understanding star grounding

Jump to Latest
A

AndrewT

R.I.P.
Joined 2004
454Casull said:


An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


The threaded part with the outside barrel is connected to the shield. As you can see, though, the inner barrel which is directly adjacent to the black insulator in the receptable cannot touch the housing until it is fully inserted, at which point the pins will already have been seated. The threads cannot be relied on to contact before the pins.
Click to expand...
go and have another look at that plug and socket combination.
See the plug surround with the semicircular protusion, opposite the middle pin.
Now look at the socket and it has a complementary rectangular recess in the same location.
There should be a metal spring connected to the shell in that recess and it should contact the matching protusion as the plug and socket are mated together.
Is the spring contact still there or is it missing?
 
4

454Casull

Member
Joined 2004
AndrewT said:
go and have another look at that plug and socket combination.
See the plug surround with the semicircular protusion, opposite the middle pin.
Now look at the socket and it has a complementary rectangular recess in the same location.
There should be a metal spring connected to the shell in that recess and it should contact the matching protusion as the plug and socket are mated together.
Is the spring contact still there or is it missing?
Click to expand...
I wondered about the same thing before. Unfortunately, the recess which the "key" inside the male shield extension slides in during insertion does not have a contact.
 
S

samsagaz

Member
Joined 2004
David Davenport said:
Okay, I'll give you two answers so you can make a choice.

First, the best solution is to galvanically isolate the computer from the amplifier with audio line transformers (one for each channel). I would suggest the EDCOR TPC10K/10K. While it is not the very best quality, it is very good and a good match for a computer sound card. It costs about $11 US each. For this solution:

- connect the signal and shield from the computer to the input of the transformer (each channel).
- connect the outputs of the transformers to the amplifier (signal and ground).
- connect the chassis to the safety ground.
- connect the chassis/safety ground to the star.
- connect the power supply common (0V) to the star.
- connect the amplifier reference to the star.
- connect the negative speaker lead to the star.

The second solution, while not as good as galvanic isolation, is less expensive and and the parts may be easier to get. That is to use a Safety Loop Breaker circuit. For this solution:

- connect the chassis to the safety ground.
- place the SLB between the chassis/safety ground and the star ground.
- connect the shield fron the computer connection to the star.
- conect the power supply common (0V) to the star.
- connect the amplifier reference to the star.
- connect the negative speaker lead to the star.

If you have only a computer and amplifier in the system, the SLB solution may work fine for you. However if you still have unaccepable hum then you will need to galvanically isolate the computer from the amplifier.

Dave
Click to expand...


Sorry for this dumb question... but
If cant get an decent Diodes Bridge to create the SLB, what kind of Diodes can i use? how i connec it with the resistor and Cap?

Regards
 
A

AndrewT

R.I.P.
Joined 2004
SLB?
Is that the Disconnecting Network between the Safety Earth and the Audio Ground?

If so, then a 35A bridge rectifier does the job.
Because the cross coupled rectifier passes fault current in both directions, the voltage rating is never used, so a 50V version is perfectly adequate.
 
D

David Davenport

Member
Joined 2004
raleighaudio.com
Yes, you may use four diodes instead of a bridge. I don't see any reason that you can't use the same kind of diodes that you use as rectifiers in the power supply.

You can make a bridge configuration, the same as with your power supply rectifiers ( I assume that you have a bridge there.) Using your power supply as an example, you connect the two AC inputs together and the two DC outputs together. So you end up with four diodes in parallel, two facing one way and two facing the opposite way.

Since the four diode arrangement is symmetrical, it doesn't matter which end goes to the chassis/earth and which goes to the star ground. Then the resistor and the capacitor each go in parallel with the diodes, from earth/chassis to the star.

Dave
 
D

David Davenport

Member
Joined 2004
raleighaudio.com
The Safety Loop Breaker circuit needs to protect against two things: a short between the primary and secondary windings of the power transformer and a short from circuitry on the output of the power supply to the chassis.

The Safety Loop Breaker Circuit was introduced in Rod Elliott's article for a 100W guitar amplifier. In that article he specified that the bridge rectifier for the SLB be rated at 5A, the same as the fuse rating for the amplifier. This makes sense because the diodes need only conduct until the fuse blows.

In Rod Elliott's Earthing article he suggests a 35A bridge. In this article he assumes a generic power supply design and does not show any fuse. The 35A bridge will conduct until the circuit breaker in the mains panel trips.

Therefore, to protect for a short between the primary to secondary windings of the power transformer, the SLB diodes need not be 35A and may be chosen to match the equipment fuse rating.

To protect against a short from circuitry on the output of the power supply, the SLB diodes need not be any larger than the power supply rectifier diodes.

Since the power supply voltage for a chip amplifier is lower than the mains voltage (step down transformer) and the current higher than the mains current, the rectifier diodes are rated to support this higher current.

Therefore, the same type diodes used as rectifier diodes may be used as SLB diodes.

Dave
 
A

AndrewT

R.I.P.
Joined 2004
David Davenport said:
The Safety Loop Breaker circuit needs to protect against two things: a short between the primary and secondary windings of the power transformer and a short from circuitry on the output of the power supply to the chassis.

The Safety Loop Breaker Circuit was introduced in Rod Elliott's article for a 100W guitar amplifier. In that article he specified that the bridge rectifier for the SLB be rated at 5A, the same as the fuse rating for the amplifier. This makes sense because the diodes need only conduct until the fuse blows.

In ................................................trips.

Therefore, to protect for a short between the primary to secondary windings of the power transformer, the SLB diodes need not be 35A and may be chosen to match the equipment fuse rating.

To protect against a short from circuitry on the output of the power supply, the SLB diodes need not be any larger than the power supply rectifier diodes.

Since the power supply voltage for a chip amplifier is lower than the mains voltage (step down transformer) and the current higher than the mains current, the rectifier diodes are rated to support this higher current.

Therefore, the same type diodes used as rectifier diodes may be used as SLB diodes.
Click to expand...
David,
I think you are wrong in almost every statement you have made.

I will restate my opinion:

The Disconnecting Network (otherwise known as a Safety Loop Breaker) must survive longer than it takes the mains fuse to blow and for the arc to extinguish.

Do you have any idea of how many amps can flow from Live to Protective Earth during a mains fault condition?
I believe that even in a domestic situation this Fault Current can approach kA.
 
kvholio

kvholio

Member
Joined 2005
The disconnecting network must be designed to withstand the worst possible fault-condition.
The worst possible i can think of is when the audio/powersupply ground becomes Live.

To protect against such a condition the mains fuse must rupture.
If the disconnecting-network fails before the fuse ruptures,
the audio/powersupply ground stays Live.
Now imagine someone touching the (conductive) rca-plugs
(which are connected to audio/ps ground) under these conditions.

Klaas
 
A

AndrewT

R.I.P.
Joined 2004
kvholio said:
The disconnecting network must be designed to withstand the worst possible fault-condition.
The worst possible i can think of is when the audio/powersupply ground becomes Live.
Click to expand...
that's a bit odd.
You seem to think that David omitted the worst possible fault condition from his list.

I wonder how many newbies will read his post during the remainder of this decade?
 
kvholio

kvholio

Member
Joined 2005
You're right Andrew, i should have read David's post more carefully.:ashamed:

The point in my post (which i'm afraid wasn't too clear from my post) is that the worst-case current is only limited by the mains fuse.
It is this worst-case current that dictates the dimensioning of the disconnecting-network,
and i believe this current can be MANY times higher than the max. current-rating of the mains fuse.

Best regards,

Klaas
 
You must log in or register to reply here.
Top Bottom