to use the two cores for signal and signal ground, then have the shield connected to the case for emi shelding. is this a reasonable idea? that way i can follow this guide:
Grounding and Shielding for your DIY Audio Projects
it states you should connect the shield at one end only, and if necessary run another single wire for the signal ground.
question is, if that second signal ground is twisted with the signal wire, is it still considered "another single wire" or is it now acting as a second "shield" and should only be connected at one end too?
argh.. im reading tons but visualising the knot of wiring im heading for is daunting.
i have purchased some twisted 2 core shielded wire for internal signal routing. my idea was to use the two cores for signal and signal ground, then have the shield connected to the case for emi shelding. is this a reasonable idea? that way i can follow this guide:
Grounding and Shielding for your DIY Audio Projects
it states you should connect the shield at one end only, and if necessary run another single wire for the signal ground.
question is, if that second signal ground is twisted with the signal wire, is it still considered "another single wire" or is it now acting as a second "shield" and should only be connected at one end too?
argh.. im reading tons but visualising the knot of wiring im heading for is daunting.
With the PC unpluged the remain?
This is the reason that my DAC is galvanic isolated between digital inputs (and logic circuits) and the analog output part.
You try the same electrical outlet for bouth, PC and Amplifier?
Can you try with an laptop only on battery?
Conect the shield of the cable to the ground at input. For the cable, the input GND is the reference so the shield must protect the cables from EMC with reference of the input GND.
This is the reason that my DAC is galvanic isolated between digital inputs (and logic circuits) and the analog output part.
You try the same electrical outlet for bouth, PC and Amplifier?
Can you try with an laptop only on battery?
Conect the shield of the cable to the ground at input. For the cable, the input GND is the reference so the shield must protect the cables from EMC with reference of the input GND.
Last edited:
Connecting the screen at the one end only is often referred to as a pseudo balanced cable. In order for the cable to carry the signal and return the cable needs to be connected to both signal and ground at both ends. If you use a 2 core cable you can attach both the screen and one core at one end, the other end of the (2nd) core wire to grd and leave the screen floating at that end. You must make sure that the RCA sockets are isolated from the chassis (if it is metal) check this with the continuity setting on your DMM. The second wire is not a shield but a grd connection.
ok thanks, that clears up some confusion i was having.
The rca jacks are isolated from the case, but also, if i am following the guide i linked there, then the termination point for all the grounding is going to be there at the signal ground of the rca jacks. ill run a wire from that point to the chassis ground at the power socket.
another grey area for me is this:
all boards (minidsp, amps, psu, bluetooth board) will be on isolated standoffs. so the only ground connection on the power side from those elements will be the ground return mains wire of the (smps) psu. this should be connected at the signal ground, with another cable running back to the chassis ground at the iec connector? does that sound right?
The rca jacks are isolated from the case, but also, if i am following the guide i linked there, then the termination point for all the grounding is going to be there at the signal ground of the rca jacks. ill run a wire from that point to the chassis ground at the power socket.
another grey area for me is this:
all boards (minidsp, amps, psu, bluetooth board) will be on isolated standoffs. so the only ground connection on the power side from those elements will be the ground return mains wire of the (smps) psu. this should be connected at the signal ground, with another cable running back to the chassis ground at the iec connector? does that sound right?
I have never connected RCA grd to mains protective earth. Is this recommended as the norm?
all boards (minidsp, amps, psu, bluetooth board) will be on isolated standoffs. so the only ground connection on the power side from those elements will be the ground return mains wire of the (smps) psu. this should be connected at the signal ground, with another cable running back to the chassis ground at the iec connector? does that sound right?
Where are you reading that the grd from the psu should be connected to signal grd?
all boards (minidsp, amps, psu, bluetooth board) will be on isolated standoffs. so the only ground connection on the power side from those elements will be the ground return mains wire of the (smps) psu. this should be connected at the signal ground, with another cable running back to the chassis ground at the iec connector? does that sound right?
Where are you reading that the grd from the psu should be connected to signal grd?
Last edited:
im sorry ive hijacked this thread somewhat.. i will continue the discussion on a thread related specifically to my project here:
tpa3255evm spacesaver mod
tpa3255evm spacesaver mod
Thanks to all - there's been some good stuff to read and get up to speed with from links and explanations in peoples' posts. By now, I completely get what ground loop is, and how balanced cables work, etc, but, I still haven't seen something which might solve what must be a common problem...
Questions:
* Is it feasible to disconnect the shield on the cable at one end - or the other suggestion of using a 2-core shield cable, and doing as Puffin suggested and leave the shield open at one end, and the 2nd core connected as return at the other end? I don't know if that would work.
