Cortez said:Hi Mike !
Obviously you didnt noticed my Q to you while I was OFFing, sorry!
Hopefully this image is readable and will answers some questions. Please give it some thought and suspend reality for a bit and assume it might work. I'll try to answer any questions to help clarify what I'm up too.
You might notice I try to route any returns from regulation and decoupling apart from the gain stages through their own low-Z path. This you can see in the previous images I posted.
Mike.
Attachments
Well, I'm getting confused... 
Unfortunetly I really dont get this.
- every wire is now just ground wire, or there is no ground wire from the source components ?
- are the sources safety earthed too ?
- in every component, the ground is earthed too ?
- what wires are shielded and what aren't and where to the shields are connected ?
Really sorry Mike, I'm trying to understand it, but isnt easy.
Unfortunetly I really dont get this.- every wire is now just ground wire, or there is no ground wire from the source components ?
- are the sources safety earthed too ?
- in every component, the ground is earthed too ?
- what wires are shielded and what aren't and where to the shields are connected ?
Really sorry Mike, I'm trying to understand it, but isnt easy.
Cortez said:Well, I'm getting confused...
1 ) every wire is now just ground wire, or there is no ground wire from the source components ?
2 ) are the sources safety earthed too ?
3 ) in every component, the ground is earthed too ?
4 )what wires are shielded and what aren't and where to the shields are connected ?
Really sorry Mike, I'm trying to understand it, but isnt easy.
I'm trying here as well, thanks for the patience.
1 ) I just tried to depict mostly the signals and the grounds; focusing on the loops.
The circuitry (Amp driver PCB, Lineamp, line amp PS regulators) are basically blocks showing the input and it's ground, the output (which doesn't need a ground from the board).
The regulators, which show a raw dc in, an out to the line amp and a path through any on card regulation, bypass caps etc. back to the ground in to complete the loop.
The safety grounds, just between the amp and preamp to indicate there is no voltage potential if they are hooked to the same voltage source.
2 ) No, the sources are connected into the preamp safety ground by their interconnects. The red lines show the path of any noise currents picked up by the shield or chassis currents, which bypasses the signal loops
3 ) I'm not sure I understand your question here. I don't see a ground and an earth. I see every component having aground, or zero volt reference at one point at the transformer. Every ground ends connected to it through controlled paths.
4 ) I forgot the shields on the source components (sorry), but they would be the same as on the between the amp and preamp. The connectors are all bonded to either the chassis or a copper ground plane which takes a low-Z path back to the PS central ground. The interconnects inside the preamp show the shields which are bonded to the chassis at the jack only.
Amplifier layout and grounding is all about knowing where the power is flowing and where it returns and the path of the signal.
Hopefully this is getting clearer. I might get a chance to take a pic of my preamp to show a real world implementation. Time is at a minimum
Mike.
> The safety grounds, just between the amp and preamp to indicate there
> is no voltage potential if they are hooked to the same voltage source.
But could have difference at the chassises already, due to interferences, couldnt ? Like an antenna, or capacitively.
> The red lines show the path of any noise currents picked up by
> the shield or chassis currents, which bypasses the signal loops
But if a noise is travelling through the signal GND wire, why wont we hear it ?
> > in every component, the ground is earthed too ?
> I'm not sure I understand your question here.
> I don't see a ground and an earth.
I meant the safety earth. In the amp and in the preamp the grounds are
connected to the mains safety earth, but this is also true at the soures ?
Are their signal GNDs also connected to their chassis and to the mains safety earth ?
So:
- every chassis is earthed
- every ground is earthed
- every incoming and outgoing signal ground is earthed
- every signal cable is shielded, connecting to this earth
- every ground starts from the main star GND wich is after tha trafo
Right ?
> is no voltage potential if they are hooked to the same voltage source.
But could have difference at the chassises already, due to interferences, couldnt ? Like an antenna, or capacitively.
> The red lines show the path of any noise currents picked up by
> the shield or chassis currents, which bypasses the signal loops
But if a noise is travelling through the signal GND wire, why wont we hear it ?
> > in every component, the ground is earthed too ?
> I'm not sure I understand your question here.
> I don't see a ground and an earth.
I meant the safety earth. In the amp and in the preamp the grounds are
connected to the mains safety earth, but this is also true at the soures ?
Are their signal GNDs also connected to their chassis and to the mains safety earth ?
