Hello Gajanan and also sesebe & sangram
For all you three, to answer your questions that have common points...
"All the input signal gnds should be brought back to the input ground only..."
If you look at my post#85, the one with the new Gnd schematic, input(s) ground(s) are shown the way you suggest.
But I have a plan B for them if it does not prove to be good as I will have space on the Star point for these to connect, eventually.
I actually have a pure cooper bar of ½"x½" to create this Star Point.
I will start this job tomorrow.
Ok now...about the PCB grounds layouts you want to have a look at...
They won't be too much in the equation right now because I changed my mind with the schematic's layout and so decided to run six separate gauge10 wire from amp's circuit's main grounds TO the star point.
And 6 other gauge 16 wire from the Amp's circuit's conjugates TO the star point again.
No more amp grounds connected to PCB anymore. Zilch!
So finally the PCB's outputs exists only for the three per board Vdc+ and Vdc- and same thing config for RegPSU.
And their source which are the two main Elcaps on each board will be runned by Gauge 10 wires ( 4 inches long) to star grounds.
So at the end I will have 25 wires connected to this new star point.
It will be less job than I first thought.
I wish it'll be OK!
I will be back with whatever result I end with.
Luke
Last edited:
Is it possible to get a picture of the top and bottom of the Mains PSU board?
Increasing the distance of the bulk capacitors from the star point ground is not going to do much good.
The ideal way to locate the return ground is close to the point where the zero volt connection of the capacitors will be.
But in your case, it at least needs to be closer to, if not right at the bulk capacitors' zero volt point.
Increasing the distance of the bulk capacitors from the star point ground is not going to do much good.
The ideal way to locate the return ground is close to the point where the zero volt connection of the capacitors will be.
But in your case, it at least needs to be closer to, if not right at the bulk capacitors' zero volt point.
Hello Pixo
On each MainPSU PCBoards, there will be two 4 inches long Ga10 Wires (350µOhms each...) that will connect directly from the cap's pads to the star point.
There will be more resistance on the solders and contacts than the wire itself so I should not care too much.
Should I?
But like I told you indirectly the PCB's ground is not used by amp's circuit anymore.
From by the star point view, thje PCB is only used to "hold the caps".
The amps are direct to StarP.
By the way did you see my Rev.4 at post #103?
Luke
Last edited:
Hello Bonsai and thanks
For the sake of discussion...
Actually and like any other amplifier, I guess, the main caps have their center terminals coupled together by a PCB track and in the center of this track or if you want, between these the two caps's pads exists what we agree to call 0 ohms point, right?
1) So is there a difference between soldering a solid wire junction between these two caps to connect a wire in the middle of this solid junction that will be send to the Star point...
AND
2) Despite the fact that while there exists the PCB track between the caps, that a pair of wires be soldered to each of these caps pads, to finally be sent to the star point.
I will add that it makes me feel that the extra solid wire junction between the two caps is redondant.
Am I right?
Don't forget...it's for sake of discussion!
If not I'll simply adopt #1".
Luke
I forgot to mention...
I can't give this picture right now as I'm far from the computer that has the required folder.
I don't have my PCB software in the present computer.
Luke
Here is a representation of how current flows through the entire circuit.
The current flow is always from higher potential to lower. But with respect to the speaker, you see it goes forwards and backwards which is what causes the speaker to vibrate, producing sound.
The speaker return IS the high current ground return that completes the amplifier circuit.
The ground return that comes from your amplifier module is important, but very little current flows through that compared to the speaker's return wire.
When connected as shown, the current is returned back to the power supply taking the path of least resistance. This keeps the current loop consistent.
The current flow is always from higher potential to lower. But with respect to the speaker, you see it goes forwards and backwards which is what causes the speaker to vibrate, producing sound.
The speaker return IS the high current ground return that completes the amplifier circuit.
The ground return that comes from your amplifier module is important, but very little current flows through that compared to the speaker's return wire.
