AndrewT said:
2 separate power transformers and 2 separate PCBs should do it.
janneman said:
I've done that, floating from the chassis. It often works, but you sometimes have RFI. One solution that *sometimes* work is to connect the signal ground to chassis via a 1nF capacitor, sometimes with 10 ohms or so in parallel.
I never have discovered a firm rule for this, but I have a couple of tricks and so far there's always one that does work
Jan Didden
Indeed, it seems that general rules cannot be applied here. A different solution may be needed in each case.
The Audio Ground is the main point where all the audio returns converge. This is the reference for all the circuits tied to it.
The Power Zero Volt will also connect to the Audio Ground.
The Safety Earth will also be continuous with this reference, except if there are NO EXPOSED CONDUCTIVE PARTS.
Unfortunately, most equipment is connected to other gear with interconnects. If these interconnects can be easily touched or removed from their sockets then the exposed (remote) end will defeat the requirement for deleting the Safety Earth connection.
Some refer to the Audio Ground as Star Ground, or Central Star, or Zero Volt Reference.
The Power Zero Volt will also connect to the Audio Ground.
The Safety Earth will also be continuous with this reference, except if there are NO EXPOSED CONDUCTIVE PARTS.
Unfortunately, most equipment is connected to other gear with interconnects. If these interconnects can be easily touched or removed from their sockets then the exposed (remote) end will defeat the requirement for deleting the Safety Earth connection.
Some refer to the Audio Ground as Star Ground, or Central Star, or Zero Volt Reference.
AndrewT said:
I find this all very confusing. The point where all audio returns converge has been called signal ground for the last century or so. I don't think it is helpful to introduce yet another term here. Then instead of Audio ground we can also have Star ground? Or Central star? Or zero volt reference? While the last seems to denote a voltage reference, the others are more physical implementations.
Can we please, please use the terms we are used to, and of which we know what they mean? It's difficult enough the way it is.
Jan Didden
I don't see the Signal Ground as equal to the Audio Ground.
It's difficult/confusing because we have grown used to everyone using the term "Ground" for a whole range of different uses/purposes as far as the circuitry is concerned.
If we strove to call each part by it's unique name, then the confusion would disappear.
It's difficult/confusing because we have grown used to everyone using the term "Ground" for a whole range of different uses/purposes as far as the circuitry is concerned.
If we strove to call each part by it's unique name, then the confusion would disappear.
yes, the isolated dual secondaries in the two transformers achieves exactly the same end as the isolated secondaries in a quad output transformer.Joshua_G said:
The problem stems from requiring two Audio Grounds inside the same chassis. That's where the disconnecting Networks come to help, but in my limited experience, cannot completely eliminate the downside of multi-channel amplifiers.
Jan,
I really don't want to belabor this at all.
But you said:
I said what I said, then you said what you said:
If I didn't understand, please reiterate.
I just want to know if you have ever heard preamplifiers or power amplifiers sound different to your ears, especially the case where on some technical level (you define that from your own personal experience and knowledge) in your opinion you did not expect to hear any differences??
Furthermore, if you (or anyone else) has ever put something in or out, or done something else that was absolutely not expected to make any audible sonic difference and was confounded by the result being an aubible sonic difference - again IN YOUR PERSONAL EXPERIENCE.
To Glenn: the overt hostility is not appreciated.
Maybe you've spent too much time near automotive power tools and engines without mufflers, and sadly damaged your hearing.
If one values one's hearing, one should always wear hearing protection around loud sounds.
_-_-bear
PS. I don't really care that much if anyone agrees or disagrees with my posts, including Curl. Happy birthday Curl?
I really don't want to belabor this at all.
But you said:
I said what I said, then you said what you said:
If I didn't understand, please reiterate.
I just want to know if you have ever heard preamplifiers or power amplifiers sound different to your ears, especially the case where on some technical level (you define that from your own personal experience and knowledge) in your opinion you did not expect to hear any differences??
Furthermore, if you (or anyone else) has ever put something in or out, or done something else that was absolutely not expected to make any audible sonic difference and was confounded by the result being an aubible sonic difference - again IN YOUR PERSONAL EXPERIENCE.
To Glenn: the overt hostility is not appreciated.
Maybe you've spent too much time near automotive power tools and engines without mufflers, and sadly damaged your hearing.
