Andrew,
Maybe I misunderstand something completely, but as far as I can see, what you are saying is simply that the connection, by whatever means, between audio ground and safety ground must keep connection long enough for the mains fuse to trigger. That is how I understand you, and I fully agree with that.
If the connection is direct, there isn't much that can fail. Most likely some part of the audio ground would break open first (a thin wire or a PCB track), but that is about as good as we can make it.
Now, if we insert some circuit inbetween audio and safety ground, then this circuit must live long enough to trigger the fuse. That is requirement of this circuit, but it is not the reason we put it there. usually this circuit is a small resistor, or a small resistor in parallel with two diodes, with the intention of being resistive enough to break hum loops, but sufficiently condcutive to trigger the mains fuse (if surviving long enough to achieve that).
Maybe I misunderstand something completely, but as far as I can see, what you are saying is simply that the connection, by whatever means, between audio ground and safety ground must keep connection long enough for the mains fuse to trigger. That is how I understand you, and I fully agree with that.
If the connection is direct, there isn't much that can fail. Most likely some part of the audio ground would break open first (a thin wire or a PCB track), but that is about as good as we can make it.
Now, if we insert some circuit inbetween audio and safety ground, then this circuit must live long enough to trigger the fuse. That is requirement of this circuit, but it is not the reason we put it there. usually this circuit is a small resistor, or a small resistor in parallel with two diodes, with the intention of being resistive enough to break hum loops, but sufficiently condcutive to trigger the mains fuse (if surviving long enough to achieve that).
Hi,
for audio purposes we do not need ANY connection between Safety Earth and audio ground.
The ONLY reason for connecting audio ground to safety earth is to make the amp safe during a fault condition that could make the "amplifier" LIVE.
If you make a direct connection between the Safety Earth and the audio ground then you open the door to hum loops.
We do our best to keep the hum loop open, but we are forced for safety reasons to make the disconnecting network link.
for audio purposes we do not need ANY connection between Safety Earth and audio ground.
The ONLY reason for connecting audio ground to safety earth is to make the amp safe during a fault condition that could make the "amplifier" LIVE.
If you make a direct connection between the Safety Earth and the audio ground then you open the door to hum loops.
We do our best to keep the hum loop open, but we are forced for safety reasons to make the disconnecting network link.
Andrew,
OK, I think I understand how you mean now. What we both said seems consistent with each other then.
OK, I think I understand how you mean now. What we both said seems consistent with each other then.
O.K., I hope I got this right -
Christer -
“Now, if we insert some circuit in-between audio and safety ground, then this circuit must live long enough to trigger the fuse.”
Andrew –
“We do our best to keep the hum loop open, but we are forced for safety reasons to make the disconnecting network link”.
Would this connecting link/circuit between audio ground to safety be a short wire/tag off an insulated audio star collecting point to chassis? If this is the case would it be two (inverted) diodes in parallel with a resistor? And if in fact this could be the case would (2) 1N4001’s and a 100 ohm ½ watt resistor do the trick??
Stan
Christer -
“Now, if we insert some circuit in-between audio and safety ground, then this circuit must live long enough to trigger the fuse.”
Andrew –
“We do our best to keep the hum loop open, but we are forced for safety reasons to make the disconnecting network link”.
Would this connecting link/circuit between audio ground to safety be a short wire/tag off an insulated audio star collecting point to chassis? If this is the case would it be two (inverted) diodes in parallel with a resistor? And if in fact this could be the case would (2) 1N4001’s and a 100 ohm ½ watt resistor do the trick??
Stan
Hi,
I have seen various suggestions and I now give the following advice.
The disconnecting network can consist of some or all of the following connected in parallel:-
Power resistor.
Power Thermistor.
Power Bridge diode connected in inverse parallel.
High frequency capacitor.
I choose to omit the Thermistor and downgrade the resistors (to 1.2W) due to the kA capacity (for a few mS) of the paralleled power diode assembly.
