Hi all,
I know this topic has been discussed a lot, but I am yet to across the specific issue that I am posting here. I am attaching the images here for visual aid in understanding .
Case 1> When I just implement a power amp with only volume control pot, there is no hum of any sort. Cant here any even if I put my ear to speaker.
Case 2> when I introduce a pre-amp (be it a tone control or active sub x-over), there is a lot of hum noise. I think it is probably caused by the loop that exists between the signal gnd of pre-amp , preamp , PSU.
My specific question would be, can I introduce a 10 ohm resistor inline at the pre-amp signal out GND wire to solve hum issue?
regards
Prasi
I know this topic has been discussed a lot, but I am yet to across the specific issue that I am posting here. I am attaching the images here for visual aid in understanding .
Case 1> When I just implement a power amp with only volume control pot, there is no hum of any sort. Cant here any even if I put my ear to speaker.
Case 2> when I introduce a pre-amp (be it a tone control or active sub x-over), there is a lot of hum noise. I think it is probably caused by the loop that exists between the signal gnd of pre-amp , preamp , PSU.
My specific question would be, can I introduce a 10 ohm resistor inline at the pre-amp signal out GND wire to solve hum issue?
regards
Prasi
Attachments
First question, if you remove the RCA connection from the source(CD player etc.) to the preamp, do you still have hum? that's is when the integrated amp is all by itself. I want to make sure it's not the ground from the source that cause the problem.
Assuming you have the hum even if the amp is alone without input connection. My first suspicion is your star ground. You have the power amp and preamp to the PSU, then from PSU to the star ground. This wire carry a lot of current from the output of the power amp to the speaker, through that wire before going back to the power amp. That can cause voltage bouncing of the whole PSU....which to the preamp.
I don't understand your PSU. seems like it's only consists of the bridge rectifiers and filter caps. I would have the star ground right on the CT of the toroid transformer. All your ground of the filter caps should go to the star ground. Your two speaker return should go to the star ground. Then both the power amp and preamp ground goes to the star ground also. There should be no ground wire from the power amp to the speaker binding post( only the output wire goes to the speaker binding post, the speaker return goes back to the star ground).
As for isolation between the grounds of the power amp and preamp, I would have a 3.3ohm on the power amp side to the ground. So the grounds between the preamp and power amp is isolated by the 3.3ohm.
Just a suggestion.
Assuming you have the hum even if the amp is alone without input connection. My first suspicion is your star ground. You have the power amp and preamp to the PSU, then from PSU to the star ground. This wire carry a lot of current from the output of the power amp to the speaker, through that wire before going back to the power amp. That can cause voltage bouncing of the whole PSU....which to the preamp.
I don't understand your PSU. seems like it's only consists of the bridge rectifiers and filter caps. I would have the star ground right on the CT of the toroid transformer. All your ground of the filter caps should go to the star ground. Your two speaker return should go to the star ground. Then both the power amp and preamp ground goes to the star ground also. There should be no ground wire from the power amp to the speaker binding post( only the output wire goes to the speaker binding post, the speaker return goes back to the star ground).
As for isolation between the grounds of the power amp and preamp, I would have a 3.3ohm on the power amp side to the ground. So the grounds between the preamp and power amp is isolated by the 3.3ohm.
Just a suggestion.
Last edited:
It would be difficult to think of a worse place to put a clean audio signal star ground than right in the middle of the dirty PSU reservoir capacitor charging pulse loop.Alan0354 said:
Only slightly better is to put the star ground on the chassis, shared with the safety ground, which is what the diagrams appear to show.
Get rid of the loops; don't try to patch them up with 'loop breakers'. I am tempted to say get rid of the star too - two circuits in one box with a common PSU naturally indicate a bus ground, not a star ground. The ground route should mimic the signal route.
The "ground" routes you have shown should read Signal Return.
Every Signal Hot MUST have a Signal Return.
ALWAYS keep the signal Hot and Return close coupled and preferably a twisted pair, or a screened twisted pair, or a coax.
NEVER separate the signal Hot from it's Return. This faulty assembly technique is guaranteed to pick up interference.
Hi Andrew,
I read the document mentioned in your post. However I need to go over it in more detail to understand completely and its implementation in my case.
Hi DF96,
I have attached another scheme which I dont think would be practical as per your suggestion to mimic signal routes. Any way here it goes.
TO others,
Also does any one have grounging schemes for the above mentioned config. which they have tried and works without any glitch?
the config is as follows,
1. One toroid with 2 secondary o/p (one for amp and one for pre-amp.
2. single PSU supplying power to both main amp and pre-amp (tone control/x-over).
