OK, I have Joffe's article open and looking at fig4 and fig5.
The external loop in the interconnect ground has R13.
the internal return circuit has R11 in the right channel signal return and R12 in the left channel signal return. Fig4 without the solution has this input loop linked to Power Ground via an inductive interference source. L1 in the right channel and L3 in the left channel.
As shown interference current passes around the interconnect return loop.
Interference current (Ii) times R11 generates an interference voltage Vi in the signal circuit. Similarly Ii*R12 for Left.
The way to reduce Vi is to remove the interference voltage and/or reduce Ii.
One can increase R13 to attenuate Ii, or add a new resistor to the loop outside the signal route. This new added resistor is shown in fig5. labeled HBRR (right) & HBRL (left).
This added resistor is NOT in the input signal flow and return route. It reduces Ii. It reduces Vi.
One could increase R13 if one had access to change the grounding scheme inside the Source component. But this is generally not feasible.
Alternatively one could increase R11 and/or R11. This would reduce Ii, since it has increased the loop resistance. The interference voltage on the input circuit will increase because Ii (lower value) * R11(higher value) results in a higher proportion of the interference voltage from L1 appearing along the signal return route. Increasing R11 does not attenuate the interference. Similarly increasing R12 does not attenuate Vi.
Another alternative would be to reduce the proportion of interference voltage in the signal route by reducing R11 and R12. But this route also has reactive impedance. Simply increasing the copper of the return will not give much attenuation. But as H.Ott often shows it does have some effect in attenuating interference.
For this diagram only increasing HBRR or increasing R13 attenuates Vi in the input circuit.
The external loop in the interconnect ground has R13.
the internal return circuit has R11 in the right channel signal return and R12 in the left channel signal return. Fig4 without the solution has this input loop linked to Power Ground via an inductive interference source. L1 in the right channel and L3 in the left channel.
As shown interference current passes around the interconnect return loop.
Interference current (Ii) times R11 generates an interference voltage Vi in the signal circuit. Similarly Ii*R12 for Left.
The way to reduce Vi is to remove the interference voltage and/or reduce Ii.
One can increase R13 to attenuate Ii, or add a new resistor to the loop outside the signal route. This new added resistor is shown in fig5. labeled HBRR (right) & HBRL (left).
This added resistor is NOT in the input signal flow and return route. It reduces Ii. It reduces Vi.
One could increase R13 if one had access to change the grounding scheme inside the Source component. But this is generally not feasible.
Alternatively one could increase R11 and/or R11. This would reduce Ii, since it has increased the loop resistance. The interference voltage on the input circuit will increase because Ii (lower value) * R11(higher value) results in a higher proportion of the interference voltage from L1 appearing along the signal return route. Increasing R11 does not attenuate the interference. Similarly increasing R12 does not attenuate Vi.
Another alternative would be to reduce the proportion of interference voltage in the signal route by reducing R11 and R12. But this route also has reactive impedance. Simply increasing the copper of the return will not give much attenuation. But as H.Ott often shows it does have some effect in attenuating interference.
For this diagram only increasing HBRR or increasing R13 attenuates Vi in the input circuit.
Last edited:
Moving the added resistor from one side of the Signal Ground Node @ junction of R9,10,12 to the Power Ground side of the same Node has changed the way the interference attenuation works, or does not work. The location of the added resistor makes a difference to the attempt at attenuation of the interference.
Last edited:
Thanks Bonsai,
I will adhere to your suggestion. I do not have problems with ground loops.
I lived in a few places and even in an old building with crappy power wirings and a big power hungry company nearby.
I had never ground loops problems. My home stereo system is powered from the same power outlet.
Only the antenna cable is galvanically isolated (otherwise the roof antenna RF-amplifier would introduce a really big ground loop).
The components are all Class-II. Some older Amps are Class-I.
I think there are at least two different mechanism involved.
1.) Different ground potentials due to wire resistance
(e.g. if a power amp needs 1 A, its ground reference jumps by a few mV).
2.) Loop currents due to magnetic induction.
Class-II or your Ground Lift Switch helps with 1.
Single ended designs can not cope with 2,
even if the current is not flowing through the internal ground
(your 15R helps here).
Minimizing the cable ground loop and a solid separate ground connection (as sometimes done on Phono inputs) helps a lot.
Greetings, Udo
Hey bentwang, no one is in this thread to read your opinion on anything, I think it's been well established that you're crazy as hell. What's it been, 2 months and you can't figure out how to put together a chipamp but you want to tell these guys how to lay out an amplifier?
