Hello everyone.
I'm currently designing some amplifier with mains-powered linear transformer supply.
While working on my design, the idea hit my head and I suddenly created Ground Breaker circuit.
Here's the simplified schematic of my design.
In here above schematic,
GND is reference zero-volt of amplifier circuits.
Earth is earth from mains power, connected to chassis of appliance and can be used to shield EMF.
R1 and R2 represents to load of external circuits.
Note that transformer T1 is linear step-down transformer, with power rated up to 56 VA.
Mains power input would be 230VAC 60Hz.
C1 is Class-Y2 rated Film type capacitor. It bypass high frequency noises. Also would be helpful to suppress voltage spikes.
When voltage difference between Earth and GND is small than Forward voltage of Diodes(D1, D2), in general situation, small current pass through R3.
D1 and D2 would bypass high current when appliance encountered to failure. It should have ability to drain hundreds amps continuously. Also it's good to use low Vf diode as possible.
These diodes are critically important to protect user and device itself to safe.
Lastly, D3 is Bi-directional TVS diode, dissipate short-period voltage spike as heat when ESD event happens.
If my design above works as I intended, it will provide noise reduction from mains earth, while also keep user safe when appliance fails.
I'm really curious about performance and safety of this schematic.
Would this design pass any regulation of international or local standards?
It will be thankful to get any kind of advice here.
Thanks in advance.
I'm currently designing some amplifier with mains-powered linear transformer supply.
While working on my design, the idea hit my head and I suddenly created Ground Breaker circuit.
Here's the simplified schematic of my design.
In here above schematic,
GND is reference zero-volt of amplifier circuits.
Earth is earth from mains power, connected to chassis of appliance and can be used to shield EMF.
R1 and R2 represents to load of external circuits.
Note that transformer T1 is linear step-down transformer, with power rated up to 56 VA.
Mains power input would be 230VAC 60Hz.
C1 is Class-Y2 rated Film type capacitor. It bypass high frequency noises. Also would be helpful to suppress voltage spikes.
When voltage difference between Earth and GND is small than Forward voltage of Diodes(D1, D2), in general situation, small current pass through R3.
D1 and D2 would bypass high current when appliance encountered to failure. It should have ability to drain hundreds amps continuously. Also it's good to use low Vf diode as possible.
These diodes are critically important to protect user and device itself to safe.
Lastly, D3 is Bi-directional TVS diode, dissipate short-period voltage spike as heat when ESD event happens.
If my design above works as I intended, it will provide noise reduction from mains earth, while also keep user safe when appliance fails.
I'm really curious about performance and safety of this schematic.
Would this design pass any regulation of international or local standards?
It will be thankful to get any kind of advice here.
Thanks in advance.
Last edited:
D3 is not required.
You cannot reinvent the wheel.
Many commercial manufacturers already use a pair of 1N540* diodes back to back shunted with a low value ceramic capacitor (as an RF filter) and a low value resistor to reduce stray voltage charges.
Nothing new here, this is already in use and not new, sorry.
You cannot reinvent the wheel.
Many commercial manufacturers already use a pair of 1N540* diodes back to back shunted with a low value ceramic capacitor (as an RF filter) and a low value resistor to reduce stray voltage charges.
Nothing new here, this is already in use and not new, sorry.
Agree, it's a pretty common configuration (except the unnecessary TVS and slightly higher than usual resistor).
For example, from Elliot Sound Projects #127:
For example, from Elliot Sound Projects #127:
Hello.
Thank you for your replies.
I've never seen that before. I'll read the article that avtech attached.
But still I can't sure separating ground with mains earth is legal or meets related international standards.
Anyone have information about regulations of these kind of construction?
Thanks.
Thank you for your replies.
I've never seen that before. I'll read the article that avtech attached.
But still I can't sure separating ground with mains earth is legal or meets related international standards.
Anyone have information about regulations of these kind of construction?
Thanks.
Rod Elliot drew that?
I would connect the network from the audio circuit common to the chassis near the input connectors.
