thank you. This slang term is probably not the most common in the english language and thus it was new for me.
Google provides except this thread this URL:
Understanding Star Grounding
Dynaco-Doctor Forum
by follow topic it was call "Ground loop Isolator":
http://www.diyaudio.com/forums/chip-amps/186538-gainclone-direct-coupled-appropriate-preamp-4.html
What happens, if the diodes from the pic in post #31 about
http://www.diyaudio.com/forums/chip-amps/186538-gainclone-direct-coupled-appropriate-preamp-4.html
are not in use, and for the cap only a very small 63V version and for the resistor only 1/8 watts?
If there is a failure, the resistor fired up and the cap explodes (this I observe in several cases - very often by both first Cyrus integrated amps).
But I don't understand, why now the safety requirements are violated. The mains earth is still connected with the metal envelope. Only the PCB-GND isn't now connected with the metal envelope, which serves as a shield against electrical interference.
The only unwanted effect is an audible hum. Isn't it?
Also by shorting of that cap. Now one hear a GND loop hum. Also in this case the mains earth is still connected.
What do I overlooked?
Nevertheless the diodes in this topology seems to be a good solution; until now I know this diode topology by using for a DC Filter - go to
http://www.diyaudio.com/forums/solid-state/2080-dc-filter.html
Last edited:
The FAULT current cannot flow from Audio Ground to Chassis and thence to PE, because Tief has blocked the route with a 100r resistor.
That resistor will limit the Fault current to ~230Vac/100r ~ 2.3Aac until it blows up. The fuse is unlikely to blow and the Audio Ground will now be sitting at 230Vac until someone touches the speaker terminal or the input terminal or some other bit of connected equipment that is not "earthed".
Once the Fault current route has broken, the equipment becomes a death trap.
Tief should never have posted that schematic on a public Forum. Tief is behaving irresponsibly by leaving the dangerous schematic on this Forum. He was told, but he did not accept advice and here we are thirteen days later and still the dangerous schematic is showing for fools waiting to copy and kill themselves.
That resistor will limit the Fault current to ~230Vac/100r ~ 2.3Aac until it blows up. The fuse is unlikely to blow and the Audio Ground will now be sitting at 230Vac until someone touches the speaker terminal or the input terminal or some other bit of connected equipment that is not "earthed".
Once the Fault current route has broken, the equipment becomes a death trap.
Tief should never have posted that schematic on a public Forum. Tief is behaving irresponsibly by leaving the dangerous schematic on this Forum. He was told, but he did not accept advice and here we are thirteen days later and still the dangerous schematic is showing for fools waiting to copy and kill themselves.
This has NOTHING in common with the DC blocker that can reduce transformer hum due to non symmetric AC waveforms.
This is all very exaggerated. A straight connection from the IEC sockets PE pin to the chassis and from there a 10 Ohm/5W resistor paralleled with 100nF 275V X2 rated capacitor connected to audio GND will do the job as mentioned before. Don't use 100 Ohm. And certainly do not connect audio GND to PE directly if you like your equipment to sound good.
Last edited:
Why you don't understand, that the circuit from attachement in my post #1 comes not from me ??? - go to
http://www.diyaudio.com/forums/chip-amps/64200-decibel-dungeon-gc-w-lm3875.html
From there I have upload that file.
BTW - I have inform the moderator "SY" from AUSTIN/Texas and ask for instructions to remove this attachment.
This I know.
Nevertheless - the topology from diodes of both (DC blocker and the so called "Disconnecting Network") is the same, even if the purpose is completely different.
Last edited:
Don't bother. The drawing is good for the discussion about possible side effects. Please note that regulations differ in various european countries. Here devices under 5 VA don't need to be earthed, in Germany 230V AC wiring may be inside plastering without PVC piping which is forbidden here etc. etc. One union but not in electricity. This while we are all connected to the european grid. It seems that in countries where building reliable equipment is difficult most attention is paid to safety regulations 
Just kidding.
Metal chassis should be connected to PE. That is where the primary concern is. Ground Fault Circuit interruptors will do their job in the aforementioned unrealistic "death trap" scenario. PE was invented to protect us from touching live conductors/wiring/cases under voltage.
Just kidding.Metal chassis should be connected to PE. That is where the primary concern is. Ground Fault Circuit interruptors will do their job in the aforementioned unrealistic "death trap" scenario. PE was invented to protect us from touching live conductors/wiring/cases under voltage.
