G5 1.8dp pci-x

It's probably dead. Not kidding.

That was the model I owned; absolute heat pig.

I'd start with the PSU, you can also check the caps on the board. Most likely is that the U3 memory controller on the back broke from its solder, with the PSU being next.

I thought these only had the ram slot problems after the board had been over flexed. This still has the stock amount of ram512mb.

They all had the memory controller issue, excepting the final multicore machines which have the far superior U4 memory controller. Let me put it this way; your model here had a 40% failure rate in the first year.

Well thanks for the information. Figured I will still pull out the multimeter tomorrow and check the power supply but if that's not the problem then I guess see what I can throw in the case.

Bummer they are nice machines. I have the original 1.6 single CPU machine and it works great, but slower then my other DC 2.0GHZ PCIE models. So many decent motherboards end up dead because of the switch from lead to silver solder and bad caps. Capacitors can be easily replaced but I don't have the skills of equipment to do reballing (could save a few video cards I have as well if I did).

Sad to hear the demise of such a machine.

Q: Cant the U3 memory controller desoldered, the pads cleaned up and fixed and the chip soldered back on?

Yep it's almost guaranteed that the PSU is cactuar, and this will be the primary issue that prevents it powering on... However it is likely that it has logic board issues or will have them eventually. Unfortunately, the PCI-x G5's were complete and utter garbage. I have a 2.0DP that makes a good shelf for this reason... It's unstable and unreliable as hell, and sometimes dumps a CPU when it feels like it, rendering it only a single-processor machine. In the end I opted not to fix it and simply replaced it with a PCI-e 2.0 dual-core instead given how far the prices of 2nd had G5's have dropped. I can't say a bad word about this machine... it's fantastic, and appreciably quicker than the one it replaced.

Not sure about the U3 controller... does it use normal feet or is it a BGA arrangement? If it is BGA then I beleive it has to be essentially flowed onto the board with localised heat. If its normal pins, well one may be in luck, however I beleive part of the issue was simply that the memory controller itself was complete and utter rubbish... just like the IDE controllers on early B+W G3's.

Pretty sure it's BGA.

I like that G5/Mac Pro case design.

I like the case too, which is why I bothered getting something that I can't boot OS 9 directly from.

Well I just checked for 5v standby and get nothing so I was thinking about doing this atx hack http://damntechnology.blogspot.com/2009/02/rebuilding-powermac-g5-power-supply.html?m=1

Anyone ever try it here?

I reckon that ATX hack is worthy of it's own thread and a sticky if ever anything was!