Last edited:
Now that you've gotten that ridiculous thought off you chest, look at what you said:I hilited in red what I was referring to.
An AC voltage does not couple to another wire when they are twisted, unless you count tens of picofarads. The discussion is about current.
I do not engage in debating tricks, so get off that horse.
jn
I caution all against using 25 amp bridge rectifiers as safety devices. They are not rated for bolted fault.
Three things can happen to a bridge under bolted fault conditions.
1. The bridge survives and protects the user.
2. The bridge fails, but since it has the silicon chips at a right angle to the mounting surface, the aluminum case it is encapsulated in retains it's physical integrity, therefore the silicon will fail into a dead short. This will kill the ground lift capability, but will protect the user from lethal ground voltages.
3. The bridge fails, but because the silicon is oriented parallel to the mounting surface, there is only the encapsulation epoxy to prevent physical destruction. Since all encapsulation epoxies which have a thermal coefficient of expansion close to aluminum will be filled with ground alumina powder, the epoxy will NOT have good tensile strength, it will fail and the device will suffer a containment breach. This failure leaves the bridge in an open condition. Should the fail to trip the panel breaker, there will be a lethal condition remaining.
Absolute agreement.
from my gallery:
ground_loop_calculation - My Photo Gallery
jn
Wow. A truly amazing new theory. It does make me wonder how a transformer works, since a twist is a coil longitudinally extended. Well, at least there's no need to worry about loop area anymore.
.
Last edited:
i was commenting on the Bonsai article which Leach used more than 20 years ago and which i also implement.....if you say my post is dangerous, then Bonsai must be wrong....is that what you are saying?
i stand by what i posted, as long as you know which wire is "neutral" and which one is "line", then color is not an issue to me...
I think in quite a few cases here, we are assuming that there is no fuse in the equipment, or that the mains breaker does not work. You have to assume these things are in place and working correctly.
I am able to build very quiet systems with the techniques shown. For those of you that don't like using a ground lifter . . . Don't. In the practical wiring diagram, I show a version without the ground lifter.
I am able to build very quiet systems with the techniques shown. For those of you that don't like using a ground lifter . . . Don't. In the practical wiring diagram, I show a version without the ground lifter.
Pictures versus words
As a person who makes his living out of manipulating words as well as a non-technical amateur, it is my experience that words and sentences can be misinterpreted, and many writers do not practice "plain language."
A properly drawn, accurately and completely labelled drawing/diagram/schematic is much better for amateurs such as I. Professor Leach's articles in Audio were read repeatedly by me and I never really understood them the way that i do with the advent of the internet and the help of the many high-quality contributors to DIYaudio. The best writing is accompanied by good diagrams and schematics, well labelled, and presented in a progression from simple to complex. Nelson Pass' writing and his diagrams are models, in this regard.
Geary
As a person who makes his living out of manipulating words as well as a non-technical amateur, it is my experience that words and sentences can be misinterpreted, and many writers do not practice "plain language."
A properly drawn, accurately and completely labelled drawing/diagram/schematic is much better for amateurs such as I. Professor Leach's articles in Audio were read repeatedly by me and I never really understood them the way that i do with the advent of the internet and the help of the many high-quality contributors to DIYaudio. The best writing is accompanied by good diagrams and schematics, well labelled, and presented in a progression from simple to complex. Nelson Pass' writing and his diagrams are models, in this regard.
Geary
Hi Geary, you are absolutely correct.....
one important point that hasn't been mentioned explicitly here, although obvious in the schematic and picture diagrams:
never use the chassis metal work as return path for power, a separate power return lead must be used....
in the tube radio era, you will see component grounds returned to chassis, some were even welded to the chassis,
and the reason for this is because of the very low currents used back then...
in today's solid state era, where currents can be in the tens of amperes, this practice will never work...
one important point that hasn't been mentioned explicitly here, although obvious in the schematic and picture diagrams:
never use the chassis metal work as return path for power, a separate power return lead must be used....
in the tube radio era, you will see component grounds returned to chassis, some were even welded to the chassis,
and the reason for this is because of the very low currents used back then...
in today's solid state era, where currents can be in the tens of amperes, this practice will never work...
Note very carefully the words used by jneutron.
You are confusing current and voltage. The coupling mechanisms are different.
This is a complex field - there is much to learn from the experts here.
- Status
- Not open for further replies.