If one values one's hearing, one should always wear hearing protection around loud sounds.
_-_-bear
PS. I don't really care that much if anyone agrees or disagrees with my posts, including Curl. Happy birthday Curl?
Happy birthday, Mr. Curl
PMA, I don't see some important part in your ground drawing.
In the amplifier (X1 triangle), the output transistors will have local decoupling (maybe medium elko + nF caps). This output transistor's local decoupling should be placed as close as possible to output transistor's collector (if the output is common emitor).
There will be 2 local decoupling caps, one from +rail to ground, and one from -rail to ground. Lets call the ground point junction of these 2 caps = junction1.
Power amp also usually have output zobel (usually 10ohm+100nf) connected between output node and ground. Lets call the ground point of this zobel = junction2.
I know designing PCB that meets all criteria is impossible, but junction1 and junction 2 should be tied together, and I would put the loudspeaker ground (return path) to this point, where junction1 and junction2 also meets.
PMA, I don't see some important part in your ground drawing.
In the amplifier (X1 triangle), the output transistors will have local decoupling (maybe medium elko + nF caps). This output transistor's local decoupling should be placed as close as possible to output transistor's collector (if the output is common emitor).
There will be 2 local decoupling caps, one from +rail to ground, and one from -rail to ground. Lets call the ground point junction of these 2 caps = junction1.
Power amp also usually have output zobel (usually 10ohm+100nf) connected between output node and ground. Lets call the ground point of this zobel = junction2.
I know designing PCB that meets all criteria is impossible, but junction1 and junction 2 should be tied together, and I would put the loudspeaker ground (return path) to this point, where junction1 and junction2 also meets.
Hi Janneman & Alan,
are we all talking about the diagram in post14195?
If so, then PMA appreciates the difference. Post14207 refers to the separation of these various grounds.
The red dot next to PWRGND is the Power ground or Power Zero Volts. The Safety Earth Disconnecting Network can be connected here.
The red dot next to 0V is the Audio Ground or voltage reference for all the audio circuits. Here the Power Ground comes in with the decoupling return, with the Zobel return, with the speaker return, with the Signal Ground and optionally the Safety Earth disconnecting Network.
The Signal Ground to Audio Ground can be a direct low resistance trace or wire, but many builders find that using a resistor between Signal Ground and Audio Ground offers benefits. Leach explains his reasoning for it.
I adopted Leach's optional Signal Ground connection from the Input RCAs rather than from the PCB Signal Ground. This option seems to more consistently avoid/lessen the hum problems.
When the Signal Ground comes from the other side of that resistor (often 10r) it becomes more obvious that SG and AG are not equal.
are we all talking about the diagram in post14195?
If so, then PMA appreciates the difference. Post14207 refers to the separation of these various grounds.
The red dot next to PWRGND is the Power ground or Power Zero Volts. The Safety Earth Disconnecting Network can be connected here.
The red dot next to 0V is the Audio Ground or voltage reference for all the audio circuits. Here the Power Ground comes in with the decoupling return, with the Zobel return, with the speaker return, with the Signal Ground and optionally the Safety Earth disconnecting Network.
The Signal Ground to Audio Ground can be a direct low resistance trace or wire, but many builders find that using a resistor between Signal Ground and Audio Ground offers benefits. Leach explains his reasoning for it.
I adopted Leach's optional Signal Ground connection from the Input RCAs rather than from the PCB Signal Ground. This option seems to more consistently avoid/lessen the hum problems.
When the Signal Ground comes from the other side of that resistor (often 10r) it becomes more obvious that SG and AG are not equal.
AndrewT said:
OK, so then like you say Power Ground. Is it really necessary to use two terms for the same thing? Probably you feel I am splitting hairs, but IME the best way to derail a discussion is to introduce new terms, that have to be explained etc. Sometimes necessary, but not if it is just the same as an existing term everyone is familiar with. But whatever floats your boat....
Jan Didden
now that we sort of understand each other, what about your first question to me, namelyjanneman said:
I replied. What's your comment?
BTW,
I have, for the past few years, consistently referred to Signal Ground, Audio Ground and PSU Zero Volts, Safety Earth, simply because most readers understand what is meant, rather than calling them all "grounds".
I know they are different and most readers have come to realise they are different.
Why call them all "grounds" if it increases confusion, particularly among newcomers?
- Status
- Not open for further replies.