Badcaps.net Forum
Go Back   Badcaps Forums > Motherboards with Bad Capacitors > Soyo - Shuttle - Supermicro
Register FAQ Calendar Search Today's Posts Mark Forums Read

 
Thread Tools Display Modes
Old 10-23-2011, 12:22 AM   #1
momaka
Badcaps Veteran
 
momaka's Avatar
 
Join Date: May 2008
City & State: VA (NoVA)
My Country: U.S.A.
Line Voltage: 120 VAC, 60 Hz
I'm a: Hobbyist Tech
Posts: 8,387
Default Need help with Shuttle SB83G5A - fans spin but that's all

Iíve been trying to fix this Shuttle SB83G5A computer for a while now. When it was given to me, the motherboard had 1 visibly bad cap on V_DIMM and all 4 bad caps on what I believe is the Northbridge/Southbridge VRM (3 on low side, 1 on high side). The bad caps were all OST. I replaced them with equivalent/superior Nichicon caps. The motherboard is a FB83 I think.

The problem is that the fans spin and nothing else happens.

What Iíve tried so far (with no further success):
- replaced a few more caps other than the obviously failed ones, but that didnít help.
- tried using 2 other known-good power supplies (one is a fully recapped 300W Macron Power, the other one is this) besides the original one, which has no visibly-bad caps.
- I reset the CMOS Ė in fact, after the first few tests, I completely removed the CMOS battery.
- tried the motherboard with and without RAM. The memory sticks Iíve used for the test are known good, and I am using them in other computers right now without a problem.

What seems to be working so far:
- V_DIMM voltage Ė it shows 2.5V stable, with and without RAM. Some of the other smaller linear voltage regulators also seem to be working fine.
- The Southbridge, since I can shut down and start the computer from the button successfully, as well as restart it.

Iíve measured some of the derived voltages on the motherboard. Thereís a number of them that arenít coming up (they jump up a bit then settle to a low value), including the CPUís V_CORE. Iíve checked all MOSFETs on these rails, and none are shorted.

I think the problem is with Northbridge/Southbridge VRM, since it shows 0.6V only and Iím almost certain 0.6V canít be high-enough for the NB and SB. My logic is that if the NB/SB voltage is not right, then perhaps the other voltages on the motherboard wonít be either. The only reason I suspect the NB/SB VRM is because the lower MOSFET *seems* to have a slight bubble on it as best as I can tell.

Iíve circled the NB/SB VRM with red. The lower MOSFET is connected between 3.3V rail of the PSU and NB/SB VRM. The upper MOSFET is connected between ground and NB/SB VRM. Caps with red dots on them are on NB/SB VRM low. Yellow Ė NB/SB VRM high. Green Ė V_DIMM 2.5v

In any case, my question is, does anyone know what that NB/SB voltage should be? Perhaps someone who has this motherboard can measure it and tell me if itís not too much of a hassle?

I would really like to get this PC fixed, since it has a PCI-E slot (a first for me). I like the small form factor of the case as well.

Also, if anyone has any other suggestions, please let me know. At this point Iím open to any and all ideas. In fact, Iíve very tempted to just insert 1.5V on that spot and see what happens.
Attached Images
File Type: jpg IMG_0343 (2).jpg (593.6 KB, 88 views)
momaka is offline   Reply With Quote
Old 10-23-2011, 12:55 AM   #2
retiredcaps
Badcaps Veteran
 
Join Date: Apr 2010
City & State: Canada
Posts: 9,063
Default Re: Need help with Shuttle SB83G5A - fans spin but that's all

Quote:
Originally Posted by momaka View Post
The lower MOSFET is connected between 3.3V rail of the PSU and NB/SB VRM. The upper MOSFET is connected between ground and NB/SB VRM.
1) You know more about this repair than I since I'm still learning, but what is the part number of the mosfet?

2) I assume all these tests were done with a cpu installed?
__________________
--- begin sig file ---

If you are new to this forum, we can help a lot more if you please post clear focused pictures (max resolution 2000x2000 and 2MB) of your boards using the manage attachments button so they are hosted here. Information and picture clarity compositions should look like this post.

We respectfully ask that you make some time and effort to read some of the guides available for basic troubleshooting. After you have read through them, then ask clarification questions or report your findings.

Please do not post inline and offsite as they slow down the loading of pages.

