What's wrong here

[lynx]

Following clocker's recent puzzle, I got another for some of you to try to figure out.

I'm building a new system for my brother, he only wants it for internet/business so no fancy graphics required.

So I got him a Biostar 6100/M9 motherboard, Athlon64 3200+, 1GB Corsair VS cas2.5 memory, 80GB Maxtor Sata drive, NEC 3550 DVD writer, Cheap aluminium case (but seems well designed), Akasa Paxpower 400W PSU (Which I believe is a rebadged Seasonic S12-430).

I initially put the processor, memory, motherboard and PSU together to bench test these components. All the fans powered up but it wouldn't post. I took my own system apart and substituted the cpu and memory, both were fine. Next I tried the PSU in my system, that too was fine, so the only thing left was the motherboard. I quickly checked that the DVD writer and HDD were ok, and headed back to my supplier with the motherboard.

Of course it is Saturday morning, and they've only got a skeleton staff working in technical support (most of their business if Mail Order, Mon to Fri), so their stated policy is book it in and leave it for testing. However, I know from experience that if it is easy to diagnose they will often test it while you wait. I'm lucky, they aren't too busy. They test it and it works perfectly.

I take it back home, plug it together again, it doesn't work. However, I've seen it working (with their cpu and memory). So I put the cpu and memory from my own system in, it still doesn't work. I remembered at this point that they had been using a PSU with a 20 pin plug, so I get out the cheap and nasty PSU that came with the case, plug it in, and straight away the system posts.

The puzzle for you to work out is why it posts with the cheap PSU, but not an expensive one. I warn you, I hadn't actually done anything wrong, so it's going to be a little harder to work out than clocker's problem.

[Darth Sushi]

Because the dylithium crystals were not properly aligned with the flux-capacitors and the matter/anti-matter chamber coils were overloading the warp nacelles!

BTW, did you plug in the aux 4-pin connector? ... Did you use a 24-pin connector?

[lynx]

Because the dylithium crystals were not properly aligned with the flux-capacitors and the matter/anti-matter chamber coils were overloading the warp nacelles!

BTW, did you plug in the aux 4-pin connector? ... Did you use a 24-pin connector?

Yes, I remembered to plug in the 4-pin connector. The system wouldn't post using the psu with the 24-pin connector, but it did post with a cheap'n'nasty psu with a 20-pin connector.

[Darth Sushi]

Yes, I remembered to plug in the 4-pin connector. The system wouldn't post using the psu with the 24-pin connector, but it did post with a cheap'n'nasty psu with a 20-pin connector.

Thank God for cheap 24 to 20-pin adapters.

[clocker]

Not enough power on the 5v standby?

Edit: Do you already know the answer or are we all shooting in the dark?

[lynx]

Not enough power on the 5v standby?

Edit: Do you already know the answer or are we all shooting in the dark?

Nope. Afaik nothing uses standby rail power on the motherboard itself. Your thinking is in the right area, but it is not +5v standby.

Edit: Hmmm, that's interesting, you may not see that.

[fkdup74]

these friggin puzzles are getting old

btw, I vote for the dylithium crystal/flux capacitor fix

or....

Thank God for cheap 24 to 20-pin adapters.

or....

I warn you, I hadn't actually done anything wrong

[zaphodiv]

I like these threads.

More information on exactly what it did at switch on might help.
Did you get a beep code or see anything on the monitor?

The manual dosn't mention a jumper setting for 20pin versus 24pin.

Using a 20pin psu will mean pins 11,12, 23,24 on the socket not being connected which are additional +12v, +3.3v, +5v and ground connectors.
Since this board is intened to have those supplys that dosn't make much sense. At a total guess its related to the jumper for keyboard power.

[spirex]

The only thing that I can find is that the motherboard states that pin 20 needs -5v and the power supply doesn't supply -5v.

[lynx]

It didn't post - in other words no beeps, nothing on the monitor, no vga sync, nothing.

There are no jumpers to adjust.

As you say, the extra 4 pins supply extra 12v, 5v, 3.3v and ground. Since there was nothing plugged into the board there was nothing needing for that extra power supply.

I repeat, I had done nothing wrong. For an extra clue, I suggest you look at my last post very carefully.

Perhaps I should also say that the system is now up and running - with a 24-pin psu!