[holland]

the trabsformers are r core, there is one under the left channel and one under the right , no the star g is not connected to earth. should i try ampheads mod?

 
well, you're clearly getting noise from proximity to the transformers.
 
you should earth the amp chassis for safety reasons.  what is it connected to, since you're doing star grounding?  the chassis is connected to the source's shield?  and if your amp faults with nothing plugged in, what happens?
 
you can try ampheads mod.  i use dc heaters only, but i don't think this is relevant.  worth a try any way.

 

[yiancar]

ive connected earth to chassis , noise still there. amohead mod tomorow since i need the resistors, anything else i can try? or my only chance is moving the amp further appart from the psu?

 

[holland]

ive connected earth to chassis , noise still there. amohead mod tomorow since i need the resistors, anything else i can try? or my only chance is moving the amp further appart from the psu?

 
earthing the chassis is for safety and grounding issues.  neither is the noise you are talking about.
 
distance, more or less, if that's what you have it isolated down to.  it's unlikely you'll be able to do anything else except shield your enclosures with mu-metal.  you can try to put a carbon steel enclosure around the transformers to see if it helps.
 
psu in a separate chassis is generally a sucky idea.  people always try to stack, and some transformers radiate in height more than they do on a parallel plane.

 

[yiancar]

tried everything i could thing of, even a steel sheet (3mm thick) on top of the transformer did not help. any other suggestions ?

 

[nikongod]

Have you moved the tubes from one channel to the other? 
 
If the hum follows the tubes you have a microphonic tube. Get a different tube to test that it is in fact the problem, then crush the bad tube with a brick. 
 
While DC heaters will not (usually) set off microphonic tubes they are really just a bandaid for the real problem of microphonic tubes... Fixing the real cause of the issue (the tube) is so much moar better than changing the heater. 
 
I guess both is sort of nice, but usually so far overkill as to be relegated to commercial amps where its better to have a happy customer with a dodgy tube than an amp that hums out of the box, but only when your customer uses the best tube he owns... Whats wrong with your amp anyways? 
 
Quote:

  psu in a separate chassis is generally a sucky idea. 

 
Tubes on long leads is another.

 

[balderon]

Unlikely but wanted to ask. Is it possible some of the hum is from the proximity of the transformer to the power supply umbilical cord? Is it shielded?

 

[yiancar]

Tubes on long leads is another.

will check the tubes as soon as i go home, i thought that the extention cable of the tubes may picking up noise i moved the tubes around and i get no difference so i guess i can rule that out.

i get noise / hum on the left channel ( balanced amp btw) which is not affected by changing the volume. when i move the amp case off the psu case the noise dissapears . i use r core transformers and the transformers are directly below one of the two amp bored of each channel

Unlikely but wanted to ask. Is it possible some of the hum is from the proximity to the transformer power supply umbilical cord? Is it shielded?

no its not shielded. i cannot understand how noise is picked up only from the right channel, i mean both the psu and amp are symmetrical.

 

[nikongod]

After seeing the quoted passage below I dont think that the long leads are the cause of the issue. 
 
Quote:

i get noise / hum on the left channel ( balanced amp btw) which is not affected by changing the volume. when i move the amp case off the psu case the noise dissapears . i use r core transformers and the transformers are directly below one of the two amp bored of each channel
no its not shielded. i cannot understand how noise is picked up only from the right channel, i mean both the psu and amp are symmetrical.

 
So when you move the amp away from a device that one could safely assume radiates EMF the noise goes down.
 
Not much question what is wrong.
 
Build a long umbilical, like you should have as the only real reason to build an external PSU, and put the PSU 6ft from anything you care about.
Alternately you could try lining the lid of the PSU with mu-metal foil, but for the cost I would not bother. 
 
Why one channel and not the other: one of your transformers leaks more flux than the other. 