* Is there an isolator which is good enough quality to think about using, which isn't too expensive? There's a Behringer MicroHD HD400 on ebay for around £30.
* Is it that the Sansui AU-2900 (from 1977) is so old that it lacks some circuitry which might make it better equipped to deal with problems thrown up by computers - eg EMI, shielding, grounding of internet or cables etc? Or is it just that it's 3-pin power, with a ground? Because my Technics amp (2-pin, class-II) didn't appear to have ground loop problems.
* Could I use a fat earth cable between the amp and the computer, to help force earth parity?
* Is the problem in the multi-plug power board? Is there a type of multi-plug power board which has better filters etc and might make things behave better?
Clutching at straws here.
Thanks in advance.
John
Questions:
* Is it feasible to disconnect the shield on the cable at one end - or the other suggestion of using a 2-core shield cable, and doing as Puffin suggested and leave the shield open at one end, and the 2nd core connected as return at the other end? I don't know if that would work.
* Is there an isolator which is good enough quality to think about using, which isn't too expensive? There's a Behringer MicroHD HD400 on ebay for around £30.
* Is it that the Sansui AU-2900 (from 1977) is so old that it lacks some circuitry which might make it better equipped to deal with problems thrown up by computers - eg EMI, shielding, grounding of internet or cables etc? Or is it just that it's 3-pin power, with a ground? Because my Technics amp (2-pin, class-II) didn't appear to have ground loop problems.
* Could I use a fat earth cable between the amp and the computer, to help force earth parity?
* Is the problem in the multi-plug power board? Is there a type of multi-plug power board which has better filters etc and might make things behave better?
Clutching at straws here.
Thanks in advance.
John
i can't speak for standards in your part of the world but in North America at that time there wasn't full conformity to current grounding practices and polarized plugs where slowly being introduced and a good many amp "cases" where floating, and some where capacitor coupled to the power supply ground.(that right no AC ground)
Thanks Puffin
But just to clarify...
You said 'yes' to this
* Is it feasible to disconnect the shield on the cable at one end - or the other suggestion of using a 2-core shield cable, and doing as Puffin suggested and leave the shield open at one end, and the 2nd core connected as return at the other end? I don't know if that would work.
Were you saying 'yes' to
1 disconnecting the shield at one end (and if so at amp or source end?) - or
2 the 2-shielded-core idea with shield disconnected at one end, and 2nd-core (acting as return for the other end) disconnected at the opposite end. Would you have an article that explains how this works? I also wouldn't mind seeing the best way to set it up - before I start cutting my leads up.
In answer to the other points - that Behringer HD400 - I just don't want to be putting anything in which would degrade the signal. I am growing fond of this AU-2900 amp, but it would defeat the purpose of it to have something there making it less good. I might as well just get a decent 2-pin class-II amp instead and avoid all this b*11*cks.
The other point - I was wondering if putting a fat wire from chassis ground to chassis ground - amp to PC - would equalise and take away the discrepancy between the two earth voltages, which is causing the ground loop hum.
John
But just to clarify...
You said 'yes' to this
* Is it feasible to disconnect the shield on the cable at one end - or the other suggestion of using a 2-core shield cable, and doing as Puffin suggested and leave the shield open at one end, and the 2nd core connected as return at the other end? I don't know if that would work.
Were you saying 'yes' to
1 disconnecting the shield at one end (and if so at amp or source end?) - or
2 the 2-shielded-core idea with shield disconnected at one end, and 2nd-core (acting as return for the other end) disconnected at the opposite end. Would you have an article that explains how this works? I also wouldn't mind seeing the best way to set it up - before I start cutting my leads up.
In answer to the other points - that Behringer HD400 - I just don't want to be putting anything in which would degrade the signal. I am growing fond of this AU-2900 amp, but it would defeat the purpose of it to have something there making it less good. I might as well just get a decent 2-pin class-II amp instead and avoid all this b*11*cks.
The other point - I was wondering if putting a fat wire from chassis ground to chassis ground - amp to PC - would equalise and take away the discrepancy between the two earth voltages, which is causing the ground loop hum.
John
A signal connection needs 2 wires.
That applies to the source to receiver link and the receiver to speaker link.
You must NOT break one of those connections.
A shield is an extension of the enclosure. It can be connected at one end, or both ends.
A shield that is connected at one end only does not give as much attenuation of interference. For effective screening at very high frequencies one should connect the shield at both ends to the enclosure.