I hope I have got if safe and never have to test it.
I have seen various suggestions and I now give the following advice.
The disconnecting network can consist of some or all of the following connected in parallel:-
Power resistor.
Power Thermistor.
Power Bridge diode connected in inverse parallel.
High frequency capacitor.
I choose to omit the Thermistor and downgrade the resistors (to 1.2W) due to the kA capacity (for a few mS) of the paralleled power diode assembly.
I hope I have got if safe and never have to test it.
AndrewT said:
What about the reason of shielding the audio circuitry via the enclosure?
AndrewT said:
Hum loops aren't guaranteed by doing this. A little effort can prevent or minimize loop problems.
Hi Pooge,
I am moving out of my area of experience here so please correct me if I get this twisted.
SCREENING
The chassis and the Safety Earth must be connected in mains operated equipment. We have no choice in that.
The amplifiers that we build do not have a need for a screening shield it appears. I base this statement on my experience of testing amplifier set-ups with open cases and without any cases and none suffer from interference in this state. I do not believe my house has any more or less electromagnetic/static than any other nomal house. One area I have not tested is what happens when I am arc-welding in the workshop when the stereo is on! That is extreme but I can't hear the stereo then, they are either side of an intervening room.
An open case is just as bad as no case, for the field simply wraps around and washes over the circuitry anyway ( it seems to work much the same as bass frequencies being omnidirectional around the box and corners etc). How many times have you taken the lid off a case and listened while you look/probe/etc?
If there is a particularly sensitive circuit that needs shielding one often finds a separate shield around that circuit. I think the best place to reference that local shield is the ground for that particular circuit, not the chassis.
I have seen some documents detailing the extent of shielding necessary to be effective and the enclosure becomes 100% and ALL penetrations have to be detailed to prevent EMI getting in (or out).
Our perforated/ventilated/holy enclosures are at best poor shields and I suspect they are actually very poor as a shield. I further suspect they are a very effective transmitting aerial for all the mains borne interference that must be coupled in via the Safety Earth. A perfect enclosing aerial would not interfere with the items inside (they would be bombarded equally from all directions) but our typical enclosures are far from perfect so they do interfere with the circuits mounted inside. So we end up trying to make our circuits resistant to the enclosure and not vice versa. (now, there's a reversal of roles! What do I hear?)
But, let's move inside the box. The worst offenders are probably the transformer and the mains incomer. We put our amplifier right in there beside these interference producers and the amplifiers continue to work. Yes, we can orient the components to lessen some audible effects (mostly hum) but the two halves co-exist.
HUM LOOPS
The amplifier is always connected to a source and for this argument both are mains powered.
I live in the UK where three wire connections are always provided at the socket outlet.
I have used double insulated equipment but I am not competent to design nor build such equipment.
That further restricts us to discuss the connection of conventional three wire source and amplifier.
Let's take the simple situation of both units using the "safe" way, bolting the Safety Earth to chassis AND to audio ground.
Now plug in an RCA interconnect.
Both audio grounds are connected via the second wire in the interconnect.
Both Safety Earths are connected via the third wire back to the distrbution board.
We have a hum loop.
Wiring two units with the audio ground directly connected to safety earth (the "safe" way) guarantees a hum loop.
There are many that seem unaware of this and manufacture audio equipment the "safe" way.
If one wires the amplifier the same easy way then eventually one other source will cause a problem. It does not matter that the offending source is at the remote end of a long string of other units/interconnects. It just needs two returns to the distribution board and the second wire between audio ground and the problem is there.
Wire up the amplifier correctly with the audio ground separate from the Safety Earth and the LOOP cannot be made through the amplifier third wire. There is still the possibility that two sources could be mis-wired (the "safe" way). and then one has to resort to modification or isolation.
Don't wire your amplifier or pre-amplifier with the Safety Earth connected to Audio Ground.