When connected as shown, the current is returned back to the power supply taking the path of least resistance. This keeps the current loop consistent.
Attachments
This advice is absolutely critical to good performance.
The Junction between the Main smoothing caps (PSU Zero Volts) passes enormous peak charging current.
This junction must NOT be used as the Main Audio Ground
Connect the Audio circuits to the MAG when these other audio circuits NEED a common reference.
Connect the MAG to the PCB (or amplifier if P2P) Power Ground.
Connect the amplifier Power Ground to the PSU Zero Volts.
BUT !!!!!!
many of the Return wires that connect ultimately to the MAG carry AUDIO signal Currents. These Return Wires are part of a two wire connection circuit. The Flow wire of that two wire circuit must remain close coupled or twisted with it's Return wire.
This regularly means that the two wires approach the MAG, the signal Flow and the Signal Return. The Signal return taps into the MAG and then couples back to the signal flow to continue it's two wire connection to it's destination.
An example.
the speaker is a two wire connection, the Flow and the Return.
The Flow leaves the amp PCB and heads for the MAG passes over the MAG and continues it's journey to the chassis mounted terminal.
The Return leaves the Chassis mounted terminal and is twisted with the flow until it reaches the MAG. Here Return makes it's tapping.
But the Return has NOT yet finished it's journey. It MUST return all the way to the Source of that Speaker current.
The remainder of the Return route is the amp PCB Power Ground to MAG wire. This wire MUST be twisted with the speaker Flow wire that is the other half of the speaker connection.
You now have a twisted two wire connection that goes from amp PCB to Chassis terminals, with a tapping into the MAG.
At no stage does the amplifier have a big loop in the speaker two wire connection. You MUST endeavor to maintain the twisted and close coupled Flow and Return as much as is physically possible in the space available. The space around the MAG gets very busy.
Last edited:
The picture here shows what you're trying to do.
Although the current loops have been shown circling within the circuits, it does not happen that way anymore.
The current through the speaker's return wire has to go to the central star point and then travel back to the zero-volt connection of the bulk capacitors.
Then there are six return flows that will have to be balanced out and since the flow is never constant within the circuit, it creates ripples.
You're listening to music after all, not trying to power a light bulb.
As the ground connections are now spanned out and connected to each other, this creates differences in potential within different points in the ground due to uneven current flow, giving rise to noise.
Although the current loops have been shown circling within the circuits, it does not happen that way anymore.
The current through the speaker's return wire has to go to the central star point and then travel back to the zero-volt connection of the bulk capacitors.
Then there are six return flows that will have to be balanced out and since the flow is never constant within the circuit, it creates ripples.
You're listening to music after all, not trying to power a light bulb.
As the ground connections are now spanned out and connected to each other, this creates differences in potential within different points in the ground due to uneven current flow, giving rise to noise.
Attachments
"1) So is there a difference between soldering a solid wire junction between these two caps to connect a wire in the middle of this solid junction that will be send to the Star point..."
Yes - a big difference. You have the capacitor chargingf currents flowing in the 0V connection between the two caps. If you make this your Star, you will be injecting all sorts of gunk into the low level signal wiring.
you must 'T' off the 0V connection between the caps and make you STAR ground at the end of the 'T'. 1 CM is all that it takes . . . .
Yes - a big difference. You have the capacitor chargingf currents flowing in the 0V connection between the two caps. If you make this your Star, you will be injecting all sorts of gunk into the low level signal wiring.
you must 'T' off the 0V connection between the caps and make you STAR ground at the end of the 'T'. 1 CM is all that it takes . . . .
PIXO,
those diagrams are theoretically right, in that it connects the correct wires to the correct nodes.
But, it is performance wise WRONG.
The explanation of the current routes in post109 is WRONG.
The PSU Zero Volts must NEVER be used as the Audio Ground.
Read the last two paragraphs of post 110.
Read Bonzai.
those diagrams are theoretically right, in that it connects the correct wires to the correct nodes.