So:
- every chassis is earthed
- every ground is earthed
- every incoming and outgoing signal ground is earthed
- every signal cable is shielded, connecting to this earth
- every ground starts from the main star GND wich is after tha trafo
Right ?
Attachments
Hi Mikeb,
give me more time to print your diagram off at home and I'll be back.
But one question first
At item 7 in your diagram you say
give me more time to print your diagram off at home and I'll be back.
But one question first
At item 7 in your diagram you say
But the loops set up between the equipment should/will generate fields that do create voltages and currents between the different reference grounds. Don't they?
And what about a layout like this ?
- every chassis is earthed
- every cable is shielded, but the shields are only connected to the chassis, both side
- the signal grounds arent connected to the shields
- the gnds arent connected to the earth
- every signal ground goes firstly to the main GND, and not directly to the circuit
So the earth from the mains is a whole different and separate line from
our grounds, even so our chassises are earthed, and our signal cables are shielded...
- every chassis is earthed
- every cable is shielded, but the shields are only connected to the chassis, both side
- the signal grounds arent connected to the shields
- the gnds arent connected to the earth
- every signal ground goes firstly to the main GND, and not directly to the circuit
So the earth from the mains is a whole different and separate line from
our grounds, even so our chassises are earthed, and our signal cables are shielded...
Attachments
I will answer more completely a bit later. I'll see if I can produce a readable schematic. Just one quick comment. The safety ground has taken on more importance than I intended. In an earlier post I said that the only place that a safty ground should be connected is at the preamp, and I implied I did not feel the need for one at all. I'll leave it out of my next attempt.
My point in trying to add it was to answer a question and illustrate you could control the path of the noise so that even if there was noise it would not end up in the signal path.
One other concept, Noise coupllng follows rules, there are coupling methods and susceptablity parameters. The two wires being electrically connected at the plug and looping between the chassis does not mean that it will act like an antenna, and in the small chance it could, there needs to be a path for coupling to the signal path. A method to avoid this is what I'm trying to describe.
I'll send more later.
Mike
My point in trying to add it was to answer a question and illustrate you could control the path of the noise so that even if there was noise it would not end up in the signal path.
One other concept, Noise coupllng follows rules, there are coupling methods and susceptablity parameters. The two wires being electrically connected at the plug and looping between the chassis does not mean that it will act like an antenna, and in the small chance it could, there needs to be a path for coupling to the signal path. A method to avoid this is what I'm trying to describe.
I'll send more later.
Mike
> you could control the path of the noise so that even if
> there was noise it would not end up in the signal path
Lets clear this: if there is any noise voltage and current on our
signal cable, why wouldnt it appear on the signal at the input ?
Noise is like a voltage source in series with the signal source, isnt it ?
Or how could we modell it with RLCs and voltage generators ?
> there was noise it would not end up in the signal path
Lets clear this: if there is any noise voltage and current on our
signal cable, why wouldnt it appear on the signal at the input ?
Noise is like a voltage source in series with the signal source, isnt it ?
Or how could we modell it with RLCs and voltage generators ?
Hi Cortez,
post88 appears to work IF you keep the safety grounds SEPARATE from all the other audio grounds.
Are we right?
Mikeb.
do this in the UK
IT MUST BE SAFE, no other choices.
post88 appears to work IF you keep the safety grounds SEPARATE from all the other audio grounds.
Are we right?
Mikeb.
do this in the UK
and you will be breaking the law if the case has any exposed conductive parts that are not earthed. or it is designed/tested to meet the double insulated standard.
IT MUST BE SAFE, no other choices.
> IF you keep the safety grounds SEPARATE from all the other audio grounds
Yes, thats important. I wouldnt mix safety earth, chassis and the shieldings with the signal, if its possible.
> Are we right?
I hope so...
An even better solution would be this layout with a "ground-independent" symetrical signal-wiring...
Otherwise I never use shielding anywhere, and have no problems
with noise, even with asymetrical wires, what I'm mainly using.
Yes, thats important. I wouldnt mix safety earth, chassis and the shieldings with the signal, if its possible.
> Are we right?
I hope so...
An even better solution would be this layout with a "ground-independent" symetrical signal-wiring...
Otherwise I never use shielding anywhere, and have no problems
with noise, even with asymetrical wires, what I'm mainly using.
AndrewT said:
I'm not designing anything to be sold, I'm experimenting with grounding, shielding and noise and it's affect on the sound. I hate to think of a world where the subtlety and beauty are choked out of everything because we have to idiot proof it.
