[funch]

Is earth ground tied to the chassis? If not, you might try that. Also look into a ground loop breaker.

 

[stixx]

'But carry on, I really need to stop posting here'

No need to get upset, I didn't say that your suggestions were wrong. You have your findings based on your experience building gear, and so do I. And in about 15 headphone amplifier, chip amp and DAC builds I never had to deal with even the faintest of hums (but one--> see my first post here), despite the transformer being close to the circuit.
Again, in this special case I might be wrong.

 

[jdkJake]

I would also recommend you try the alternate wiring scheme for the signal ground as outlined on Alex's site.

I personally had better success with the alternate wiring scheme.

 

[Sathimas]

Is earth ground tied to the chassis? If not, you might try that.

 
Earth ground is tied to the star ground - which is of cause connected to the
bottom of the case, since the screw goes through.
 
 
I just tried varying the distance of the torroids.
Had the heater-torroid behind the back panel and the other one standing
upright to the back. Even moved them closer and back off while listening.
No audible change ...
 
Going to try the alternate wiring now.
Connecting all those tube-socket wires is a pain in the ***

 

[Sathimas]

Let's go on with the one man show
Think you're all sleeping in other time zones
 
 
Alternate wiring gave me a huge step forward.
 
 
Only a faint hum is left - now equally loud on both channels.
 
 
Just now I'm listening to the first tunes of Master of Puppets from the S&M Album
with my ATH W1000. As soon as music is playing, I can't detect any disturbing noise.
 
Yet it's there - might an ground loop breaker solve that problem?
 
 
Another problem is that the amp happily picks up noise from my computer.
At the moment the source is my gamma1dac. I still have to try out some other
place for it, next to the external harddrive seems not that good :-D
 
 
I'm also happy having the gain at 5, 8 would have been to high.
(Just using high-efficiency cans as the ATH, Beyerdynamic T90, DT1350)
 
 
And - by the way - sound is great, as expected 

 
 
Thanks to everyone helping me here, and especially to Alex, for sharing this
great amp design with the DIY-community.

 

[jdkJake]

Alternate wiring gave me a huge step forward.


Only a faint hum is left - now equally loud on both channels.

Yet it's there - might an ground loop breaker solve the problem?

I would advise you first verify the hum is not internally generated before moving outwards.

Be advised, I would NOT use your good headphones for ANY initial testing. There are plenty of cheap, disposable headphones on the market for this task.

First, ground the inputs. I like to use a pair of shorted jacks (ground and signal connected), dressed with a piece of heat shrink.




Then, using your initial test headphones, see if you still hear a hum. Start with the signal fully attenuated and work your way up. Ideally, everything should be very much hum free. If there is hum, then something internally is amiss or some internal component is picking up on some external component inducing the hum.

After this step you can start to add external components and consider the use of a ground loop breaker.

 

[Sathimas]

I used some cheap earbuds for initial testing, no need to worry.

I'll try what you suggested later this evening.

 

[Sathimas]

The hum is still there with grounded inputs, but not changing with volume.

 

[Sathimas]

Thought about the CLB now.
 
I also found some other thread here in the forum about that.
(Also read some facts on AMBs website.)
 
Here's what I think I should do:

 
 
Red are the ground wires, all connected to star ground, which is tied to earth ground.
Yellow is input-ground.
 
 
Are my thoughts correct?

 

[stixx]

Your image is really tiny this time... hard to decipher anything

 

[Sathimas]

Image is 2000px wide - just click on it

 

[muskyhuntr]

Hi,
A couple more suggestions.
 
1) Look on page 101 of this thread and you will see some pictures of my build.  Note that I have only 2 wires going to a "star ground" (from SG on the input connector) and those 2 are going through a ground loop breaker before going to chassis ground..  Mains ground is tied directly to the chassis. Your build has 7 wires.  All these are possible points to inject hum into your project..  My only complaint with this amp is that there is too much external wiring to the board.
2) The left 2 RCA's on your build are wired to ground, though the positives are not wired.  Try disconnecting those for the time being.
3) Take one channel out of the equation.  Disconnect every thing related to that channel. Trouble shoot just one channel at a time.  Make sure nothing from that channel is connected. Start from the beginning.    Remove all the major possibilities out of the case (transformers, switch wiring, etc).  Remove all your ground wiring and then re attach one at a time and in different combinations.  Are you twisting your wires?  If you are trouble shooting without a source, follow jdkJake's suggestion and short your inputs.  Have you tried swapping the tubes to see if louder hum moves with the change?  Once you solve the problem on one channel, work on the other.
4) On your next build, mount everything on a piece of plywood and trouble shoot from there.  You can see where Stixx and I did this earlier in the thread.
 
Keep at it, this amp is worth it!
Jim

 

[stixx]

"Image is 2000px wide - just click on it "

Really strange... I did click it... still tiny. Guess my computer here at work is blocking something

 

[Sathimas]

Hi,
A couple more suggestions.
 
1) Look on page 101 of this thread and you will see some pictures of my build.  Note that I have only 2 wires going to a "star ground" (from SG on the input connector) and those 2 are going through a ground loop breaker before going to chassis ground..  Mains ground is tied directly to the chassis. Your build has 7 wires.  All these are possible points to inject hum into your project..  My only complaint with this amp is that there is too much external wiring to the board.

 
Let's clarify some vocabulary first, since all those different "grounds" confuse me:
 
Mains Ground
Earth ground
Chassis Ground
Star Ground
 
I think mains ground = earth ground on IEC
 
Chassis ground = connection to metal bottom of the case.
Im my case, star ground is connected to chassis ground.
 
Where did you connect the grounds of the heater and amp power supplies
if you have only wired the SG-points on the amp boards?
(I can't make it out on the pictures)
 
Ground of power supply and amp board seems directly wired,
I routed those wires to star ground and back, according to the scheme
on the website.
 

2) The left 2 RCA's on your build are wired to ground, though the positives are not wired.  Try disconnecting those for the time being.

 
I'll try - but I'll need them as soon as I have the e12 here.
 

3) Take one channel out of the equation.  Disconnect every thing related to that channel. Trouble shoot just one channel at a time.  Make sure nothing from that channel is connected. Start from the beginning.    Remove all the major possibilities out of the case (transformers, switch wiring, etc).  Remove all your ground wiring and then re attach one at a time and in different combinations.  Are you twisting your wires?  If you are trouble shooting without a source, follow jdkJake's suggestion and short your inputs.  Have you tried swapping the tubes to see if louder hum moves with the change?  Once you solve the problem on one channel, work on the other.

 
1. wires from the 30V torroid are braided, those from the 9V are not.
Input wiring is fully shielded
2. Tubes don't make a difference. As I said before, there's no difference betwenn
the channels anymore since I use the alternate wiring scheme.
 

4) On your next build, mount everything on a piece of plywood and trouble shoot from there.  You can see where Stixx and I did this earlier in the thread.

 
Did this with the amps before this one, but since I never had
those problems before I was kind of too optimistic here.

Keep at it, this amp is worth it!
Jim

 
I definitley will!

 

[UKToecutter]

Whilst I hope I won't be shot down in flames.....
 
I believe SG is Signal Ground.
Star Ground is the designator of the topology where grounds are 'tied' together at a single point to ensure there is no potential difference between any of the grounds.