But, it is performance wise WRONG.
The explanation of the current routes in post109 is WRONG.
The PSU Zero Volts must NEVER be used as the Audio Ground.
Read the last two paragraphs of post 110.
Read Bonzai.
Last edited:
Bonsai
I don't understand what you say.
What means "T" off?
Some of you guys have suggested through papers to have "between caps" the place where to "install" the star point but things changed...
You suggest not to use caps junction while papers suggests to do it.
Kind of mixed up..Sorry...
Luke
Soooooo...finally,
according to AndrewT & Bonsai...I was on the right track with the schematic on my post#103?
Right?
Or should I change something?
Bonsai...if you would have seen #103 you would have remarqued that the Star point is NOT on the caps's junction but connected from it like your DRW on 115.
Luke
according to AndrewT & Bonsai...I was on the right track with the schematic on my post#103?
Right?
Or should I change something?
Bonsai...if you would have seen #103 you would have remarqued that the Star point is NOT on the caps's junction but connected from it like your DRW on 115.
Luke
Last edited:
Hello
More than two weeks after...
I planified to implement a star-ground system in my amp.
Here is what I did:
Guess what? No improvement at all!
I tried to bring input Cable-Shied to Star...no!
Take all shield together and ground them on earth....BAD!
...all shield together on star...no improvement.
I think I faced all possibilities right now and think about putting input transformers on each input.
Luke
More than two weeks after...
I planified to implement a star-ground system in my amp.
Here is what I did:
An externally hosted image should be here but it was not working when we last tested it.
Guess what? No improvement at all!
I tried to bring input Cable-Shied to Star...no!
Take all shield together and ground them on earth....BAD!
...all shield together on star...no improvement.
I think I faced all possibilities right now and think about putting input transformers on each input.
Luke
Hi Luke, sorry you are having such a bad time. I am curious about the original problem. You said when all channels are hooked to the same tone generating source, that is when the problem occurs? Do you mean you are Y-jacking all the in's to one out? If so, have you tried different sources but still using all in's, like the analog out of a DVD/bluray/AVR? Same problem?
I understand your problem, believe me. I had a preamp give me so many noise problems so I tried a transformer out of desperation and guess what? It didn't work at all to fix my problem. I had to track down what the problem was. In my case it was a defective input jack which wasn't making good contact on the ground connection.
Something must be wrong it's a matter of tracking it. My friend found my problem for me by leaving everything on and hooked up at a low level while we experimented with attaching and disconnecting ground connections. All of a sudden he connected one and... sweet silence!
I understand your problem, believe me. I had a preamp give me so many noise problems so I tried a transformer out of desperation and guess what? It didn't work at all to fix my problem. I had to track down what the problem was. In my case it was a defective input jack which wasn't making good contact on the ground connection.
Something must be wrong it's a matter of tracking it. My friend found my problem for me by leaving everything on and hooked up at a low level while we experimented with attaching and disconnecting ground connections. All of a sudden he connected one and... sweet silence!
Hello shredhead
Who told you I have a bad time?
I have an excellent time...in fact I'm enjoying life like never before!!!
Nobody in the world can have a better time than me as justified by passion.
Do you believe this?
For a little less than $2K a fun project extends for months and months!
Anyway!
I've explained the problem in few of the 100 or more threads before but I'm becoming an expert for explainings.
To make it short: every channel has great noise figure when tested individually.
But if you dare connect two or more channels to a preamp or just paralleled together then buzzes and hum occurs to a totally unacceptable point.
This happens to any 2 CH. unbalanced power amp but six is another story, I'm telling you.
Frankly...?...I really consider a set of Jensen or Cinemag 10k-10k line Xfo's.
Galvanic isolation is in my point the only solution.
Check in the past replies as I tried everything.
I've had a lot of great suggestions and I tried them all but none worked, even at the end converting my ground system did not bring the solution...not even a glimpse of solution or improvement!
Luke
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.