3. single stereo amp PCB with common return (power) for both left and right channels
4. seperate signal returns for left and right channels
reg
prasi
Attachments
I better draw the circuit out.
1) I put the star at the CT. I put the star as shown so the noise of the charging pulse going back to the CT of the transformer. This is where the ground goes to preamp, power amp and speaker returns.
2) The key is the interface between the preamp and power amp. I represent the power amp as opamp. Notice The opamp is referenced to the ground provided by the preamp, not the ground of the power amp. The preamp has the signal output and the ground return. Both the +ve and -ve input of the opamp(power amp) is REFERENCED to the ground of the preamp. Resistor R1 is to break the ground loop between the power and preamp.
The star ground is referenced to the earth ground that has no current between the amp ground and earth ground. Large current return from the speaker going back to the star. Any noise on the star is common mode to to the power and preamp.
1) I put the star at the CT. I put the star as shown so the noise of the charging pulse going back to the CT of the transformer. This is where the ground goes to preamp, power amp and speaker returns.
2) The key is the interface between the preamp and power amp. I represent the power amp as opamp. Notice The opamp is referenced to the ground provided by the preamp, not the ground of the power amp. The preamp has the signal output and the ground return. Both the +ve and -ve input of the opamp(power amp) is REFERENCED to the ground of the preamp. Resistor R1 is to break the ground loop between the power and preamp.
The star ground is referenced to the earth ground that has no current between the amp ground and earth ground. Large current return from the speaker going back to the star. Any noise on the star is common mode to to the power and preamp.
Attachments
NO !
keep EVERY Signal Return Close coupled with it's Signal Hot.
DO NOT use the PSU smoothing caps junction as a voltage reference. NEVER !!!!!!!!!
keep EVERY Signal Return Close coupled with it's Signal Hot.
DO NOT use the PSU smoothing caps junction as a voltage reference. NEVER !!!!!!!!!
If you put small resistances into the wires in that schematic, the problems with that arrangement will be clearer.
The pulse current return directly back to the CT, it does not affect the rest of the circuit. the star is referenced to the earth ground.
Draw out the current path of the filter cap and you'll see.
Do what SY suggests.
A perfect star (i.e. a single point - about the size of an electron) could be included in a superconducting charging pulse loop. The rest of us have to keep them separate, unless we want buzz with our music.
Also, the safety ground and signal ground are best kept apart - apart from a single connection between them. Don't put the 'star' ground on the chassis! Don't put the star ground right in the PSU!
Remember: a star ground is a (misleading) theoretical concept. All real grounds are busses.
A perfect star (i.e. a single point - about the size of an electron) could be included in a superconducting charging pulse loop. The rest of us have to keep them separate, unless we want buzz with our music.
Also, the safety ground and signal ground are best kept apart - apart from a single connection between them. Don't put the 'star' ground on the chassis! Don't put the star ground right in the PSU!
Remember: a star ground is a (misleading) theoretical concept. All real grounds are busses.
Hi all,
I forgot to mention here (though the diagrams show it) that the signal gnd and power gnd are connected through a trace both on amplifier PCB as well as pre-amp PCB.
hope this will help in understanding the scheme.
the hum is such ***, almost takes the joy out of DIY;-(
reg
prasi
I forgot to mention here (though the diagrams show it) that the signal gnd and power gnd are connected through a trace both on amplifier PCB as well as pre-amp PCB.
hope this will help in understanding the scheme.
the hum is such ***, almost takes the joy out of DIY;-(
reg
prasi
Last edited:
I would ditch the 'star' ground. Let the preamp gets its DC 'ground' supply via the return of the signal connection to the amp.
You need to take a deep breath and rethink the grounding. Don't look for a recipe or rules. Think about where the currents flow (always, without exception, in loops). This will tell you where the voltages are developed. Keep unwanted voltages away from signals.
When we draw a circuit most people automatically draw it with a ground bus (often doubling as the negative supply). Why is it that when implementing a circuit most people automatically introduce a 'star' ground - which is actually a short unplanned bus? Then wonder why they get hum?
You need to take a deep breath and rethink the grounding. Don't look for a recipe or rules. Think about where the currents flow (always, without exception, in loops). This will tell you where the voltages are developed. Keep unwanted voltages away from signals.
When we draw a circuit most people automatically draw it with a ground bus (often doubling as the negative supply). Why is it that when implementing a circuit most people automatically introduce a 'star' ground - which is actually a short unplanned bus? Then wonder why they get hum?
Hi all,
went through the following article http://www.diyaudio.com/forums/diya...udio-component-grounding-interconnection.html here on diyaudio.