 

[duncan1]

You have certainly tried out a variety of suggestions . In my own design 15W RMS class A  mono-block stereo amp with separate -outboard power supplies[two off] with separate umbilical cords . I have NO hum and thats using standard transformers. In a letter to Electronics World I said that [by experiment] I found that if I lifted the centre tap of the mains transformers so that they werent terminated in the power supply boxes but ran [floated] through   each cord to the star earthing that I had located directly underneath each amps  PCB [for a very short run to star earth for each component requiring a star earth] on the PCB. This worked amazingly well for me and got approval from some EEs in the mag.I do not agree with the point of -external PSs are not as good as internal . That conflicts with my own findings into them as I have found a PS that is not near a power amp is much better for lower noise  / by induction/radiation. You could  if needed separate the star earth return wires from the sensitive input to output by a 10 ohm resistor but they still both go back to the main star earth. I didnt need to do that. I proved this works and up to now nobody has challenged that. But knowing Head-Fi somebody will --as long as they understand/realise it WORKS! I did not need to screen the cable.        

 

[holland]

  external PSs are not as good as internal . That conflicts with my own findings into them as I have found a PS that is not near a power amp is much better for lower noise  / by induction/radiation. You could  if needed separate the star earth return wires from the sensitive input to output by a 10 ohm resistor but they still both go back to the main star earth. I didnt need to do that. I proved this works and up to now nobody has challenged that. But knowing Head-Fi somebody will --as long as they understand/realise it WORKS! I did not need to screen the cable.        

 
They are sucky because people try to stack, which is a really bad reason to do it.  I run external supplies now and then, not PSUs, just the transformer over a longish distance from the amp.  I always have regulation very close to the amp itself.  I run the secondaries in an umbilical.  If you run regulation over long distances, it generally defeats the purpose because the longer the leads the more "stuff" it can pick up along the way, higher resistances, etc. etc.
 
His problems are not ground loop related.  I always use a ground loop breaker, some say it's due to bad practice, but I like nodes/zones of ground "planes" or loops.  i.e., star ground nodes.

 

[duncan1]

People try to stack them ? If you mean  put them straight on top of the amp. I would have though after posters had been on Head-fi a while they would know that -that is one sure way of causing trouble. You would then be as well without an outboard PS due to closeness of the transformers.and the heat would cause components to overheat and cause instability. Anybody with a bit of common sense would put them on a lower shelf with a foot spacing.  And thats what I did. As far as putting the regulators  on the PCBs all the amps I built had heavy smoothing at the point where the outboard PS contacted the PCB. I am not alone in this some very up market UK designs do the same.If you extend the secondary coil side of the mains transformers via the umbilical cord you then have  a coil circuit that can pick up interference. Whereas if all the rectifiers and regulators  are on the transformer box then the output impedance is much lower with less likely hood of induced/ radiated  pick up >If his problem is not ground related then why have the previous suggestions not worked?  and have you the answer?

 

[holland]

  And thats what I did.

 
And that's not what the other poster did (he stacked), and hence the comment, and your reference to it....which is completely tangential.  Read it again.  His problem is that he stacked and is having flux induced noise.  He noted moving the PS section away, the noise went away.  There's not much to discuss.
 
BTW, white space will make your posts more readable.

 

[AnakChan]

Chaps, I've just cleared up a bunch of off-topic bickering. Honestly if you chaps just can't conduct a mature conversation, please just block each other.

 

[bhjazz]

look at (and reheat, if necessary) Quote:

   
after the amp is left to complete cool down also the right channel experiences hum, 1-2 minutes powered up and the right channel hum disappears.

 
That's a cold solder joint, pure and simple.  I know I've done it myself.  While you're working through some of the other proximity tests, take some time to look at (and reheat, if necessary) as many solder joints as possible.  Do you have a magnifier?  10x would do it.  Check them all.  
 
There are multiple things going on with your build.  My suggestion is only a small part of the fix for your original problem, but you'll need it eventually and it will certainly fix your right channel issues.  You've got to start with solid connections before you can get to these other fixes.  

 

[RAZRr1275]

Would someone build one of these for me? I'm not looking for one right now but closer to Christmas I'd like to work something out. Feel free to pm me if interested. I will pay of course.