I've also come to realize that, what is perfectly clear to me has become that way because it's been my work and hobby for so many years. Someone else understanding it would require them to suspend their belief system and replacing it with mine.
The more I think about it, the more I remember all of the small hurtles along the way to refine this. Trying to throw it out in an easily digestable form is doomed to failure.
It has forced me think and for that, thanks. If anyone does check this out, at their own risk I will add, and has any questions about it or problems I would be happy to help.
Mike.
> I hate to think of a world where the subtlety and beauty are
> choked out of everything because we have to idiot proof it.
The more one can understand something, the better he can accept it.
I really want to understand your layout, dont give up explain it to us, please !
> Someone else understanding it would require them to
> suspend their belief system and replacing it with mine.
For example I dont have beliefs like this, just want to really understand the point, and the brackground.
Otherwise what is your opinion on my last figure / idea ?
> choked out of everything because we have to idiot proof it.
The more one can understand something, the better he can accept it.
I really want to understand your layout, dont give up explain it to us, please !
> Someone else understanding it would require them to
> suspend their belief system and replacing it with mine.
For example I dont have beliefs like this, just want to really understand the point, and the brackground.
Otherwise what is your opinion on my last figure / idea ?
Cortez said:
Your figure shows all of the grounds daisy chained from one gain block to the next. If you add the effect of the wiring and you have essentually the modified version of your figure that I've attached. All of the circuit support return currents, the garbage shunted around your circuit by bypass and decoupling caps, any RFI, etc. creates a voltage drop across the resistors I've inserted. The ground is no longer ground but ground plus whatever voltage is present (which is a form of unintended feedback). Think of it real terms. It means the reference used by both the input and feedback is comprized of zero volts at the transformer plus whatever voltage is developed along the way. This is way over simplified because we are not taking into effect frequency, phase, magnetic or electrical coupling and their impedance, the frequency related effects of the components, etc.
Not to mention that each stage in the chain is affected by the demands of the stages before and after it. Better stock up on decoupling caps...
Mike
Some people claim sonic superiority and less potential hum problems with two bridges. See:
http://www.diyaudio.com/forums/showthread.php?postid=143874#post143874
http://www.diyaudio.com/forums/showthread.php?s=&threadid=49327&perpage=10&highlight=&pagenumber=3
for some examples, searching will find you more. In my own experiments I found the dual bridge configuration superior.
Take care,
Doug
Mike:
Please, draw me a very simplified figure like mine, to help me understand your statements.
I think 2 stage would be way enough, no need for a preamp yet, it just confuses me.
And you dont have to draw the mains, the whole PS, the rails, just the main star ground point, the
chassis, all the signal wires, grounds, shieldings and the connections between these.
Then we could discuss everything precisely. This should be max 4-5 minutes with paint.
For example I still dont know, what layout do you use modeled with this figure:
http://www.diyaudio.com/forums/attachment.php?s=&postid=900011&stamp=1145856645
#3 + #2 without #1 ?
> Your figure shows all of the grounds daisy chained from one gain block to the next.
Not exactly, every stage's chassis and shields are earthed individualy.
> the garbage shunted around your circuit by bypass and decoupling caps,
> any RFI, etc. creates a voltage drop across the resistors I've inserted.
What resistors do you mean, where are they ? For example now, it would be
great to have a good figure to define exactly what we are talking about.
Otherwise the same doesnt appears in your design ?
Please, draw me a very simplified figure like mine, to help me understand your statements.
I think 2 stage would be way enough, no need for a preamp yet, it just confuses me.
And you dont have to draw the mains, the whole PS, the rails, just the main star ground point, the
chassis, all the signal wires, grounds, shieldings and the connections between these.
Then we could discuss everything precisely. This should be max 4-5 minutes with paint.
For example I still dont know, what layout do you use modeled with this figure:
http://www.diyaudio.com/forums/attachment.php?s=&postid=900011&stamp=1145856645
#3 + #2 without #1 ?
> Your figure shows all of the grounds daisy chained from one gain block to the next.
Not exactly, every stage's chassis and shields are earthed individualy.
> the garbage shunted around your circuit by bypass and decoupling caps,
> any RFI, etc. creates a voltage drop across the resistors I've inserted.
What resistors do you mean, where are they ? For example now, it would be
great to have a good figure to define exactly what we are talking about.
Otherwise the same doesnt appears in your design ?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.