Hi,
On some of the websites I read that bridge rectifier in power supply unit will have floating points.
Also I have gone through some power supply designs, many of them have designed with hot and neutral mains power cables with fuse protection.
However I feel important to provide earth ground to whole circuit. I have got below scematic from web.

I am not clear on earth connection to transformer.
Do I need to connect earth ground to the metal chessy of transformer?
Also, Do I need to connect earth ground to metal chessy of power supply?

Any kind of suggestions are welcome.
Thanks in advance.

 

The earth ground should be connected to the transformer frame you can use the same point to earth connect the -ve, only as long as the AC secondary is not connected to earth GND in any way.
All metal parts of enclosures should be also connected to earth GND.
However, never use the earth Ground as a conductor, IOW the DC -ve should have its own terminations and discrete conductors.
Max.

 

Yes that is a safe connection in your diagram.

The box around the transformer is shown earthed. This box will be the frame of the transformer and any metal shrouding.

Some transformers also have an electrostatic screen between primary and secondary. This should also be earthed.

It is not necessary to earth the output from your bridge, unless the voltage is high.
In fact you may want to avoid this to create a floating power supply.
Many designs do this and provide an earth terminal at the output so that the user can earth either positive or negative or neither if he so wishes.
Your circuit is capable of this very useful level of sophistication, perfectly safely for voltages up to about 150V.

Note the mains fuse should only be in the line. The neutral should not be fused.

It is also conventional to separately fuse the output in many cases.

 

I have also seen a Megohm resistor connected from bridge negative to Earth ground. This stops static charges from accumulating but makes it entirely clear that the Earth conductor is NOT a current carrying ground for your circuits. In some cases, no connection at all is the right thing to do. When you accumulate enough experience, you will know when to use this completely isolated style.

 

Often when testing circuits, you don't want the power supply output connected to earth ground, since the various instruments (oscilloscopes, signal generators, etc.) already have their input/output common going to earth ground. Connecting the power supply to earth ground in that case just creates an added ground loop which can sometimes generate signal noise.

 

Also much depends on whether this is a small self contained application, or if it is part of a larger system where other supplies are incorporated and also when AC operated devices are part of the same system.
Max.

 

Elaborating: A good, low impedance ground on the power supply output side (negative side, positive side, or center tap of the transformer secondary) and a good, low impedance ground on the oscilloscope probe will cause a short circuit in many places you would want to measure the circuit. This is one place where you would want complete isolation. Some people use a 3 wire to 2 wire converter on the power cord to the oscilloscope, but that violates safety rules in most places and might get a moderator rapping my knuckles for even mentioning it. A better way would be to use an isolation transformer on the power supply to get it disconnected from any Earth ground circuits while measuring it.

You see, it gets sticky trying to obey all safety rules and Wiring Codes while working on electronics. Experience and accumulated education will get you through these sticky situations.

 

Yes that is a safe connection in your diagram.

The box around the transformer is shown earthed. This box will be the frame of the transformer and any metal shrouding.

Some transformers also have an electrostatic screen between primary and secondary. This should also be earthed.

It is not necessary to earth the output from your bridge, unless the voltage is high.
In fact you may want to avoid this to create a floating power supply.
Many designs do this and provide an earth terminal at the output so that the user can earth either positive or negative or neither if he so wishes.
Your circuit is capable of this very useful level of sophistication, perfectly safely for voltages up to about 150V.

Note the mains fuse should only be in the line. The neutral should not be fused.

It is also conventional to separately fuse the output in many cases.

Thank you for your reply.
What do you mean by separately fuse the output??
Do I need to add a fuse in transformer secondary?

 

Also much depends on whether this is a small self contained application, or if it is part of a larger system where other supplies are incorporated and also when AC operated devices are part of the same system.
Max.

Elaborating: A good, low impedance ground on the power supply output side (negative side, positive side, or center tap of the transformer secondary) and a good, low impedance ground on the oscilloscope probe will cause a short circuit in many places you would want to measure the circuit. This is one place where you would want complete isolation. Some people use a 3 wire to 2 wire converter on the power cord to the oscilloscope, but that violates safety rules in most places and might get a moderator rapping my knuckles for even mentioning it. A better way would be to use an isolation transformer on the power supply to get it disconnected from any Earth ground circuits while measuring it.