Last edited:
Several URLs I have found in this matter by the term "safety earth"
Earthing Your Hi-Fi - Tricks and Techniques
Earthing (Grounding) Your Hi-Fi - Tricks and Techniques
Resistor-GND
http://www.diyaudio.com/forums/chip-amps/112278-resistor-ground.html
Safety Question Concerning GND Loops
http://www.diyaudio.com/forums/chip-amps/81156-safety-question-concerning-ground-loops.html
Poweramp Starground and mains earth
http://www.diyaudio.com/forums/chip-amps/61900-poweramp-starground-mains-earth.html#post3068312
connecting audio ground to Safety Earth
http://www.diyaudio.com/forums/power-supplies/112359-connecting-audio-ground-safety-earth.html
Testing low current diodes for fault current survivability
http://www.diyaudio.com/forums/chip-amps/92673-yet-another-3886-gain-clone-5.html#post1357794
Ground Loops Tech Notes
Ground Loops
Breaking The Loop
Solving Computer Audio Problems
GND/Earth Wiring Management by Power Amplifiers - what is the royal Way
http://www.diyaudio.com/forums/soli...nagement-power-amplifiers-what-royal-way.html
http://valvewizard2.webs.com/Grounding.pdf
http://blueguitar.org/new/articles/other/ground_loop.pdf
Balanced Connectios by Audio (Jensen)
http://www.aes.org/sections/pnw/pnwrecaps/2005/whitlock/whitlock_pnw05.pdf
Star-earthing and "3rd party" Naim power supplies
Star-earthing and "3rd party" Naim power supplies - pink fish media
Mains and Chasis Star Grounding in AKSA 100?
Mains and Chasis Star Grounding in AKSA 100?
Hum on Avondale NCC200 derivative amp
http://www.diyaudio.com/forums/solid-state/212935-hum-avondale-ncc200-derivative-amp.html
Starground and mains earth at Gainclone
Building a Gainclone chip amp power supply.
Earthing Your Hi-Fi - Tricks and Techniques
Earthing (Grounding) Your Hi-Fi - Tricks and Techniques
Resistor-GND
http://www.diyaudio.com/forums/chip-amps/112278-resistor-ground.html
Safety Question Concerning GND Loops
http://www.diyaudio.com/forums/chip-amps/81156-safety-question-concerning-ground-loops.html
Poweramp Starground and mains earth
http://www.diyaudio.com/forums/chip-amps/61900-poweramp-starground-mains-earth.html#post3068312
connecting audio ground to Safety Earth
http://www.diyaudio.com/forums/power-supplies/112359-connecting-audio-ground-safety-earth.html
Testing low current diodes for fault current survivability
http://www.diyaudio.com/forums/chip-amps/92673-yet-another-3886-gain-clone-5.html#post1357794
Ground Loops Tech Notes
Ground Loops
Breaking The Loop
Solving Computer Audio Problems
GND/Earth Wiring Management by Power Amplifiers - what is the royal Way
http://www.diyaudio.com/forums/soli...nagement-power-amplifiers-what-royal-way.html
http://valvewizard2.webs.com/Grounding.pdf
http://blueguitar.org/new/articles/other/ground_loop.pdf
Balanced Connectios by Audio (Jensen)
http://www.aes.org/sections/pnw/pnwrecaps/2005/whitlock/whitlock_pnw05.pdf
Star-earthing and "3rd party" Naim power supplies
Star-earthing and "3rd party" Naim power supplies - pink fish media
Mains and Chasis Star Grounding in AKSA 100?
Mains and Chasis Star Grounding in AKSA 100?
Hum on Avondale NCC200 derivative amp
http://www.diyaudio.com/forums/solid-state/212935-hum-avondale-ncc200-derivative-amp.html
Starground and mains earth at Gainclone
Building a Gainclone chip amp power supply.
Checkout this also
http://www.diyaudio.com/forums/solid-state/211977-quick-guide-grounding.html
http://www.diyaudio.com/forums/solid-state/211977-quick-guide-grounding.html
Thank you for this URL. Because there are many different headlines to the same topic, there must actually be a lot of more threads here on diyaudio than the previously mentioned from the last two posts.
Here are another examples.
http://www.diyaudio.com/forums/solid-state/99567-what-means-ground-lift.html
http://www.diyaudio.com/forums/chip-amps/186538-gainclone-direct-coupled-appropriate-preamp.html
http://www.diyhifi.org/forums/viewtopic.php?f=11&t=2003&start=15
http://www.diyaudio.com/forums/chip-amps/118735-gainclone-hum-thump-problems-solution.html
http://www.diyaudio.com/forums/solid-state/122202-power-amp-build-bits-guidance-required.html
Last edited:
Note that I said use a 10 ohm 5W resistor. For a 230V supply, this will allow 23A to pass. The resistor should survive long enough for a fuse to rupture or a breaker to trip. If you want to be sure, a 10R NTC would be even better as it would decrease in resistance the longer the current flows.
The ground terminal of the power connector should always go direct to chassis. The ground loop breaker should be connected between the chassis ground point and the PSU 0V star.
