Tight squeeze. Need help with layout.

[maurycy]

I am trying to put my LM3875 chipamp in a new case. Originally, I had only one trafo and one power board but since I got a matching trafo, I am going with what I think it is called dual monoblock setup (though in one enclosure). Unfortunately, adding additional stuff to a small enclosure is a challenge. It is going to be a very tight squeeze. So here are few layouts that I have come up with. Any one better than the other ones? Front of the amp is on the bottom.

Layout 1
http://i1215.photobucket.com/albums/...sc/layout1.jpg

Layout 2
http://i1215.photobucket.com/albums/...sc/layout2.jpg

Layout 3
http://i1215.photobucket.com/albums/...sc/layout3.jpg

Layout 4
http://i1215.photobucket.com/albums/...sc/layout4.jpg

Please do not recommend getting second enclosure for PS because I bought this one over a year or so ago and can't find a matching one now. I also do not have the money to buy two new enclosures.

[firechief]

Heatsink really needs to be exposed to the outside air. If you enclose it inside it will not do its job of keeping the chip cool. Solve the space problem by making an opening in the box so that the heatsink is outside. Kill two birds with one stone.

[maurycy]

I was thinking about it but the amp needs to pass the WAF (wife acceptance factor). It is not going to be the case with aluminum fins sticking on the sides. Also, there is a lot of holes on the top and I can drill some on the bottom to let the air in and out. Airflow shouldn't be a problem.

[discrete]

Originally Posted by maurycy

Airflow shouldn't be a problem.

...not so sure about that.

That style of heat sink needs a fan.
The fin spacing is too dense to allow natural convection.
They will not work well, even if installed outside the case.

If the case walls are more than 2-3 mm thick , it may be better to just bolt the chips to the case, lets say one to each side wall.
Space problem solved as well....

[maurycy]

Thanks for the suggestions. I may go with a different heatsinks. As to the mounting them outside, that's most likely not going to happen. I had this amp in similar enclosure with smaller heatsinks and it never overheated.

So given that I will use better heatsinks and make sure that the airflow is not restricted inside the case, which layout would work better?

[AndrewT]

I can't see any layouts.
It appears you have not attached them to your post.

[maurycy]

Andrew, there are links in my first post at the top.

[borninthefifties]

Originally Posted by maurycy

Andrew, there are links in my first post at the top.

I would use option 3, gives a bit of separation from the toroids.
I have a similar setup using the same boards, and similar sized heatsinks and they barely get warm on an 8 ohm load!

Someone will probably disagree..................... but it works for me!

[kvholio]

I would go with option #1, but with both amplifier-pcb's/heatsinks
rotated 180 degrees so the heatsinks will be separating the transformers+rectifiers from the amplifier-pcb's.
Braid/twist all ps-wiring to minimise radiation and keep all ps-wiring well separated from the signal-wiring.
But before comitting yourself to a layout by drilling holes in the enclosure
to fasten various components, you could try various options
without fastening the components.

Regards,

Klaas

[discrete]

I would suggest one of the options attached, using the chassis as a heat sink.
(Even an 8"x4" aluminum plate/bar will work better than that PC heat sink).

The worst 100/120 Hz radiators are shown in blue.
Apart from the other benefits of adding a real filter bank between the rectifiers and the amps, these radiators are shortened in the second option shown.
(But I do not want to start an argument about the 'benefit' of low capacitance here...)