Hello
I just finished building a pre and a power amp. The problem is that upon connecting the two there is a buzz. I've done all the usual procedures to eliminate any earth loop. The strange thing is that the buzz only shows when the signal cable from the pre-amp is connected to the power amp. There is a phones output on the pre and it's completely silent. Also, when the two are not connected, the power amp is also extremely quiet. One last detail is that the buzz is only present when there is soemthing connected to the input, so if one channel is connected but the other one is not, the connected one produces buzz but the other doesn't. If this was an earth loop the buzz should be always present right? Any ideas??
The amps are based on the LM3886
I just finished building a pre and a power amp. The problem is that upon connecting the two there is a buzz. I've done all the usual procedures to eliminate any earth loop. The strange thing is that the buzz only shows when the signal cable from the pre-amp is connected to the power amp. There is a phones output on the pre and it's completely silent. Also, when the two are not connected, the power amp is also extremely quiet. One last detail is that the buzz is only present when there is soemthing connected to the input, so if one channel is connected but the other one is not, the connected one produces buzz but the other doesn't. If this was an earth loop the buzz should be always present right? Any ideas??
The amps are based on the LM3886
I have the same problem when connecting my VA versa preamp to my GC. When they are not connected, my GC is perfectly quiet. Could it be the grounding in my pream? When I connect a cd player with a line out and volume built into the cd player I don't get the buzzing sound. One thing though, the buzzing is really low level, you have to put your ear very close to the speaker to hear it. It does not increase in sound when you turn the volume up either.
A few checks:
- are the input connectors isolated in the box?
- (How) is the ground of the amp connected to the box?
- does the ground from the inputs follow the way: input ==> signal ground ==> power ground?
- do you have a cable tv or -radio connected? If yes, remove the antenna cable for test. If it helps, you must insert a small device in the antenna cable to break the ground loop (in german: Mantelstromfilter, it is a simple DC-blocker for the ground).
- are all devices (even this one before the pre) connected to the same mains connector?
- are all devices grounded?
If all fails, then you could open the ground at the cables just before the pre and the amp and use ONE separate ground connection between (all) devices (like a ground rail). Be careful, not to build a loop and to have at least one device (pre?) connected to earth via mains.
For my experience, the chipamps based on LMchips are very resistive against this style of hum, because the high PSRR. I have a configuration with my PC, TV-cable (a classical ground loop), where every commercial amp produces hum, but not the chipclones!
Franz
P.S.
all above assuming, your pre is O.K.!
- are the input connectors isolated in the box?
- (How) is the ground of the amp connected to the box?
- does the ground from the inputs follow the way: input ==> signal ground ==> power ground?
- do you have a cable tv or -radio connected? If yes, remove the antenna cable for test. If it helps, you must insert a small device in the antenna cable to break the ground loop (in german: Mantelstromfilter, it is a simple DC-blocker for the ground).
- are all devices (even this one before the pre) connected to the same mains connector?
- are all devices grounded?
If all fails, then you could open the ground at the cables just before the pre and the amp and use ONE separate ground connection between (all) devices (like a ground rail). Be careful, not to build a loop and to have at least one device (pre?) connected to earth via mains.
For my experience, the chipamps based on LMchips are very resistive against this style of hum, because the high PSRR. I have a configuration with my PC, TV-cable (a classical ground loop), where every commercial amp produces hum, but not the chipclones!
Franz
P.S.
all above assuming, your pre is O.K.!
There could be a problem with your preamp.
I recently made changes to my active crossover and preamp that caused the combination to oscillate. When connected to the BrianGT gainclone there was a strong buzz. When I got rid of the oscillation, the buzz went away.
This does not make sense to me because the oscillation was likely very high frequency, certainly above audible. I know this because it was unhearable on another amp that is just 2 prong (double insulated) without an earth ground. That amp did heat up noticably though when connected to the crossover.