* * * * * * * * * * * *
also in the first post: C1 does not need to be a Class-Y2 capacitor. If there is a problem lots of other parts will fail first.
I would connect the network from the audio circuit common to the chassis near the input connectors.
* * * * * * * * * * * *
also in the first post: C1 does not need to be a Class-Y2 capacitor. If there is a problem lots of other parts will fail first.
So, you mean you'll connect the GND directly to earth/chassis?
I've read many threads regarding GLB and it's still a mystery to me whether it can/ should be used or not. And if it's legal to do so.
If all your cables, shields are attached to the metal chassis at their connectors, then there is little need for a GLB.
Some designers float the analog input and output connectors from the chassis to reduce certain ground loop effects. Its a design tradeoff question as to chassis grounding of analog I/O connections, or else to float the active circuity and use a hum-breaker circuit.
For example, in the case of a dac with galvanic isolation of USB, it can work to float the output ground and let it be chassis grounded in a subsequent preamp or power amp. In that case a hum-breaker circuit can be useful. That way there is less of an AC line ground loop going back around to the dac.
For example, in the case of a dac with galvanic isolation of USB, it can work to float the output ground and let it be chassis grounded in a subsequent preamp or power amp. In that case a hum-breaker circuit can be useful. That way there is less of an AC line ground loop going back around to the dac.
Last edited:
Yes, I will connect GND to Earth - Metal chassis will be connected to single point of Earth.
30 years ago, Neil Muncy (RIP) wrote an Audio Engineering Society paper explaining why all cable shields need to be attached to the metal chassis at their connectors. He dubbed it "the XLR Pin 1 problem". Later the Audio Engineering Society made it AES 48 standard.
Sure, and all interconnects should be balanced. We study what the textbooks, app notes, and articles say. Except doing it by rote doesn't always give the best sound.
Sometimes certain textbook ways even damage the sound more than help. Here I'm thinking of ferrites and nonlinear capacitors in dac clock power supplies.
In the long run science is supposed to be self correcting, but whatever is easiest to measure becomes the focus of attention first.
Besides, maybe Rod Elliot had his own reasons for publishing that hum breaker circuit?
Sometimes certain textbook ways even damage the sound more than help. Here I'm thinking of ferrites and nonlinear capacitors in dac clock power supplies.
In the long run science is supposed to be self correcting, but whatever is easiest to measure becomes the focus of attention first.
Besides, maybe Rod Elliot had his own reasons for publishing that hum breaker circuit?
Last edited:
In my experience Audio GND can be connected directly to PE only in the power amplifier and the other devices can have networks from Audio GND to PE. This will have the side effect that eventually all devices will be referenced to the direct connection of the power amplifier Audio GND (with RCA) to PE connection via cabling. This way safety and no ground loops. Guaranteed.
Or have all devices fitted with a GND reference network. The absolute worst is to have all devices with Audio GND connected to PE directly. Floating Audio GND is a possibility too but it may not be allowed everywhere and leakage will possibly make unpleasant potential differences.
Of course metal cases must always be connected directly to PE in DIY devices. This is its main purpose: safety. But we want safety and no hum 🙂
Or have all devices fitted with a GND reference network. The absolute worst is to have all devices with Audio GND connected to PE directly. Floating Audio GND is a possibility too but it may not be allowed everywhere and leakage will possibly make unpleasant potential differences.
Of course metal cases must always be connected directly to PE in DIY devices. This is its main purpose: safety. But we want safety and no hum 🙂
It is international standard to connect metal cases to PE directly for safety in class I devices like DIY stuff. There are no strict/hard regulations concerning Audio GND AFAIK. In fact they are separated by transformers normally. PE is primary side of things, Audio GND is secondary side of things. We always get the same farfetched question "what happens when an L wire comes loose and touches Audio GND?". Well when you use the ground breaker circuit the fuse will burn or the breaker will shutdown electricity. Also in directly connected Audio GND to PE but possibly less so in floating situations.
Last edited: