Dell GX620

[PCBONEZ]

Re: Dell GX620

Oops, my mistake.
The bloated caps on both those are 85c 100uF 25v 6.3mm.
100% on the D850GB and ~50% on the Gateway.
Gateway also has 2 or 3 85c 5mm [no vents] that -might- be bloated.

I'll try to get a snap-shot but bloated 6.3mm don't show up well.
- They don't have -room- to bloat that much when they do bloat so the 'bump' is slight.

[PCBONEZ]

Re: Dell GX620

Dunno if you can see it but there it is.
They are noticeably plumped up [bulged] between the lines of the vent stamp.
Obvious / easy to see in person. - Hard to show in a photo.

The one showing the RAM slots is the D850GB.
The Kittyhawk is the other one.

[PCTechNY23]

Re: Dell GX620

I saw the post and asked the stupid question because I figured midibob may have had a unique situation or possibly just a fluke. These board are almost identical to the SX280s that I have been doing a lot of since our Dell warranty ran out ( my recaps last longer than their child labor ones anyway ) and I never had a problem with the 16v 220uFs 85c on those. That coupled with the fact that I looked at every board replaced by Dell techs and I didn't see any blown 16v 220uFs I am somewhat surprised. Now that our warranty is up on most of these and our asset manager will finally let me take a crack at them I will have to test some of the 16v 220uFs and let you know what I find.

[PCBONEZ]

Re: Dell GX620

The only thing unique about midibob's post is that he's one of the few that understands the whole problem.
Most techs are so 'conditioned' as to what to replace they don't look around for odd problems.
I suspect a lot of repairable GX620 boards are getting written off and trashed because the techs don't think to check the 220uF [or even 100uF].
-
Absolutely not a fluke.
Not uncommon for 85c caps to dry out or bloat on a mobo with poor airflow.
Thing is they aren't heavily loaded with ripple like the 'popular' replacement caps so it [usually] takes them longer to dry or bloat and cause problems.
-
So I understand the GX620 [I don't have one to play with yet but I'm looking] puts a bunch of 220uF right under the hard drive which limits airflow and is source of heat on the 220uF that SX270/SX280 don't have.

Personally if it's USFF or a server board [or for ME] and it has 85c caps I replace them.
Buying in bulk [100pc+] the cost difference is usually < $3 per board [for Dell GX/SX anyway] and that's cheap insurance.
- It's also handy to have a handful of 100uF and 220uF around to replace 85c caps in LCD screens when they show up so I just keep at least ~50pc of those sizes in stock now.
.

[PCTechNY23]

Re: Dell GX620

The SX280 DO have at least 8X 16v 220uFs under the hard drive, the ones I have seen are Rubycon YK

The GX620 USFF has 11 16v 220uFs under the hard drive and those have been chemi-con SMG series on the machines I have seen so far.

Look at the picture:
http://www.flickr.com/photos/3082593...n/photostream/

Full list of GX620 USFF caps from the cap king:
7X 6.3v 2200uF
13X 6.3v 1800uF
5X 6.3v 820uF
4X 16v 1000uF
3X 16v 1500uF
19X 16v 220uF
4X 25v 22uF

[PCTechNY23]

Re: Dell GX620

Still trying to wrap my head around this one

"So I understand the GX620 [I don't have one to play with yet but I'm looking] puts a bunch of 220uF right under the hard drive which limits airflow and is source of heat on the 220uF that SX270/SX280 don't have."

Are you saying the SX270/SX280 do not have hard drives?

[Topcat]

Re: Dell GX620

So I understand the GX620 [I don't have one to play with yet but I'm looking] puts a bunch of 220uF right under the hard drive which limits airflow and is source of heat on the 220uF that SX270/SX280 don't have.

The SX280 and GX620 are almost IDENTICAL systems, only a few variations in their boards, cases are identical. The 220uf caps don't fail on them, its the little 22uF 25v caps that will go from time to time. The SX270, I've never seen anything lower than the 680uF caps fail on.

FWIW on the sx280 and gx620, I've stuck 220uF 16v caps in place of the failed 22uF 25v caps many times, never had any problems doing that.