Our WiFi, mobile phones and computers are very high frequency polluters.
Where one uses a screened single core cable to make a signal connection, both parts (the screen and the core) must be connected at BOTH ends. The screen is the return route for the signal. The fundamental rule/law is: The current MUST return to the source.
Do not split the screen from the core and take one on a route that increases the loop area between core and screen. Any loop will act as an aerial and will pick up interference. The bigger the loop area, the bigger the interference.
A coaxial cable is a signal connector that uses the outer for two functions. as a signal return and as a screen to attenuate interference on the core. But to be effective the screen must be connected to a low impedance sink.
That applies to the source to receiver link and the receiver to speaker link.
You must NOT break one of those connections.
A shield is an extension of the enclosure. It can be connected at one end, or both ends.
A shield that is connected at one end only does not give as much attenuation of interference. For effective screening at very high frequencies one should connect the shield at both ends to the enclosure.
Our WiFi, mobile phones and computers are very high frequency polluters.
Where one uses a screened single core cable to make a signal connection, both parts (the screen and the core) must be connected at BOTH ends. The screen is the return route for the signal. The fundamental rule/law is: The current MUST return to the source.
Do not split the screen from the core and take one on a route that increases the loop area between core and screen. Any loop will act as an aerial and will pick up interference. The bigger the loop area, the bigger the interference.
A coaxial cable is a signal connector that uses the outer for two functions. as a signal return and as a screen to attenuate interference on the core. But to be effective the screen must be connected to a low impedance sink.
Last edited:
The shell of an RCA plug/jack MUST connect to Protective Earth *somehow*.
It usually do not connect *direct*. The RCA goes to the amplifier. The amplifier gets power from a power supply. The power supply, or some similar node, is bonded to the metal chassis. The touchable metal must go to Protective Earth.
Except indoors on dry wood floors with no inspector around, some hi-fi systems play hum-free without any connection to Earth/Ground. That was common in the US from the beginning even into the 1990s. Fine as long as there is NO insulation breakdown which might cause the case to become "live" relative to radiators, sinks, concrete, etc.
It usually do not connect *direct*. The RCA goes to the amplifier. The amplifier gets power from a power supply. The power supply, or some similar node, is bonded to the metal chassis. The touchable metal must go to Protective Earth.
Except indoors on dry wood floors with no inspector around, some hi-fi systems play hum-free without any connection to Earth/Ground. That was common in the US from the beginning even into the 1990s. Fine as long as there is NO insulation breakdown which might cause the case to become "live" relative to radiators, sinks, concrete, etc.
Hi all
I did some experiments here to clarify some aspects of this ground loop between PC and amp (Sansui AU-2900)...
1 - I took the USB DAC (Edirol UA-25) - off the PC which was causing ground loop hum, and ran it from a 2-pin powered laptop. No ground loop hum. --It's as though running two devices in tandem which both have 3-pin wired ground is the source of the problem here--. The PC PSU is a new-ish Seasonic - it should be good quality.
2 - I put the PC's plug next to the amp's plug on the same power-strip. I also used a short RCA lead between UA-25 and amp. The hum was still there - maybe slightly less (but that could be an illusion).
3 - I checked the condition and polarity of the amp's power plug - all ok.
Could anyone please suggest a way to remove a ground loop without compromising or degrading sound quality? I'm drawing a blank here.
It seems that running two units together which both have 3-pin power/wired ground is the culprit here.
If I can't solve this I may have to get another amp (specifically 2-pin, class-II) which is a shame because this AU-2900 is very good.
Thanks
John
I did some experiments here to clarify some aspects of this ground loop between PC and amp (Sansui AU-2900)...
1 - I took the USB DAC (Edirol UA-25) - off the PC which was causing ground loop hum, and ran it from a 2-pin powered laptop. No ground loop hum. --It's as though running two devices in tandem which both have 3-pin wired ground is the source of the problem here--. The PC PSU is a new-ish Seasonic - it should be good quality.
2 - I put the PC's plug next to the amp's plug on the same power-strip. I also used a short RCA lead between UA-25 and amp. The hum was still there - maybe slightly less (but that could be an illusion).
3 - I checked the condition and polarity of the amp's power plug - all ok.
Could anyone please suggest a way to remove a ground loop without compromising or degrading sound quality? I'm drawing a blank here.
It seems that running two units together which both have 3-pin power/wired ground is the culprit here.
If I can't solve this I may have to get another amp (specifically 2-pin, class-II) which is a shame because this AU-2900 is very good.
Thanks
John
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.