--- end sig file ---
retiredcaps is offline   Reply With Quote
Old 10-23-2011, 10:38 PM   #3
momaka
Badcaps Veteran
 
momaka's Avatar
 
Join Date: May 2008
City & State: VA (NoVA)
My Country: U.S.A.
Line Voltage: 120 VAC, 60 Hz
I'm a: Hobbyist Tech
Posts: 8,387
Default Re: Need help with Shuttle SB83G5A - fans spin but that's all

Thanks for the reply, retiredcaps.

Quote:
Originally Posted by retiredcaps View Post
1) You know more about this repair than I since I'm still learning, but what is the part number of the mosfet?
Good call. Don't know why I didn't think about looking it up.
Both MOSFETs on the NB/SB VRM are Alpha and Omega 4404 N-channel MOSFETs.
I found the data sheet. Says it's rated for 8.5A continuous, and 24 mOhms min Rds(on). Will look through my junk pile and see what I can find as I do have a few dead motherboards and video cards. I'm still very tempted to just insert 1.5V with an external source on the NB/SB VRM low side since replacing these MOSFETs just seems like a pain in the butt.

Quote:
Originally Posted by retiredcaps View Post
2) I assume all these tests were done with a cpu installed?
Correct.
momaka is offline   Reply With Quote
Old 10-23-2011, 11:19 PM   #4
retiredcaps
Badcaps Veteran
 
Join Date: Apr 2010
City & State: Canada
Posts: 9,063
Default Re: Need help with Shuttle SB83G5A - fans spin but that's all

Quote:
Originally Posted by momaka View Post
I think the problem is with Northbridge/Southbridge VRM, since it shows 0.6V only and Iím almost certain 0.6V canít be high-enough for the NB and SB.
I'm not sure if my understanding of how this works is correct or not, but the AO4404 datasheet says that the gate threshold voltage minimum is 0.7V, typical is 1.0V, and max is 1.4V.
retiredcaps is offline   Reply With Quote
Old 10-24-2011, 07:23 AM   #5
kc8adu
Super Moderator
 
kc8adu's Avatar
 
Join Date: Nov 2003
City & State: dayton ohio
My Country: U.S.A!
Line Voltage: 12vdc,120/240vac,480vac 3ph on my bench
I'm a: Professional Tech
Posts: 8,254
Default Re: Need help with Shuttle SB83G5A - fans spin but that's all

look at that vrm again.iirc 915 uses 1.5v core.could also be shorted.does nb or vrm get hot?
kc8adu is offline   Reply With Quote
Old 10-24-2011, 11:01 PM   #6
momaka
Badcaps Veteran
 
momaka's Avatar
 
Join Date: May 2008
City & State: VA (NoVA)
My Country: U.S.A.
Line Voltage: 120 VAC, 60 Hz
I'm a: Hobbyist Tech
Posts: 8,387
Default Re: Need help with Shuttle SB83G5A - fans spin but that's all

Quote:
Originally Posted by kc8adu
look at that vrm again.iirc 915 uses 1.5v core.could also be shorted.does nb or vrm get hot?
Correct, 915 uses 1.5v core. Nothing gets hot, though, and that's what worries me a bit. I measured resistance between the NB/SB VRM low and ground and the lowest I got was 40 Ohms. Switching the probes around gives higher readings. In any case, I can't get a short circuit across either of the MOSFETs.

Quote:
Originally Posted by retiredcaps View Post
I'm not sure if my understanding of how this works is correct or not, but the AO4404 datasheet says that the gate threshold voltage minimum is 0.7V, typical is 1.0V, and max is 1.4V.
Vgs(th) indicates just what voltage the MOSFET needs in order to turn on.

Good thing you brought it up, though. I measured the Gate voltage on both MOSFETs and here's what I got:
- the lower (supposedly bad) MOSFET that's connected between 3.3v rail and NB/SB VRM low has 1.4v on its Gate.
- the upper MOSFET that's connected between ground and NB/SB VRM low has 0.0v on its Gate.

So if I'm not mistaken, it looks like the PWM controller controlling these MOSFETs is trying to make the lower MOSFET conduct, but perhaps the lower MOSFET is open?
As for the 0.6v present on the NB/SB VRM, my only guess is that this is caused from leakage current from other components. Any thoughts on this?