You see, it gets sticky trying to obey all safety rules and Wiring Codes while working on electronics. Experience and accumulated education will get you through these sticky situations.

Thanks for the info.
I am designing 24 volts, 2 amps power supply for hobby projects.
This is my first independent project and as I need to handle mains supply, I thought ground connection is required for safety.

 

Do I need to add a fuse in transformer secondary?

Most people don't fuse the secondary because a properly sized fuse in the primary (125% slow-blo) will be sufficient to take the supply out in a few tenths of a second. However, the secondary is very predictable in a power supply, so sometimes it's convenient to fuse that, especially on very low power applications where you can get fuses down to 1/10th of an amp.

 

Also much depends on whether this is a small self contained application, or if it is part of a larger system where other supplies are incorporated and also when AC operated devices are part of the same system.
Max.

Also much depends on whether this is a small self contained application, or if it is part of a larger system where other supplies are incorporated and also when AC operated devices are part of the same system.
Max.

Thanks for the info.
Am designing 24 volts, 2 amps power supply unit for hobby projects.

 

Thank you all for your feedback.

 

This is my first independent project and as I need to handle mains supply, I thought ground connection is required for safety.

You should as a minimum ensure the transformer frame is earth grounded, for small contained projects it is not imperative to ground the low voltage DC side.
Keep in mind that if you hook any project up to a P.C., the DC common of your project with be earth grounded via the PC DC supply.
Max.

 

Hi,
On some of the websites I read that bridge rectifier in power supply unit will have floating points.
...
I am not clear on earth connection to transformer.
Do I need to connect earth ground to the metal chessy of transformer?
Also, Do I need to connect earth ground to metal chessy of power supply?
...

Most folks that build a lab supply do it like this with three output terminals:



Then they can decide at the last min if they want a "positive voltage with respect to earth ground", "negative voltage with respect to earth ground", or a "floating supply with respect to earth ground" by jumpering between + and Gnd, - and Gnd , or leaving out the jumper, respectively.

 

MileML, as usual, talks a lot of horse sense.

That is exactly the option I was describing.

Fuses are intended to protect something.
So you should ask yourself "What am I protecting?"

The fuse usually protects something on the supply (incoming) side of the fuse and makes the outging side dead so that it is safe to touch.

So a mains fuse protects the building wiring from overload due to the connection of faulty equipment.
It also makes that faulty equipment dead so that it will not presnet a safety hazard.

But a mains fuse can be inconvenient, especially for experimental equipment.

So some equipment, such as a bench power supply, has a cutout or fuse in series with the output to protect the connected equipment if faulty yet allow ease of reset once the fault has been eliminated.

A second reason for an output fuse would be that if the power supply is regulated the output may be able to withstand a momentary short circuit, but a sustained short circuit will cause overheating and failure.
Such power supplies have a very fast acting electronic protection accompanied by a much slower fuse that will only act if the short is sustained.

 

MileML, as usual, talks a lot of horse sense.

That is exactly the option I was describing.

Fuses are intended to protect something.
So you should ask yourself "What am I protecting?"

The fuse usually protects something on the supply (incoming) side of the fuse and makes the outging side dead so that it is safe to touch.

So a mains fuse protects the building wiring from overload due to the connection of faulty equipment.
It also makes that faulty equipment dead so that it will not presnet a safety hazard.

But a mains fuse can be inconvenient, especially for experimental equipment.

So some equipment, such as a bench power supply, has a cutout or fuse in series with the output to protect the connected equipment if faulty yet allow ease of reset once the fault has been eliminated.

A second reason for an output fuse would be that if the power supply is regulated the output may be able to withstand a momentary short circuit, but a sustained short circuit will cause overheating and failure.
Such power supplies have a very fast acting electronic protection accompanied by a much slower fuse that will only act if the short is sustained.

Ohk. Thanks for short circuit info.

 

Most folks that build a lab supply do it like this with three output terminals:

View attachment 72868

Then they can decide at the last min if they want a "positive voltage with respect to earth ground", "negative voltage with respect to earth ground", or a "floating supply with respect to earth ground" by jumpering between + and Gnd, - and Gnd , or leaving out the jumper, respectively.

Thanks Mike