[Tabascosauce]

Most pictures I've seen show the trafos mounted at 45 degrees from the case edge, so that they are 90 degrees from each other.

I recall reading somewhere that Hugh recommends you have a compass when you construct the GK-1.  I take it that this is to properly orientate the trafos to reduce hum?

If so, what is the process for orientation?

Thanks.

[jules]

Can I add a supplementary question to this ... ?

If the trafos are orientated at 90*, do they also need to be close together to take advantage of the [field cancelling?] effect?

and ... as with all these things, on a scale of 1 to 10, does this really matter very much  or might proximity of the trafos to critical components be of more importance?

Jules

[AKSA]

Hi Paul,

Ah, the old trafos facing each other trick!!

You can use any angle between 30 and 60 degrees, it's not critical.

You can have primary connections at the outside, secondaries to the inside, or the reverse.

Some commercial gear I've seen (with tubes also) puts the trafo on a circular plate with pinch screws so you can orient it for minimum hum.  That's pretty extreme, but it shows that other designers take the orientation of the trafos pretty seriously.

Jules,

Just saw your post.  Yes, to fully exploit this, the trafos would need to be close together, as they often are on the top cover of tube amplifiers, but not here.  It's actually wise to keep them far apart - I mount them at opposite ends of the Swift enclosure - and as far as possible from the tube circuitry on the main analog board.  The highest impedance point on the GK1 is the tube grid at 4.3 megohms, and this makes it highly susceptible to hum induction.

Cheers,

Hugh

[jules]

thanks Hugh ...

that's very useful in terms of design layout priorities

Jules

[Tabascosauce]

Thanks.

I was proposing to mount the trafos on either side of the case, with the pcb's in between.  I have a peice of galvanised steel that I will be cutting into two sheilds to place between both trafos and the PCB's.

I was considering mounting the trafos an equal distance from the analog pcb and the switching pcb.  I was concerned about the proximity of the trafos to the switching pcb, which has the signal running through the tracks and wire links.  Seems like you disagree Hugh, and the analog PCB is the bigger concern for potential hum.

[fajimr]

is thre any interference with the rca wiring?  or is this just an issue with the PCB?

if I put the trafos on either side one would be right next to the selector switch and RCA wiring....  any comments on this?

edit- I just saw this is similar to tabacosauce except that I am using a stand alone selector switch rather than the pcb..

[AKSA]

Hi Jim,

Shouldn't be - but good wiring practice is to keep all RCA and shielded cable far from AC lines and transformers.  As mentioned, the tube grid on the analog board and associated wiring to and from the volume control is quite sensitive to RFI, but this is normally at the front of the installation, where the power trafos are at the rear, so this is not an issue.  On the Swift, I shroud the transformers, the digital 5V supply and the digital control board under a large U shaped stainless steel shield, which runs the whole width of the preamp, with power and control wiring passing through holes in the shroud.  This makes it VERY quiet, as the shroud is bolted to the base and thus at chassis ground.  A true Faraday cage....

Cheers,

Hugh