[PCTechNY23]

Re: Dell GX620

Here is a not very good side by side ( GX620XSX280 ) but it would make a great nerd edition photohunt screen. These boards will both fit on the same tray and into each others case. Aside from a chip switch, led upgrade, Chipset heat sync upgrade ( those retaining clips are prone to pop off the board when they get hot), and a couple cap substitutions ( the 2200 next to the battery to a 16v 220uF, 1 extra 16v 1000uF filling a blank, 3 extra 6.3v 1800uFs filling blanks on the left of the heat sync, and leaves a blank next to the printer port), the SX280 is identical to the GX620as far as I can see. I am sure I missed something and it will be pointed out but as far as trends I have noticed on the SX280 I expect them to transfer to the GX620. The optiplex 745 does have a fan under the hard drive and I will let you know next week if the fans can be fitted to the GX620 and SX280. I borrowed the pics from thecapking.com, hopefully he won't mind.

[PCTechNY23]

Re: Dell GX620

This would make more sense if I included the link: http://www.flickr.com/photos/43646305@N04/4018780162/

[Topcat]

Re: Dell GX620

^
You must be the capking guy... haha

I have some 620's and some 280's in the shop, I'll try to get a better comparative photograph.

[PCTechNY23]

Re: Dell GX620

No, but I have purchased caps from him and bounced a lot of questions off him. I couldn't find any other pics online. I suppose I could take a couple pics myself but no need to take a picture that has been taken.

[PCTechNY23]

Re: Dell GX620

Do you have any 745s in? I would love to see what those fans look like. I am sure Dell will charge $50 for them; maybe we can get them to admit their mistake and give them out to SX-280 and GX-620USFF owners' lol. Or we can figure out what fan and source a bracket maker. After the GX270, GX280, SX270, SX280, A couple precision duel Zeon machines and for the last straw GX-520 failures like most govt and non profits the company I work for has switched to HP. Dell has taken the greatest nosedive I have ever seen. I am one capacitor away from officially declaring Acer more reliable than Dell. The only worse piece of hardware I have seen lately is the Xbox. Those things look like they were designed and put together by a secret groups of Encino men in a secret plot assume the identity of a OEMs and manipulate their way to the bottom of the lot only to blame Dan Marino for the entire thing.

[Topcat]

Re: Dell GX620

I haven't seen a 745 in the shop yet......I sense another train wreck, knowing dell though. The SX/GX270 and 280 are low to mid-grade systems, so I can see them cutting a few corners here and there to keep the prices down. However, the precision line just bewilders me. These are systems are 'high-end workstations' according to them, that start out at 3000 bucks and rapidly went up when new (depending on what bells and whistles you buy with it). It boggles the imagination why they cheap out on those... Getting them to admit they screwed up? Good luck!! Even when they lost a class action lawsuit and were forced to extend warranties on GX270 and GX280 models, they still never admitted they did anything wrong...

[PCBONEZ]

Re: Dell GX620

Interesting, looks like SX280 and GX620 are built on the same reference board but GX620 uses the newer chipset.

I get Dell boards now and then but not the whole chassis and until recently I've been trying to avoid them simply because I don't like dealing with proprietary layouts.
Curiosity finally killed the cat and I've thought up a little experiment I wanna do on some of the models that I need a complete chassis for so I've been shoppin' for some complete units [cheap - LOL].

I've been saying "220uF 16v" when what I mean is the small uF caps in general.
When I rebuild a board that I'll have to support for any length of time, if those caps are 85c [or from certain crap brands] they go bye bye.
Replacing ALL of those caps on a board costs a lot less in the long run than wasting several hours [or days] chasing screwy problems.
They don't go bad often but when they do the problems they create can cost you a ton of time going in circles before you get it tracked down to the caps.
- Hard drives that keep getting corrupted on a particular board but work fine on others is a good example. [Same-same with RAM that tests good in other boards but fails in a suspect board.]
- Here's one experience of that exact problem with HDD controllers:
Just before I heard of the whole bad caps 'thing', I built a system for my niece using a brand new Soyo SY-P4VGA. At about a year old it started corrupting the hard drive and if I put -any- drive [HDD or optical] on the secondary controller it wouldn't detect ANY drives present on either controller. I went around and with it wiping and changing out drives and so forth trying to 'get it right' but it kept losing data a week or two after each 'fix'. Since she was off to college soon and I was due to move out of State, her folks had me refurbish a laptop for her instead. [Which was the plan anyway, we just switched her to a laptop 6 months sooner than planned.]
The Soyo based system got shoved in a corner in their garage where it sat for a couple years. I eventually got it back [my wife wanted the 'girly' case for her new system] and [now knowing about caps issues] what I found wrong was two bad 100uF 16v between the chipset and the IDE ports.
They did say 105c on them but they were Sacon [formerly GSC] so they are probably more like 50c.
Recapped the whole board then everything worked fine again.