If not, I guess I'll remove the lower MOSFET, test it out of circuit, and then proceed from there.

Last edited by momaka; 10-24-2011 at 11:05 PM..
momaka is offline   Reply With Quote
Old 11-11-2012, 09:54 PM   #7
momaka
Badcaps Veteran
 
momaka's Avatar
 
Join Date: May 2008
City & State: VA (NoVA)
My Country: U.S.A.
Line Voltage: 120 VAC, 60 Hz
I'm a: Hobbyist Tech
Posts: 8,387
Default Re: Need help with Shuttle SB83G5A - fans spin but that's all

It's really been a while since I last updated the info in this thread. I removed the lower 4404 MOSFET a long time ago and tested it out of circuit - there was a short circuit (6-8 Ohms actually) not only between Source and Drain of that MOSFET but also between Gate and Drain! Checking the resistance between the Gate pad on the motherboard for that MOSFET and ground revealed a short circuit too. So looks like the MOSFET took out the controller with it after it shorted. GREAT! The controller that's responsible for driving this buck circuit is a RichTek RT9203.

And now for some more bad news...
Before I even found the gate drive on the RT9203 controller for that MOSFET was shorted, I tried wiring up a 60N03 MOSFET in place of the dead 4404 to see if the NB voltage would come up. And in this process, I managed to rip the pad on the board for the Gate of that MOSFET. So now even if I do get a replacement RT9203 controller and 4404 MOSFET, I would still have to do some ghetto hacks to wire the Gate drive.

I've been eager to test that computer so I've also been working on a 1.5V DC-DC power supply. Finally made a design not too long ago that actually works well enough with more than 0.5A of load, but I think it's oscillating at high loads (2A and above). It's a linear setup with a single MOSFET and XRA10358 (equivalent to LM358) op-amp. I'll leave that for another thread, though. Maybe if I get that working, I'll use it instead of buying replacement MOSFET and controller. Not that the controller is that expensive or anything. Just eager to build my own working DC-DC low-voltage power supply. Right now, the circuit I have works fine with 0.5A of load or less. When I go above 1A, the voltage goes up to 1.6V - I'm not sure if the Intel 915G NB will like that. Finally, when I increase the load to 5A, the output drops to 1.42 to 1.48V (it's fluctuating all the time).

Last edited by momaka; 11-11-2012 at 10:00 PM..
momaka is offline   Reply With Quote
Old 01-01-2013, 04:57 PM   #8
momaka
Badcaps Veteran
 
momaka's Avatar
 
Join Date: May 2008
City & State: VA (NoVA)
My Country: U.S.A.
Line Voltage: 120 VAC, 60 Hz
I'm a: Hobbyist Tech
Posts: 8,387
Default Re: Need help with Shuttle SB83G5A - fans spin but that's all

And another update for this thread (if anyone cares, lol ).
.
.
I bought a replacement RT9203 controller from eBay. Since the gate trace to the upper MOSFET was ripped on the board, I decided I might as well just kludge it with a DPAK MOSFET (I have quite a few of them). Ended up using a 60N02 MOSFET (25V, 62A, 8.4 mOhm Rds_on). Pictures of the handy work are attached below . Yeah, I know... as usual. The braid soldered to the MOSFET's tab is its heat sink and holder, BTW .

Surprisingly when I powered it on, nothing smoked . However, the motherboard is still dead as a stone . I checked the voltage output of the NB buck regulator and it's putting out 1.44-1.45V. That seems pretty close to the required 1.5V for the NB, don't you think?

Also, now that the NB voltage rail is present, all of the other major voltage rails are present too. Before this, the CPU Vcore and RAM VTT were not present along with a few other voltage rails. So looks like there IS some progress after all.

CPU Vcore is at 1.57V, though. Isn't that a bit high? FYI, CPU is a 2.8 GHz Pentium 4 (socket 775, Prescott??). I just looked on CPU-World website and all Pentium 4 for the socket 775 are rated for up to 1.4V. So something looks wrong here I think . Caps on the CPU VRM low side seem fine though (there's 4x 6.3V 3300uF Ost RLS and 8x 4V 560uF Sanyo SEPC polymers). If anything, those polymers should be keeping everything stable. In addition, I added an extra 4V 820uF Fujitsu FP-CAP RE series, so I think there shouldn't be any excess ripple on the CPU VRM output.