The ground terminal of the power connector should always go direct to chassis. The ground loop breaker should be connected between the chassis ground point and the PSU 0V star.
"Note that I said use a 10 ohm 5W resistor. For a 230V supply, this will allow 23A to pass. The resistor should survive long enough for a fuse to rupture or a breaker to trip. If you want to be sure, a 10R NTC would be even better as it would decrease in resistance the longer the current flows."
I think this is very poor practice.
The connection from the chassis safety ground to the supply ground star is to lower noise, no real current should ever flow through it. A pair of inverse-parallel connected diodes bypassed by a parallel resistor and parallel cap offer the best noise performance and safety. If one of the other pieces connected to the unit in question has a ground problem it will buzz like heck, letting you know that you have a problem to run down.
The diodes are best be ones rated at 400A surge, and ordinary (not fast type) meet this requirement. If the fuse/circuit-breaker is not fast enough on a fault, the diode will fail as a short, so you still have protection.
23A will cause the 10Ω/5W resistor to dissipate about 5.3KW, and possibly burst into flames and maybe produce shrapnel.
I think this is very poor practice.
The connection from the chassis safety ground to the supply ground star is to lower noise, no real current should ever flow through it. A pair of inverse-parallel connected diodes bypassed by a parallel resistor and parallel cap offer the best noise performance and safety. If one of the other pieces connected to the unit in question has a ground problem it will buzz like heck, letting you know that you have a problem to run down.
The diodes are best be ones rated at 400A surge, and ordinary (not fast type) meet this requirement. If the fuse/circuit-breaker is not fast enough on a fault, the diode will fail as a short, so you still have protection.
23A will cause the 10Ω/5W resistor to dissipate about 5.3KW, and possibly burst into flames and maybe produce shrapnel.
In truth 23A would not flow, though. The primary side fuse is going to be what, 2A at the most in a domestic amp ? Even a time-delay type would blow quickly.
Agreed, if this were a PA amp where you might have a 1000VA transformer or so with a large primary fuse, and likely not running from a circuit protected by an RCD, I would use the diodes.
In the UK we have extra protection as our wall plugs have their own fuses. A typical IEC cable is protected by a 13A fuse in the plug. A primary-secondary short with a 10 ohm resistor would blow this very quickly.
Agreed, if this were a PA amp where you might have a 1000VA transformer or so with a large primary fuse, and likely not running from a circuit protected by an RCD, I would use the diodes.
In the UK we have extra protection as our wall plugs have their own fuses. A typical IEC cable is protected by a 13A fuse in the plug. A primary-secondary short with a 10 ohm resistor would blow this very quickly.
Most fuses will stand 200% of rated current for two minutes, so that 13A fuse may not blow as fast as you think. You are also assuming a short to ground, and near the top of the winding, none of which should be assumed.
Fast-Acting Ceramic Tube Fuses
http://www.cooperindustries.com/content/dam/public/bussmann/Electrical/Resources/Data%20Sheets/Bus_Ele_DS_2000_ABC_ABC-V_Series.pdf
Fast-Acting Ceramic Tube Fuses
http://www.cooperindustries.com/content/dam/public/bussmann/Electrical/Resources/Data%20Sheets/Bus_Ele_DS_2000_ABC_ABC-V_Series.pdf
Last edited:
I want to know more exactly informations arround this topic. E. g. the fact, from where comes the high current flow (i. e. the potential difference between the mains earth and the audio signal GND) under certain conditions through the so called "Disconnected Network" in usual domestic environments.
By several service work I must replace such devices, but after replace this parts, it's never happened that they blown away once again.
Thus it is impossible for me to determine an exact cause.
Therefore the same question as in one of a previous post from me:
What is the most usual english term for "Disconnecting Network" ??
The provided results from Google by this term is too low.
By several service work I must replace such devices, but after replace this parts, it's never happened that they blown away once again.
Thus it is impossible for me to determine an exact cause.
Therefore the same question as in one of a previous post from me:
What is the most usual english term for "Disconnecting Network" ??
The provided results from Google by this term is too low.
Last edited:
Just as a hypothetical worst case example, suppose the Live wire feeding the transformer pops out of it's screw down terminal block, or the stranded wire just breaks.
The loose end springs away and briefly flicks past an audio connection, blowing up the audio stage and finally lands on the Main Audio Ground wire attached to the speaker terminal.
Many of the audio side components are burnt out, some have turned into a carbon mess. All the audio grounds are at LIVE voltage and the mains fuse has not blown yet.
The mains wiring impedance from Distribution board to amplifier and back again is ~0.2ohms.
The chassis impedance from LIVE to Earth is ~0.01ohms.
The voltage applied across these is 240Vac. A current of ~1142Aac starts to pass from LIVE to PE. The Mains fuse blows in <1ms.