Slightly OT.
Soyo boards are usually great boards - IF you replace EVERY flippin' cap on them.
Short of that they tend to degrade fast after 6-18 months use.

~~

About Dell 'cheaping out' on their boards.
It's not Dell, it's Intel, the mobo manufacturer [often built by Foxconn for Intel on some 'Partnership' contract.]
I'm saying that because Intel does the same thing on boards that aren't specifically built for Dell.

- I got this from reading some Intel marketing crap.
It applies to PC's and Workstations, not outright Servers.
[I'm not saying I agree. - I'm just relating how Intel's 'Suits' think.]

- Intel has this idea that the 'maximum useful lifetime' [as in 'practical useful lifetime'] of a system is 7 years assuming a 40 hour work week and that the system is off when not in use. That works out to 14560 hours of 'on time' and 1820 power up/down cycles for the system.
- Based on that they analyze and test components [including mobo, CPU, RAM, drives, ICs in general, caps, thermal paste, everything] and build systems/mobos/CPU's/Chipsets to last -a minimum of- 14560 hours of 'on time' and 1820 power cycles. They don't determine the maximum time it will last, they design based on things lasting -at least- that long.
Any lifetime past that is shear luck.

Mentioned because: Those 85c caps probably WILL last for 14560 hours of 'on time' - but people [or offices] that never power down their systems eat those 14560 hours up in ~607 days. [~20 months.]
.

[PCBONEZ]

Re: Dell GX620

Still trying to wrap my head around this one

"So I understand the GX620 [I don't have one to play with yet but I'm looking] puts a bunch of 220uF right under the hard drive which limits airflow and is source of heat on the 220uF that SX270/SX280 don't have."

Are you saying the SX270/SX280 do not have hard drives?

SX270 uses a laptop hard drive that mounts under the motherboard tray.
I was in error about the SX280.
Confused myself because the GX270 and GX280 SFF use the same layout as each other and that's what I was thinking about.
Changing the USFF to GX____ from SX____ doesn't help me keep things straight either.
.

[midibob]

Re: Dell GX620

Hi all,

Further to the discussion on the 220uF 16v caps. My original point was that most people (including myself) were ignoring these caps, as visually they looked fine and you could possibly make the wrong assumption that as they didn't really need to filter out much, because of their size, they wouldn't cause too much of a problem.
It was only when I had a board in with no blown caps that wouldn't boot that I decided to test every cap on the board. It soon became obvious that the ones nearest the heat source (lack of ventilation) were all dried out to various degrees. The bottom line is, whatever they do it's pretty important!

I've seen a number of re-worked boards (Dell replacements) and they have all been changed to 105c. I've also seen one that had polys fitted for some of the larger caps.

Out of interest I always replace all the 220's around the area under the hard disk including the two by the i/o ports (11 in all). Not had any returns for about a year now so fingers crossed.

Happy recapping
Midibob

[PCBONEZ]

Re: Dell GX620

Hi Midibob.
Thanks. That's what I was trying to get across.

Would you perhaps have a cap-map of the 'usual suspect' 220's?
I don't usually see boards in the chassis so I dunno for sure where the HDD is relation to the mobo's layout.

Thanks.
.

[midibob]

Re: Dell GX620

Here we go.

The 2200uF I've only seen this one fail (venting) although there's no doubt all may be affected in that area at some time.

Bob

Forgot to add, the hard disk sits over the bottom left 3/4 of the board as you look at it.

[PCBONEZ]

Re: Dell GX620

Thank you Sir!
.

[techcaps45]

Re: Dell GX620

80% of the "inside" 11X 16v 220uFs on mine test bad as well. The "outside" 8X and the 2X 25v 22uFs all test perfect with my trusty capanalyzer 88.