Dell OptiPlex GX270 - multiple bad cap failures

[Uranium-235]

Re: Dell OptiPlex GX270 - multiple bad cap failures

well I got them all reversed, and it almost comes on, but never fully goes on

I tried, the mofsets might be burned out. oh well plan b

[Uranium-235]

Re: Dell OptiPlex GX270 - multiple bad cap failures

well, plan b was about to rock, and I f*cked it up

I got a dimension 4600c from work, it boots up to say "northbridge heatsink not found, halting"

one of the metal loops that holds the nb heatsink on broke, I'd imagine just soldering a wire on and using it to hold the nb heatsink hook would work

I was also going to replace some KGZ's on the VRM (rest of them Sanyos, which are OK). I had to take the P4 mount off, and it was being sooo stubbron, I actually cut the plastic pushers off and force it off, and mount the P4 holder from the gx270 (which uses screws). I folded a static bag to protect the board when I pryd the mount off, but I didn't fold it enough. it broke 3 traces between the cpu and nb, and the thing won't power on now. I tried to scrape some insulation off some of those traces to get a wire bypassing the broken traces but it just isn't working

damnit

[Uranium-235]

Re: Dell OptiPlex GX270 - multiple bad cap failures

well I just replaced all those polys with lytics and it does the same thing, never fully comes on then dies. what do you think it is? mosfets?

[tuskenrader]

Re: Dell OptiPlex GX270 - multiple bad cap failures

I know it's been half a year since this thread seemed to die, but just thought I'd throw out a question here. There's one 680uF 10v cap which BadCaps says can be replaced with an 820uF 6.3v on its GX270 kit page. I just wanted to confirm that here on the forum. Going down in voltage makes me wary.

[PCBONEZ]

Re: Dell OptiPlex GX270 - multiple bad cap failures

I use a 680uF, reason being those boards have 820's all over them so if an 820 was a good idea they'd just use one in the first place to reduce their costs.
And it's the same situation with that one 680uF in the same place on many Dell boards.
Obviously the uF matters on that one cap.

I tried to trace the circuit at one point.
It has something to do with the chipset but I couldn't ID the specific pin.
.

[brethin]

Re: Dell OptiPlex GX270 - multiple bad cap failures

I dont think he was questioning the uF change from 680uF to 820uF but was questioning the voltage change from 10v to 6.3v. Not personaly knowing the circuit myself I would say if Topcat has it listed as a replacement it is a safe replacememnt. I am sure he has done his homework and didnt just use the new value just to change it.

[PCBONEZ]

Re: Dell OptiPlex GX270 - multiple bad cap failures

I dont think he was questioning the uF change from 680uF to 820uF but was questioning the voltage change from 10v to 6.3v.

The voltage change is fine.
I don't have my notes handy but I do remember it's on less than 5v.

Not personaly knowing the circuit myself I would say if Topcat has it listed as a replacement it is a safe replacememnt. I am sure he has done his homework and didnt just use the new value just to change it.

Doubtful. - The methodology has been stated before.
The odd way Intel sticks solidly to 680uF on multiple Dell boards for that one cap when 820uF in the same diameter are stocked in droves indicates that particular circuit it is frequency sensitive and TC's method doesn't look for that. It's associated with the chipset so it might only have to do with sound but OTOH it might have to do with syncing the clocks between the CPU and RAM. To check for problems with the latter you'd have to actually count RAM errors at the chipset [vice at the RAM] which you can't do with normal equipment or software.
That situation is pretty rare. It's not something people think of.
.
Now, if I only saw the peculiar use of a 680uF in place of readily available 820uF's they buy in mass quantities on one model board I wouldn't think too much of it but Intel does it on multiple models with similar chipsets through the socket 478 and 775 models so for some reason the value for that particular cap is important.
And I think Intel has probably "done it's homework" too and they didn't just randomly use an odd value and then stick to it with such enthusiasm on dozens of models if the value didn't matter.
.

[tuskenrader]

Re: Dell OptiPlex GX270 - multiple bad cap failures

Good to hear that the voltage won't be a problem. If it's less than 5v, then I feel comfortable with a 6.3v cap. As far as the uF goes, moving to an 820uF is about +20.59% over the 680uF value. If the +-20% rule can be trusted in most cases I guess I can buy TC's decision to stick a 820uF there. I'll give it a shot on one I've already replaced and see if I run into troubles. But I think I will be buying a small baggie of 680uF's, though. I was hoping my big bag of 820uF's would suffice since I already have them, but no biggie. I really appreciate everyone's input!

[tuskenrader]

Re: Dell OptiPlex GX270 - multiple bad cap failures

Just got a response from the man in charge. The kit page needed updating and he now is always including the 680uF in the GX270 kit. He also said that you can always drop a 10v cap down to a 6.3v since they always run at 5v on any board. Very interesting. I asked him if he thought 820uF would work.

[PCBONEZ]

Re: Dell OptiPlex GX270 - multiple bad cap failures

The kit page needed updating and he now is always including the 680uF in the GX270 kit.

Probably happened after an old thread where I pointed out how weird it was for Intel to stick to using a 680uF there.

I really suspect 680uF is already pushing the upper design limits for that cap.
[As in Intel pushed it as far as they could.]
That cap is blown on some where between 1/4 and 1/2 the [maybe 1/3] of the Dell boards I see, even when few other caps are blown. That suggests to me that it's heavily loaded. The fact they didn't use an 820uF there for the bigger can [to up the ripple rating] tells me they can't in that spot for some reason.
.

[Gabriel]

Re: Dell OptiPlex GX270 - multiple bad cap failures

Is it possible to use a gx280 psu on a gx270 mb ?

I guess the 280 psu has more wattage and they look alike but are they interchangeable ?

[severach]

Re: Dell OptiPlex GX270 - multiple bad cap failures

GX280 tower PSU: 24 pin
GX270 tower PSU: 20 pin

The desktop and slim desktop PSU are both 20 pin with SATA power for GX280 and Molex power for GX270. It works if you can find a way to power the hard drive.

20 and 24 pin supplies are compatible if there is no physical obstruction to plugging it in.