The operator/user is exposed to danger for just that 1ms - IF THEY HAPPEN TO BE TOUCHING THE FAULTY APPLIANCE.
But they are not exposed to the full 240Vac. The faulty appliance is located roughly half way around the distribution route from LIVE to Neutral at the distribution board. At the very worst the appliance will rise to ~120Vac for that <1ms till the fuse blows. In the UK they is a further protective rule : the faulty appliance and the distribution wiring must limit the voltage on the LIVE appliance during the fault incident. I think this limited voltage is 50Vac "by design"
The design of the house wiring installation and the handover test must both show that the limiting voltage will not be exceeded.
Now, let's put a 10r resistor in that fault current return route.
The fuse could now take many hundreds of milliseconds to blow, or in some worst combination of cases even many seconds to blow.
The user is now exposed to a higher voltage and for much longer.
Now the really bad situation:- the potentially fatal version of events.
The fault occurs, the 10r resistor burns out while trying to dissipate those many kW of power. The fuse does not blow !!!! The voltage on the audio Ground is now 240Vac. The music system sounds quiet. The user gets up from his/her seat and fiddles with a control knob or checks the speaker cable or "POOF" they are gone !!!!!!!
The loose end springs away and briefly flicks past an audio connection, blowing up the audio stage and finally lands on the Main Audio Ground wire attached to the speaker terminal.
Many of the audio side components are burnt out, some have turned into a carbon mess. All the audio grounds are at LIVE voltage and the mains fuse has not blown yet.
The mains wiring impedance from Distribution board to amplifier and back again is ~0.2ohms.
The chassis impedance from LIVE to Earth is ~0.01ohms.
The voltage applied across these is 240Vac. A current of ~1142Aac starts to pass from LIVE to PE. The Mains fuse blows in <1ms.
The operator/user is exposed to danger for just that 1ms - IF THEY HAPPEN TO BE TOUCHING THE FAULTY APPLIANCE.
But they are not exposed to the full 240Vac. The faulty appliance is located roughly half way around the distribution route from LIVE to Neutral at the distribution board. At the very worst the appliance will rise to ~120Vac for that <1ms till the fuse blows. In the UK they is a further protective rule : the faulty appliance and the distribution wiring must limit the voltage on the LIVE appliance during the fault incident. I think this limited voltage is 50Vac "by design"
The design of the house wiring installation and the handover test must both show that the limiting voltage will not be exceeded.
Now, let's put a 10r resistor in that fault current return route.
The fuse could now take many hundreds of milliseconds to blow, or in some worst combination of cases even many seconds to blow.
The user is now exposed to a higher voltage and for much longer.
Now the really bad situation:- the potentially fatal version of events.
The fault occurs, the 10r resistor burns out while trying to dissipate those many kW of power. The fuse does not blow !!!! The voltage on the audio Ground is now 240Vac. The music system sounds quiet. The user gets up from his/her seat and fiddles with a control knob or checks the speaker cable or "POOF" they are gone !!!!!!!
Last edited:
A slow-blow type may pass 200% for one hour, much more over shorter periods of time. I used to use a pair of TL30 to fuse a distro, it could pass 120A of lights and sound for a live stage-show (one hour sets).
Guitar amps with hum switches (line phase-reversal) are a real problem, I've had them send 120V down a snake, through the FOH gear, back up the snake, and blow the input opamp in the electronic crossover. Designing a line level piece to handle 120V on the input is a chore.
Guitar amps with hum switches (line phase-reversal) are a real problem, I've had them send 120V down a snake, through the FOH gear, back up the snake, and blow the input opamp in the electronic crossover. Designing a line level piece to handle 120V on the input is a chore.
have a look to that 10 ohm resistor in the Cyrus ONE - go to the second picture from post #31 about
http://www.diyaudio.com/forums/solid-state/82499-cyrus-one-4.html
search for R121 (in front of the black potentiometer shaft).
It seems to be only a 0,125W resistor.
There are no additional components like diodes or caps on the solder side of the main PCB.
Hafler (and many others) have used resistors (without the diodes) in the range of 0R5~10R/1/4W in this position. They burn up instantly in the event of a fault.
Rockford Fosgate (parent company of Hafler) ceased using the 10R/1/4W in this position as they burned up so frequently, the now malfunctioning amplifier needing a warranty repair. As a consequence of replacing this resistor with a wire link the ground trace in the device driving the amplifier now vaporizes from carrying the fault current down the shield. This moved the repair cost from the amplifier to the driving device.
Rockford Fosgate (parent company of Hafler) ceased using the 10R/1/4W in this position as they burned up so frequently, the now malfunctioning amplifier needing a warranty repair. As a consequence of replacing this resistor with a wire link the ground trace in the device driving the amplifier now vaporizes from carrying the fault current down the shield. This moved the repair cost from the amplifier to the driving device.