EVGA W1 500W... Oops!

[Tator Tot]

Me, I see it as an "investment". If I pay a shit load of money up front to reduce failure rate from 5% to .01%, even on a cheap PSU like a VS, then I'll make my money back in a heart beat.

It's also an investment in consumer goodwill. Besides cutting your overhead on support and returns, you're also generating return customers who will buy your companies product again in 5-10 years when they're making a new system or recommend those products to others when they wish to build.

[ashiekh]

It is also possible that Aris got a bad unit and it is not a design flaw.

[Jon Gerow]

It is also possible that Aris got a bad unit and it is not a design flaw.

No. There are a lot of design flaws in that PSU.

[ashiekh]

Isn't the VS400,500,600 based on the same platform?

[Jon Gerow]

Isn't the VS400,500,600 based on the same platform?

Yes. But without the design flaws.

[Philipus II]

I agree. EVGA PSUs were never "popular" until the G2 came out and were sold below cost.

I don't think they were technically sold below cost, at least not most of the time. Gross margin was definately positive. Still, EVGA was more aiming for 5% gross margin, which other brands aim for 15-30%. The 15-30% you need to be profitable with your project because there is a lot of administrative overhead, cost for R&D, testing, QC, tooling, customer service, RMA, marketing, channel marketing/kickbacks/SI discounts etc.. Okay, if you don't have to spend money on R&D, tooling and testing you need much less margin, true, but you also rely on the goodwill and abilities of your OEM to deliver decent product. Worked fine for G2/P2, obviously didn't work for W1.

[Orion]

Isn't the VS400,500,600 based on the same platform?

You can see how the bridge rectifier in the Corsair does have a heatsink, right?

There are other improvements too, like moving the secondary capacitors a little to the left and cables to the right, so the caps are not hidden underneath the forest of wires like on the original design.

[Orion]

Now the supply is 500W so only 5A, so the temperature rise might be a lot
less than 50C if the board is itself a good heatsink and that would keep the
junction temperature below 100C even with a high ambient.

So I am not at all convinced that Aris is right.

Also, notice that it's not like the bridge rectifier (without heatsink) is rated at 3A@100C, and is rated for more current at lower temps.
No - it's literally rated at 3A at all temps from 0C to 100C. Of course, logic would suggest "surely if it is rated for 3A at 100C, it could do more amps below that", but the fact is the manufacturer rates it at 3A through the entire operating range. Even at 0 degrees C they don't guarantee any more than 3A.

[asf14tomcat]

don't forget the fan . this with blow some heat in diode bridge.
seem need someone make a test

[ashiekh]

Also, notice that it's not like the bridge rectifier (without heatsink) is rated at 3A@100C, and is rated for more current at lower temps.
No - it's literally rated at 3A at all temps from 0C to 100C. Of course, logic would suggest "surely if it is rated for 3A at 100C, it could do more amps below that", but the fact is the manufacturer rates it at 3A through the entire operating range. Even at 0 degrees C they don't guarantee any more than 3A.

As I keep saying, one can use the circuit board to dump heat through the legs (that is also in the specs)