Is this monoblock common?

[Jay]

I'm building a "monoblock" amplifier in one box. I'm separating the ground for both channels by letting one of the channel ground floating. Is this wrong approach?

 

[pinkmouse]

As long as the cases are both safety grounded to earth, leaving a signal ground floating is perfectly valid.

 

[PMA]

As long as the cases are both safety grounded to earth, leaving a signal ground floating is perfectly valid.

This might be questionable. There is no shielding effect if PCB "floats" against case. You will lead capacitive coupled EMI inside the case, just along signal cables. As a minimum, signal gnd should be connected via capacitor with the case.

 

[peranders]

If your pcb has groundplane you have rather good sheilding but I would connect the case becuase it's good in almost every sense. I can't see why you shouldn't if the case is of metal.

 

[PMA]

Per,

the reason may be ground-loop problem. In case that both cases are connected to protective earth (PE), you create ground-loop by signal cables shields and two monoblock cases connected to PE in case that you interconnect signal GND to the case at both monoblocks.

Pavel

 

[pinkmouse]

This might be questionable. There is no shielding effect if PCB "floats" against case. You will lead capacitive coupled EMI inside the case, just along signal cables. As a minimum, signal gnd should be connected via capacitor with the case.

Sorry, a bit early in the morning for me, I didn't make myself clear, I was talking about the ground in the signal cable. After reading about this in Self's book, I always connect mains and signal ground at the input socket, and it seems to work well.

 

[Jay]

It was not about fighting the circular ground.

Circular ground (hum) was what I didn't expect from trying to do what I'm going to do (Safety is not a big deal).

I wanted to have the lowest degree of crosstalk between channels.

I wanted to test the validity of monoblock concept from economy perspective.

Distance between channels (especially the power supply) does count indeed, but often there is something more serious than just about distance.

 

[PMA]

I can guarantee that consequentional monoblock solution leads to best achievable localization in the soundstage.

 

[peranders]

Per,

the reason may be ground-loop problem. In case that both cases are connected to protective earth (PE), you create ground-loop by signal cables shields and two monoblock cases connected to PE in case that you interconnect signal GND to the case at both monoblocks.

Pavel

There is no need to connect to protective earth as long as the mains circuitry is OK, just let the case haven the same potential as the signal ground.

NOTE: This is only valid if you don't have a mains filter which has it's case connected (normal mains filter).

(Why: If an apparatus has mains filter and the case is floating you will get 115 VAC (or 55 V) in the chassis, not good.)

 

[trwh]

There is no need to connect to protective earth as long as the mains circuitry is OK, just let the case haven the same potential as the signal ground.

NOTE: This is only valid if you don't have a mains filter which has it's case connected (normal mains filter).

(Why: If an apparatus has mains filter and the case is floating you will get 115 VAC (or 55 V) in the chassis, not good.)

Any metal chassis containing points at mains potential must be securely connected to mains earth. I don't understand what you mean about mains filters.

Sorry if that sounds harsh, but this is an important point.

Cheers,
Tim.

 

[PMA]

Tim, this may depend on local safety laws/rules. It is possible to have an instrument in metal case with the case not connected to protective earth (in Europe). This depends on transformer's class of isolation, i.e. on the tests and certifications that transformer must fulfill. Why P-A speaks about line filter? That's because this has a middle point connected to the metal case. As filter has capacitors between phase and neutral, phase and chassis and neutral and chassis, the chassis must be connected to PE for safety reason.

But as I stated, this is a little controversial topic, as countries have different mains distribution networks with different kinds of voltage/touch protection and different safety rules.

 

[trwh]

Tim, this may depend on local safety laws/rules. It is possible to have an instrument in metal case with the case not connected to protective earth (in Europe). This depends on transformer's class of isolation, i.e. on the tests and certifications that transformer must fulfill. Why P-A speaks about line filter? That's because this has a middle point connected to the metal case. As filter has capacitors between phase and neutral, phase and chassis and neutral and chassis, the chassis must be connected to PE for safety reason.

But as I stated, this is a little controversial topic, as countries have different mains distribution networks with different kinds of voltage/touch protection and different safety rules.

Thanks for your post, PMA - I see what peranders means now.

I was aware that it is in theory possible to have a safe mains powered product with an unearthed metal chassis - my CD player is an example of this. However, it was mentioned to me that the requirements for doing this legally are very stringent, and it is easier and safer for the DIYer to just earth the chassis. Any ground loops can be defeated by the use of a ground loop breaker circuit, as described here:

http://sound.westhost.com/earthing.htm

I hope that helps,
Tim.