I am moving out of my area of experience here so please correct me if I get this twisted.
SCREENING
The chassis and the Safety Earth must be connected in mains operated equipment. We have no choice in that.
The amplifiers that we build do not have a need for a screening shield it appears. I base this statement on my experience of testing amplifier set-ups with open cases and without any cases and none suffer from interference in this state. I do not believe my house has any more or less electromagnetic/static than any other nomal house. One area I have not tested is what happens when I am arc-welding in the workshop when the stereo is on! That is extreme but I can't hear the stereo then, they are either side of an intervening room.
An open case is just as bad as no case, for the field simply wraps around and washes over the circuitry anyway ( it seems to work much the same as bass frequencies being omnidirectional around the box and corners etc). How many times have you taken the lid off a case and listened while you look/probe/etc?
If there is a particularly sensitive circuit that needs shielding one often finds a separate shield around that circuit. I think the best place to reference that local shield is the ground for that particular circuit, not the chassis.
I have seen some documents detailing the extent of shielding necessary to be effective and the enclosure becomes 100% and ALL penetrations have to be detailed to prevent EMI getting in (or out).
Our perforated/ventilated/holy enclosures are at best poor shields and I suspect they are actually very poor as a shield. I further suspect they are a very effective transmitting aerial for all the mains borne interference that must be coupled in via the Safety Earth. A perfect enclosing aerial would not interfere with the items inside (they would be bombarded equally from all directions) but our typical enclosures are far from perfect so they do interfere with the circuits mounted inside. So we end up trying to make our circuits resistant to the enclosure and not vice versa. (now, there's a reversal of roles! What do I hear?)
But, let's move inside the box. The worst offenders are probably the transformer and the mains incomer. We put our amplifier right in there beside these interference producers and the amplifiers continue to work. Yes, we can orient the components to lessen some audible effects (mostly hum) but the two halves co-exist.
HUM LOOPS
The amplifier is always connected to a source and for this argument both are mains powered.
I live in the UK where three wire connections are always provided at the socket outlet.
I have used double insulated equipment but I am not competent to design nor build such equipment.
That further restricts us to discuss the connection of conventional three wire source and amplifier.
Let's take the simple situation of both units using the "safe" way, bolting the Safety Earth to chassis AND to audio ground.
Now plug in an RCA interconnect.
Both audio grounds are connected via the second wire in the interconnect.
Both Safety Earths are connected via the third wire back to the distrbution board.
We have a hum loop.
Wiring two units with the audio ground directly connected to safety earth (the "safe" way) guarantees a hum loop.
There are many that seem unaware of this and manufacture audio equipment the "safe" way.
If one wires the amplifier the same easy way then eventually one other source will cause a problem. It does not matter that the offending source is at the remote end of a long string of other units/interconnects. It just needs two returns to the distribution board and the second wire between audio ground and the problem is there.
Wire up the amplifier correctly with the audio ground separate from the Safety Earth and the LOOP cannot be made through the amplifier third wire. There is still the possibility that two sources could be mis-wired (the "safe" way). and then one has to resort to modification or isolation.
Don't wire your amplifier or pre-amplifier with the Safety Earth connected to Audio Ground.
Hi all,
I don't want to dominate this thread.
There must be other contributors just dying to come in and suggest alternatives.
I don't want to dominate this thread.
There must be other contributors just dying to come in and suggest alternatives.
Balanced or Unbalanced Equipment
I've somehow made some confusion in this thread. . .
Balanced equipment and unbalanced equipment are wired
different. The shield of balanced equipment connects the
chassis of all equipment togther. Pin 2 and pin 3 are signal while
pin 1 is the shield that connects to the chassis.
Unbalanced equipment
should not connect all chassis in the chain together.
Then there is three prong equipment vs. two prong equipment.
Then there are country safety codes.