But when connected to the Gainclone there was a strong buzz until I fixed the oscillation. It doesn't make a lot of sense to me that a high frequency large input signal would result in a lowish buzz but that is what happened. Perhaps it was the chip protection circuit kicking in, in some way? But normal music sounded Ok through it, even though there was the constant level buzz overlayed on it.
Can anyone suggest how a strong high frequency signal from oscillation could induce a buzz in the amp?
I recently made changes to my active crossover and preamp that caused the combination to oscillate. When connected to the BrianGT gainclone there was a strong buzz. When I got rid of the oscillation, the buzz went away.
This does not make sense to me because the oscillation was likely very high frequency, certainly above audible. I know this because it was unhearable on another amp that is just 2 prong (double insulated) without an earth ground. That amp did heat up noticably though when connected to the crossover.
But when connected to the Gainclone there was a strong buzz until I fixed the oscillation. It doesn't make a lot of sense to me that a high frequency large input signal would result in a lowish buzz but that is what happened. Perhaps it was the chip protection circuit kicking in, in some way? But normal music sounded Ok through it, even though there was the constant level buzz overlayed on it.
Can anyone suggest how a strong high frequency signal from oscillation could induce a buzz in the amp?
I had two issues when I first put my GC together causing buzz. . . One is that I had switched signal and ground wiring the inputs. This mostly gave me trouble if I only had one channel attached. The other is that I discovered one of my cables was in bad shape. I don't know what was wrong with it - not shorted out, but there seemed to be some signal bleed.
I'm assuming most folks know better than to make the first mistake I made (heck, *I* knew better and I did it anyhow).
C
I'm assuming most folks know better than to make the first mistake I made (heck, *I* knew better and I did it anyhow).
C
Well, I have no digital camera so there is no way I can photograph the amp... but it wouldn't do any good since it is now quite a mess.
So, here is the schematic of the present situation. The amp's power cable is connected to the pre-amp where a single switch turns on or off both. So there is only one connection to the mais. There is no more noise present when another appliance is connected to the pre-amp's input. The noise is independent of volume setting. The phones outpu on the pre is completely silent so I'm guessing the problem is with the power amp. Even still I've did some changes in the grounding of the pre-amp and there were differences although there was never buzz or hum from the phones output.
I have no idea of what's going on... I didn't use a zobel network at the amp's output. maybe there is instability since the buzz only appears when something mettalic touches the input pins of the input connector. The amp has no metal casing and the only shielding are the heatsinks.
The input ground used to be connected to the power supply caps ground but it was more noisy.
So, here is the schematic of the present situation. The amp's power cable is connected to the pre-amp where a single switch turns on or off both. So there is only one connection to the mais. There is no more noise present when another appliance is connected to the pre-amp's input. The noise is independent of volume setting. The phones outpu on the pre is completely silent so I'm guessing the problem is with the power amp. Even still I've did some changes in the grounding of the pre-amp and there were differences although there was never buzz or hum from the phones output.
I have no idea of what's going on... I didn't use a zobel network at the amp's output. maybe there is instability since the buzz only appears when something mettalic touches the input pins of the input connector. The amp has no metal casing and the only shielding are the heatsinks.
The input ground used to be connected to the power supply caps ground but it was more noisy.
Attachments
At one point when I had my gainclone initially assembled on a piece of wood I had a symptom that sounds a bit like yours. No noise until I connected inputs.
It turned out that the fix for me was to change from twisted wires connecting the input jacks to using shielded cable for both the + and minus signal ins and grounding the shield from both ends.
To clarify: Each input connection from the jack to the gainclone now uses two shielded cables, with each shield having a wire at both ends from the shield to ground. So the shields carry no input signal. They are purely shielding. There are those who would argue that grounding on one end is the right way and others who will argue that both ends works in more cases.
I do not know if I really would have needed this when i got it into a metal box, because I used the grounded shielded whan I put it into the metal box and went to shorter length connections.
Others do not seem to need to do the grounded shielded. But I have radio towers of various kinds within a short distance and it seems that I have to do more RF protection than most.