I have made a diagram based on my understanding of the article. the same is attached.
any suggestions/ corrections are welcome.
reg
prasi
PS. if this doesn't solve the problem, better I give up diy!!
went through the following article http://www.diyaudio.com/forums/diya...udio-component-grounding-interconnection.html here on diyaudio.
I have made a diagram based on my understanding of the article. the same is attached.
any suggestions/ corrections are welcome.
reg
prasi
PS. if this doesn't solve the problem, better I give up diy!!
Attachments
Stop and think.
The RCA input Jack feeds the Pre-amp.
That must be a pair of close coupled wires.
One wire is the Signal Hot the other wire is the Signal Return.
Both wires must run together from RCA to pre.
Then from pre to power amp.
Again TWO close coupled wires.
Both these active parts have a power input. That is NOT one wire. It is a Flow and Return and these are close coupled.
If the supply is dual polarity then it becomes a three wire connection: +,-,return, twisted as a triplet.
Finally you have power amp to speaker terminals. This is a TWO wire connection. It is NOT a one wire connection.
Remember what I said earlier EVERY circuit is two wires. Flow/Hot and Return
The RCA input Jack feeds the Pre-amp.
That must be a pair of close coupled wires.
One wire is the Signal Hot the other wire is the Signal Return.
Both wires must run together from RCA to pre.
Then from pre to power amp.
Again TWO close coupled wires.
Both these active parts have a power input. That is NOT one wire. It is a Flow and Return and these are close coupled.
If the supply is dual polarity then it becomes a three wire connection: +,-,return, twisted as a triplet.
Finally you have power amp to speaker terminals. This is a TWO wire connection. It is NOT a one wire connection.
Remember what I said earlier EVERY circuit is two wires. Flow/Hot and Return
Last edited:
Hi Andrew,
What you say is true. both pre and power amps have dual polarity supply. If I connect the power return of pre also, there is a loop formed as shown in the attachment(red dashed line). How to get rid of this loop?
Now for the speaker return, Both the PSU and Power amp can be though of as a single PCB having a bus connection for return path. And hence both the returns are connected to the PSU return terminal.
regards
Prasi
Attachments
here is an update!,
I tested with various config. there were two kinds of noises in each configurations. One is low freq hum which was quite loud and another is a high freq noise but quite low in loudness.
The problem was with the extension board that I was using. its earth pin wire was damaged . when I replaced the extension board with a new one (tested for continuity of earth pin), the low freq hum is gone. I mean vanished totally, but the high freq noise is still heard only when I put my ear to the speaker. dont know what else could be the problem.
regards,
Prasi
I tested with various config. there were two kinds of noises in each configurations. One is low freq hum which was quite loud and another is a high freq noise but quite low in loudness.
The problem was with the extension board that I was using. its earth pin wire was damaged . when I replaced the extension board with a new one (tested for continuity of earth pin), the low freq hum is gone. I mean vanished totally, but the high freq noise is still heard only when I put my ear to the speaker. dont know what else could be the problem.
regards,
Prasi
Cut all the grounding wires that go back to the PSU.
Draw in ALL the twisted pairs that run between the modules.
Draw in the two twisted triplets between PSU and the active modules.
Now look carefully at what you have and compare to the theoretical SCH. What voltage reference connections are missing? Insert them/it
Then read D.Joffe to see where the loop current attenuating resistors need to go without increasing the interference voltage superimposed on the wanted audio signal.
This same problem occurs in ALL multichannel amplifiers. Only mono block topology avoids the loops.
Draw in ALL the twisted pairs that run between the modules.
Draw in the two twisted triplets between PSU and the active modules.
Now look carefully at what you have and compare to the theoretical SCH. What voltage reference connections are missing? Insert them/it
Then read D.Joffe to see where the loop current attenuating resistors need to go without increasing the interference voltage superimposed on the wanted audio signal.
This same problem occurs in ALL multichannel amplifiers. Only mono block topology avoids the loops.
When you test for him/ noise do the following
1. Without any inputs - i.e nothing plugged in at all - your amplifier should be absolutely silent
2. Plug your cables in but do not connect them to anything the other end. It should be absolutely silent
3. Next, short the left and right signal input out at the end of the input cable with a short piece of wire. It should still remain absolutely silent.
4. Finally, plug into your source (I assume you are using a preamp) - it should still be silent.
D. Joffe is a very good source - I also recommend it.
1. Without any inputs - i.e nothing plugged in at all - your amplifier should be absolutely silent
2. Plug your cables in but do not connect them to anything the other end. It should be absolutely silent
3. Next, short the left and right signal input out at the end of the input cable with a short piece of wire. It should still remain absolutely silent.
4. Finally, plug into your source (I assume you are using a preamp) - it should still be silent.
D. Joffe is a very good source - I also recommend it.
- Status
- Not open for further replies.