Lastly, I tried a POST card. Doesn't look like the BIOS is trying to boot at all (POST card stays at "00"). Normally, the POST card will display some numbers as the BIOS is going through the POST process. CPU fan is still blowing on full blast (it's PWM controlled and should slow down if POST is successful).

Another abnormality I noticed is that the first few times I tried the computer after I put in the MOSFET and the RT9203 buck controller is that the POST card indicated that the PSU 3.3V rail was not present and the NB voltage was only showing 1.34V (unless I made an error measuring with my multimeter). I then connected the PCI-E auxiliary power connector on the motherboard (it's a 4 pin floppy connector) and on the next boot, the POST card indicated that the PSU 3.3V rail was present and the NB voltage rail came up to 1.45V. Ever since then, it doesn't matter if I have the PCI-E auxiliary power connected or not - PSU 3.3V rail on the POST card always shows up as present and the NB voltage rail has been fine since then too. Coincidence or weird occurrence?

So, any ideas? I believe I have tried pretty much everything on this motherboard. The only few things I could still try is a new CPU, replace the NB, and maybe reflash the BIOS or replace it altogether. Out of these 3, though, only replacing the CPU seems something I can do. I can't replace the i915 NB because I don't have another one (unless someone here has a dead board with one and doesn't mind pulling it off), and I don't have a BIOS flasher. So if anyone has any ideas, it'd love to hear them. I don't need this computer badly or anything, but it would be nice to have and also good learning experience (+ highly rewarding for my mind since I've done so much work on it ).
Attached Images
File Type: jpg Suttle XPC FB83 (7).jpg (316.5 KB, 62 views)
File Type: jpg Suttle XPC FB83 (8).jpg (273.7 KB, 52 views)

Last edited by momaka; 01-01-2013 at 05:06 PM..
momaka is offline   Reply With Quote
Old 01-04-2013, 12:23 PM   #9
Gabriel
Badcaps Veteran
 
Join Date: Jan 2010
City & State: Northern Norway
My Country: Norway
Line Voltage: 230VAC 50Hz
I'm a: Knowledge Seeker
Posts: 360
Default Re: Need help with Shuttle SB83G5A - fans spin but that's all

I really like your attention to details, wish I were 1/2 as good as you.

Let me know what plcc chip you have, I have a box of spare plcc chips and I can easily program for you and ship it out for free.

Looking forward for updates.
Gabriel is offline   Reply With Quote
Old 01-09-2013, 11:08 PM   #10
momaka
Badcaps Veteran
 
momaka's Avatar
 
Join Date: May 2008
City & State: VA (NoVA)
My Country: U.S.A.
Line Voltage: 120 VAC, 60 Hz
I'm a: Hobbyist Tech
Posts: 8,387
Default Re: Need help with Shuttle SB83G5A - fans spin but that's all

Thanks for the offer and kind comments, Gabriel. Before trying another BIOS chip, I think I have another problem I need to straighten out first, and that is the CPU voltage. I just did another test 5 minutes ago and CPU voltage is still at 1.56V - 1.57V. This is definitely not right! Thus I ended up recapping the 4x OST RLS 10V 3300uF caps with 6.3V 2700uF Rubycon MFZ (pulled from Xbox 360s... FYI these MFZ caps have much better specs than the OST RLS series). But the voltage was still the same.

I'm not sure what controls the CPU voltage, but I can only think of a few things that can affect it. Either NB is bad and not detecting the CPU properly, the CPU socket joints are cracked (I will admit that there were a few times where I wasn't very gentle with this motherboard), the CPU is bad, or the BIOS is corrupt and giving the wrong info to the CPU voltage controller.

Was going to get a cheap refurb P4 CPU at MicroCenter today, but they were out of stock. So the next option I think I will try is to do a CPU socket reflow. If that doesn't fix it either, then I might change the NB (if I find another one from a junk motherboard) or get a new BIOS chip. Unfortunately, I'm starting college again, so I don't know when I will have time to work on this. But as always, any suggestions on what to try are welcome.
momaka is offline   Reply With Quote
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump



Badcaps.net Technical Forums © 2003 - 2018
Powered by vBulletin ®
Copyright ©2000 - 2018, Jelsoft Enterprises Ltd.
All times are GMT -6. The time now is 02:52 AM.

Did you find this forum helpful?