[Felipe]

Hello all,

I need an advice on how to perform this, or rather if my thinking has any counteradvice.
I am fitting my LF55 in a 1U case so its very tidy space. Max is 40mm high.
I managed to cut the heatsink so i end up with a 40mm high, i will make sure its properly ventilated. What i am having trouble is with the ps unit.


I already have it cutted in half ( i had monoblocks) so this is a picture of 1 channel. What i was thinking was to cut this through the RED line. I would then connect the (-) and (+) with a wire.
My question:
Is that GROUND EARTH around the board necessary for any RFI rejection purpose ? Can the split mess up with that ?
I hope not....

[AKSA]

Hi Felipe,

Absolutely no problem, you will not have more hum, just keep the wires joining the cut boards nice and thick!

Cheers,

Hugh

[Felipe]

Thanks a lot Hugh.
I will assume a 2mm (square) will be suficient.
I will post results and pics when its finished.

[andyr]

Thanks a lot Hugh.
I will assume a 2mm (square) will be suficient.
I will post results and pics when its finished.

"Nothing succeeds like excess", Felipe!    So I personally would use 6.5 mm sq mains cable if I was doing what you intend ... or 30amp car wiring.

2mm sq wire is only 14awg.

Regards,

Andy

[Felipe]

Andy,

I understand that the copper junctions i am splitting are no more than the quantity present in a 14awg cable. After all its a plate that is 0.2mm thick and 3mm wide. I belive a good solder joint with lots of contact between copper plates and wire is more important...

Also this will not me more than 10cm length of wire.

[andyr]

Andy,

I understand that the copper junctions i am splitting are no more than the quantity present in a 14awg cable. After all its a plate that is 0.2mm thick and 3mm wide. I believe a good solder joint with lots of contact between copper plates and wire is more important...

Also this will not me more than 10cm length of wire.

That's very true, Felipe!    (Perhaps Hugh needs to increase the Cu area in his PS PCB tracks!  )

Regards,

Andy

[AKSA]

Hi Andy, Felipe,

1mm wide 1oz copper tracks can carry 1A continuous.  2mm is 2A, and double these ratings for the 2oz tracks I use.  Have a look at the tracks on the USB port of a PC motherboard.  These can carry up to 2A!!

The charge currents on the power supplies are pretty high, probably 50A peaks for a few microseconds each mains waveform, but any resistance in the tracks will have the effect of lengthening the duty cycle and warming the tracks.  I've carefully tested for temp increases, and found that they are not detectable above ambient at normal operation levels.

The design is well tested in the field, and seems adequate as none have caught fire or failed due to overcurrent or burned tracks.

I'm very careful about current loading on pcb tracks.  That's one of the reasons I always use 1.5mm tracks wherever possible, and more if I can manage it.  My pcb designs take weeks to complete, they drive me crazy.   

Cheers,

Hugh

[andyr]

Hi Andy, Felipe,

1mm wide 1oz copper tracks can carry 1A continuous.  2mm is 2A, and double these ratings for the 2oz tracks I use.  Have a look at the tracks on the USB port of a PC motherboard.  These can carry up to 2A!!

The charge currents on the power supplies are pretty high, probably 50A peaks for a few microseconds each mains waveform, but any resistance in the tracks will have the effect of lengthening the duty cycle and warming the tracks.  I've carefully tested for temp increases, and found that they are not detectable above ambient at normal operation levels.

The design is well tested in the field, and seems adequate as none have caught fire or failed due to overcurrent or burned tracks.

Cheers,

Hugh

Hi Hugh,

I was simply thinking about making the PS better - and therefore the LF - better!!     After all, if we think about the lifecycle of the amps and the GK-1:
*  after you designed the AKSA circuit, various "tweakers" decided to play around with certain components ... and this led on to the '+' and 'Nirvana' versions.
*  then you came out with the LF circuit ... and some people have "improved?" this with teflon output caps and other tweaks.
*  same with the GK-1 ... with all the cap changes described in one of the threads below.

Logically, it seems to me that if a LF PS "PCB" was re-created using (say, 1 mm sq) point-to-point wiring, it would have less resistance than your PCB tracks.  This should make the PS faster ... so the power amp should sound better?    But I'm not going to pull apart 6 LF PSes in my tri-amp setup, to experiment! 

(It's not about whether the copper trace will or will not get hot with the current passing through it.)

Regards,

Andy

[AKSA]

Andy,

I honestly don't believe thicker/wider tracks would make it better, or change the sonics......

However, I was wrong once before, in 1962, in fact......   

I will see you tomorrow at your esteemed 60th celebrations!!  (or should that be cerebrations?)

Hugh

[andyr]

Andy,

I honestly don't believe thicker/wider tracks would make it better, or change the sonics......

However, I was wrong once before, in 1962, in fact......   

I will see you tomorrow at your esteemed 60th celebrations!!  (or should that be cerebrations?)

Hugh

Aw gee, Hugh, don't spread it all over the Internet!    Now, Audiocircle denizens will dismiss my posts as senile ramblings!    Hang on ... maybe they already do that! 

OK, is there anybody out there willing to make up a PS using p-to-p wiring (with solid-core 1mm sq wire) instead of Hugh's PCB and test out whether it makes their amp sound better?

And with respect, Hugh ... IMO you have been proven wrong with the GK-1 ... which was in this century!     When you developed it, you didn't think people would be able to improve it - but they did, with input & output SoniCaps - and then with those other caps that West organised a group buy for. 

See ya tomorrow.  It should be a notable turn.   

Regards,

Andy

[RonR]

Gents,

In true Blue Peter style, here's one I prepared earlier:



This is on 3oz board, with 6mm minimum width tracks - that works out to 18A capacity per track.

It made a definite improvement to my LF55, but then it's close to Hugh's newer PSU design, so I'd imagine most of the improvement is down to the design. The Amp doesn't want for grunt when the going gets tough though!

Happy Birthday Andy!

Cheers,

Ron.

[AKSA]

Ron,

That is exquisite work, congratulations, you are an artisan!

It was a great party, and as one expects, Andy's wife Vicky turned his hifi down at least once that I noticed...  that made me laugh! 

Cheers,

Hugh

[andyr]

Gents,

In true Blue Peter style, here's one I prepared earlier:



This is on 3oz board, with 6mm minimum width tracks - that works out to 18A capacity per track.

It made a definite improvement to my LF55, but then it's close to Hugh's newer PSU design, so I'd imagine most of the improvement is down to the design. The Amp doesn't want for grunt when the going gets tough though!

Happy Birthday Andy!

Cheers,

Ron.

Thanks, Ron.  Sensational PS you have there!    That's exactly the sort of "heft" that I was thinking would make a difference to the sound of Hugh's amps!    A shame it wasn't a direct comparison with the old one, so that just the track cross-sectional area was different ... and you coukld've proved my point! 

Regards,

Andy