It turned out that the fix for me was to change from twisted wires connecting the input jacks to using shielded cable for both the + and minus signal ins and grounding the shield from both ends.
To clarify: Each input connection from the jack to the gainclone now uses two shielded cables, with each shield having a wire at both ends from the shield to ground. So the shields carry no input signal. They are purely shielding. There are those who would argue that grounding on one end is the right way and others who will argue that both ends works in more cases.
I do not know if I really would have needed this when i got it into a metal box, because I used the grounded shielded whan I put it into the metal box and went to shorter length connections.
Others do not seem to need to do the grounded shielded. But I have radio towers of various kinds within a short distance and it seems that I have to do more RF protection than most.
Well... I'll try shielding the input cable at both ends. But in the mean time I've done some more testing and the problem is at the input because I connected an input directly to the input on the amp's pcb and the buzz was really low. Maybe it is still there because now the grounding is a complete mess with all the tests. There may be a problem with the shielding or something. The input is very close to the transformer, but there is no way it can't be so close 
so there will probably always be some noise.
so there will probably always be some noise.
Before I changed to shielded cables on the input wiring I also added the 220pF caps across the input+ and input- as recommended in this post:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=22946
It made a noticeable difference but I still ended up needing to do the shielded. Will likely only help if your problem is RF related, but is worth trying.
I originally grabbed a cheap interconnect cable and used it for wire. It was coaxial but the outer copper was fairly sparse. It made some difference that was encouraging. When I used better shielded cable with aluminum foil + braided copper, things got quite quiet unless I get very close (10" from the speaker). This was all with the parts mounted on an open wood board.
With 1500uF as the power supply caps, this is about as good as I expect, and what I have now with it all in an aluminum box.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=22946
It made a noticeable difference but I still ended up needing to do the shielded. Will likely only help if your problem is RF related, but is worth trying.
I originally grabbed a cheap interconnect cable and used it for wire. It was coaxial but the outer copper was fairly sparse. It made some difference that was encouraging. When I used better shielded cable with aluminum foil + braided copper, things got quite quiet unless I get very close (10" from the speaker). This was all with the parts mounted on an open wood board.
With 1500uF as the power supply caps, this is about as good as I expect, and what I have now with it all in an aluminum box.
Well, I've tried putting a unetched pcb connected to ground between the transformer and the input/output connectors but it didn't made a difference.
Also, each amp has 220pF caps between the inputs.
The power supply consists of a single 400VA transformer and 3 sets of bridge + caps. The 3 different power supply grounds are all connected to the transformer center tap.
There is still buzz....
I'm getting tired of this...
Also, each amp has 220pF caps between the inputs.
The power supply consists of a single 400VA transformer and 3 sets of bridge + caps. The 3 different power supply grounds are all connected to the transformer center tap.
There is still buzz....
I'm getting tired of this...Obrigado ("thanks" for the non-portuguese speaking)
This is the present situation. I've also tried to connect the signal ground to the PSU star point but there was no significant improvement. The shield is only connected on the pre-amp side. The amps mains earth that connects through the pre-amp is presently disconnected. When connected to an amp's pcb GND there is a slight reduction in noise but in one of the pair of amps the noise gets justs different and not reduced really strange behaviour... I think there is instability in one of the other pair of amps since that particular heatsink gets warmer than the others.
If needed I can also post the pre-amp's gnd arrangement.
("thanks" for the non-portuguese speaking)This is the present situation. I've also tried to connect the signal ground to the PSU star point but there was no significant improvement. The shield is only connected on the pre-amp side. The amps mains earth that connects through the pre-amp is presently disconnected. When connected to an amp's pcb GND there is a slight reduction in noise but in one of the pair of amps the noise gets justs different and not reduced
really strange behaviour... I think there is instability in one of the other pair of amps since that particular heatsink gets warmer than the others. If needed I can also post the pre-amp's gnd arrangement.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.