Somebody stop the insanity! ; - )
I am beginning to see why it would be bad to connect
the signal bus ground to the chassis and ulimately the
earth ground. It seems if you are using two prong equipment
that is a floating ground you would not want a direct
connection to a three prong balanced piece of equipment.
You would be mixing and mismatching 'grounding' schemes.
There seems to be much mis-information in the electronics
world about the 'grounding' of equipment.
If you were using a three prong box with a floating signal
ground and something made a short between the signal
ground and chassis. . .nothing bad should happen but maybe
increased humm. If a positive or negative rail were to
touch the chassis. . .it should short to ground. . .if
enough current were flowing. . .it should trip the circuit
breaker or fuse back at the main AC service box. . .
I can see the wisdom of not connecting the signal ground
bus to the 'star ground' point on the chassis. . .but there
seems to be much 'support' for this connection and the
magic around the 'star ground.'
I will begin to redraw for balanced and unbalanced. . .
I've somehow made some confusion in this thread. . .
Balanced equipment and unbalanced equipment are wired
different. The shield of balanced equipment connects the
chassis of all equipment togther. Pin 2 and pin 3 are signal while
pin 1 is the shield that connects to the chassis.
Unbalanced equipment
should not connect all chassis in the chain together.
Then there is three prong equipment vs. two prong equipment.
Then there are country safety codes.
Somebody stop the insanity! ; - )
I am beginning to see why it would be bad to connect
the signal bus ground to the chassis and ulimately the
earth ground. It seems if you are using two prong equipment
that is a floating ground you would not want a direct
connection to a three prong balanced piece of equipment.
You would be mixing and mismatching 'grounding' schemes.
There seems to be much mis-information in the electronics
world about the 'grounding' of equipment.
If you were using a three prong box with a floating signal
ground and something made a short between the signal
ground and chassis. . .nothing bad should happen but maybe
increased humm. If a positive or negative rail were to
touch the chassis. . .it should short to ground. . .if
enough current were flowing. . .it should trip the circuit
breaker or fuse back at the main AC service box. . .
I can see the wisdom of not connecting the signal ground
bus to the 'star ground' point on the chassis. . .but there
seems to be much 'support' for this connection and the
magic around the 'star ground.'
I will begin to redraw for balanced and unbalanced. . .
Andrew,
I'm a little rusty on this, cause I haven't delved into it in a while, and don't have time now. I seem to recall that shielding works most effectively if the circuitry has the same reference potential as the shield. However, this may be more important for the output cable screen from an enclosure, i.e., to maintain screen on cable as an extension of the shielded enclosure.
As to connecting equipment together where the input references are connected to the chassis, Leach has shown a way of minimizing ground loops through the stereo channels by connecting one screen to the other channel at the input.
Also, input transformers are useful for breaking ground loops. A more expensive solution, but nevertheless one that does not
"guarantee" a ground loop if signal ground is connected to chassis ground, even if more than one chassis in the chain is safety grounded.
I'm a little rusty on this, cause I haven't delved into it in a while, and don't have time now. I seem to recall that shielding works most effectively if the circuitry has the same reference potential as the shield. However, this may be more important for the output cable screen from an enclosure, i.e., to maintain screen on cable as an extension of the shielded enclosure.
As to connecting equipment together where the input references are connected to the chassis, Leach has shown a way of minimizing ground loops through the stereo channels by connecting one screen to the other channel at the input.
Also, input transformers are useful for breaking ground loops. A more expensive solution, but nevertheless one that does not
"guarantee" a ground loop if signal ground is connected to chassis ground, even if more than one chassis in the chain is safety grounded.
Gentlemen, thank you for making the complicated topic of grounding more easily understood for we electrically-challenged individuals (meaning me). To this point I've been almost totally confused about safety ground and signal ground so I feel that AndrewT's insistence that they NOT be connected, and the subsequent discussion and drawings, has helped me to understand the differences between them - at least enough for me to confidently move ahead and build my first DIY tube preamp.
- Status
- Not open for further replies.