[jtostenr]

Inspired by Johnwmclean's PSU layout, I decided to try my hand at Illustrator and figure out how I'm going to squeeze 4 S22s, 4 trafos, 1 E24, 1 S25, 2 S24s and a relay into a 12x16x3 case....

No wiring yet, but here's the layout I have so far. I'm trying to avoid running AC all over the place, but I have not figured out an optimal layout yet.

Any advice on layout would be appreciated

Jeff

 

[johnwmclean]

Well done! I had a play, my thought was to keep the umbilicals on either side to save a messy crossover, gives the dc a direct path.

 

[jtostenr]

Well, I was hoping to preserve the heatsinks if at all possible. Not just for thermal reasons, but I don't have any way to cut them cleanly. If I can't figure something out, I may just have to bite the bullet though.

I was trying to keep all the AC on one side (and the inputs on the B22 on the other side) with the original design in order to minimize hum, but do you think it's necessary to go to that extreme? The two cases will be stacked.

 

[oneplustwo]

How about stacking the sigma boards 2x2? Not sure how tall the heat sinks are but that could significantly save some space for you.

 

[jtostenr]

I thought about that, and it may be possible, however I don't like the idea of the unequal heat distribution between the upper and lower mosfets. Good idea though.

 

[johnwmclean]

Quote:

Originally Posted by jtostenr /img/forum/go_quote.gif I was trying to keep all the AC on one side (and the inputs on the B22 on the other side) with the original design in order to minimize hum, but do you think it's necessary to go to that extreme? The two cases will be stacked.

Keep the trannies as far away from the amp boards as possible

 

[oneplustwo]

Quote:

Originally Posted by jtostenr /img/forum/go_quote.gif I thought about that, and it may be possible, however I don't like the idea of the unequal heat distribution between the upper and lower mosfets. Good idea though.

Would the heat distribution really be that different? Just imagining how the heat will "flux" I'm thinking you should be able to get the distribution pretty even. Hard to say i guess without knowing how tall your sinks are. Another idea, although harder to implement, would be to offset the top board from the bottom board so the MOSFETS have horizontal spacing as well as vertical. Not sure how that would work... maybe a piece of 1/4" aluminum about 1/2 a sigma board long that you could screw standoffs to? Also, another variable... remember you could mount some FETs above the board and some under. Just brainstorming.

 

[luvdunhill]

yes, it will be uneven.

There is pretty much no reason for four sigma22 that I can think of. Look at what those MOSFETs are rated for!

 

[jtostenr]

oneplustwo - My heatsinks are about 2.5" tall and the case is 3". If I could manage to stack the S22s somehow, I think it would make things dificult when it comes to troubleshooting. I think I'm actually leaning towards just cutting the heatsinks down. Since I am using four boards, I don't think they will ever really get pushed that hard and not generate much heat.

luvdunhill - I'd rather not get into a debate about 1 vs 2 vs 3 vs 4 S22s (or balanced/unblanced for that matter) at this point....do I really "need" 4 S22s? Probably not, but I know that by providing each B22 with it's own S22, I am presenting the smallest possible load to each S22, which I think has some benefits.

 

[luvdunhill]

Quote:

Originally Posted by jtostenr /img/forum/go_quote.gif I am presenting the smallest possible load to each S22, which I think has some benefits.

That's fair. You might find that it's difficult to match the rail voltages across all four PSU boards without a lot of frustration. Whether or not this is more important than providing the smallest load on each S22 is for you to decide. You might also find that putting more load on power supply boards might also sound better. Again, you'll have to experiment, but many times things are more stable with more current passing through them. I've done my testing and have come to my conclusions