Overclock.net › Forums › Volt Mods › General Volt Mods › Graphics Cards Volt Mods › Calling All Volt-Modders, 8800 GTS Volt Mod Gone Wrong, HELP
New Posts  All Forums:Forum Nav:

Calling All Volt-Modders, 8800 GTS Volt Mod Gone Wrong, HELP - Page 7

post #61 of 68
oh! and polarity for a coil doesn't matter its a coil no matter what way you hook it up. its Caps, Diodes, FETs/transistors, IC's you have to watch, thats when it gets important.
post #62 of 68
yeah, lead free solder can be a pain - infact any soldering can be a pain on premade boards - what I do if i have trouble melting the solder, is to Add some solder as I heat it, the fresh solder helps melt the existing stuff. smile.gif
post #63 of 68
Thread Starter 
Quote:
Originally Posted by latelesley View Post

yeah, lead free solder can be a pain - infact any soldering can be a pain on premade boards - what I do if i have trouble melting the solder, is to Add some solder as I heat it, the fresh solder helps melt the existing stuff. smile.gif

I tried that but the existing solder was so little and so well hidden underneath it wasn't working for me.
     
CPUMotherboardGraphicsRAM
Intel Core i5 2500k @ 4.9Ghz Gigabyte GA-Z77X-UP7 EVGA GTX 980Ti SC 4x 4GB Patriot Viper Xtreme 2133Mhz 
Hard DriveHard DriveCoolingOS
Intel 520 Series 120GB SSD WD Caviar Blue SATA II 640GB EK CPUWB & Rad, Temporary XSPC Pump & Res, All ... Windows 7 Home Premium 
MonitorKeyboardPowerCase
3x 1080p 24" Monitors Razer BlackWidow Ultimate Corsair HX-750 NZXT Lexa S (Highly Modified) 
MouseMouse PadAudio
Razer Naga 2014 Steelseries QCK Kenwood KA-5500 Reciever & (Temporary) ATH-M30X's 
CPUMotherboardGraphicsRAM
Intel Atom D525 2C/4T 1.8Ghz SuperMicro X7SPE-HF-D525 Integrated - Not Used 2x 2GB DDR3 Sodimms 
Hard DriveCoolingOSPower
4GB Sata DOM Passive (except for PSU Fan) Debian GNU/Linux 8.0 Jessie PicoPSU 120Watt with 60Watt EDAC Brick 
CaseOther
1U Supermicro DHCP/DNS/Munin 
  hide details  
Reply
     
CPUMotherboardGraphicsRAM
Intel Core i5 2500k @ 4.9Ghz Gigabyte GA-Z77X-UP7 EVGA GTX 980Ti SC 4x 4GB Patriot Viper Xtreme 2133Mhz 
Hard DriveHard DriveCoolingOS
Intel 520 Series 120GB SSD WD Caviar Blue SATA II 640GB EK CPUWB & Rad, Temporary XSPC Pump & Res, All ... Windows 7 Home Premium 
MonitorKeyboardPowerCase
3x 1080p 24" Monitors Razer BlackWidow Ultimate Corsair HX-750 NZXT Lexa S (Highly Modified) 
MouseMouse PadAudio
Razer Naga 2014 Steelseries QCK Kenwood KA-5500 Reciever & (Temporary) ATH-M30X's 
CPUMotherboardGraphicsRAM
Intel Atom D525 2C/4T 1.8Ghz SuperMicro X7SPE-HF-D525 Integrated - Not Used 2x 2GB DDR3 Sodimms 
Hard DriveCoolingOSPower
4GB Sata DOM Passive (except for PSU Fan) Debian GNU/Linux 8.0 Jessie PicoPSU 120Watt with 60Watt EDAC Brick 
CaseOther
1U Supermicro DHCP/DNS/Munin 
  hide details  
Reply
post #64 of 68
TBH it could be your soldering iron too - if its old, it might not be getting hot enough. All these lead-free solders have higher melting points, and need heated for longer to melt frown.gif

If you're gonna be soldering frequently, I need to make you a list of useful tools to get, for doing this stuff! smile.gif
post #65 of 68
Thread Starter 
Well thank you everyone for all of your help but this seems to be a lost cause.

With ALL of those things soldered back on I still recieve no core core voltage and no memory voltage and no post.

I'm gonna give up on this card, at this point its more trouble than its worth.
     
CPUMotherboardGraphicsRAM
Intel Core i5 2500k @ 4.9Ghz Gigabyte GA-Z77X-UP7 EVGA GTX 980Ti SC 4x 4GB Patriot Viper Xtreme 2133Mhz 
Hard DriveHard DriveCoolingOS
Intel 520 Series 120GB SSD WD Caviar Blue SATA II 640GB EK CPUWB & Rad, Temporary XSPC Pump & Res, All ... Windows 7 Home Premium 
MonitorKeyboardPowerCase
3x 1080p 24" Monitors Razer BlackWidow Ultimate Corsair HX-750 NZXT Lexa S (Highly Modified) 
MouseMouse PadAudio
Razer Naga 2014 Steelseries QCK Kenwood KA-5500 Reciever & (Temporary) ATH-M30X's 
CPUMotherboardGraphicsRAM
Intel Atom D525 2C/4T 1.8Ghz SuperMicro X7SPE-HF-D525 Integrated - Not Used 2x 2GB DDR3 Sodimms 
Hard DriveCoolingOSPower
4GB Sata DOM Passive (except for PSU Fan) Debian GNU/Linux 8.0 Jessie PicoPSU 120Watt with 60Watt EDAC Brick 
CaseOther
1U Supermicro DHCP/DNS/Munin 
  hide details  
Reply
     
CPUMotherboardGraphicsRAM
Intel Core i5 2500k @ 4.9Ghz Gigabyte GA-Z77X-UP7 EVGA GTX 980Ti SC 4x 4GB Patriot Viper Xtreme 2133Mhz 
Hard DriveHard DriveCoolingOS
Intel 520 Series 120GB SSD WD Caviar Blue SATA II 640GB EK CPUWB & Rad, Temporary XSPC Pump & Res, All ... Windows 7 Home Premium 
MonitorKeyboardPowerCase
3x 1080p 24" Monitors Razer BlackWidow Ultimate Corsair HX-750 NZXT Lexa S (Highly Modified) 
MouseMouse PadAudio
Razer Naga 2014 Steelseries QCK Kenwood KA-5500 Reciever & (Temporary) ATH-M30X's 
CPUMotherboardGraphicsRAM
Intel Atom D525 2C/4T 1.8Ghz SuperMicro X7SPE-HF-D525 Integrated - Not Used 2x 2GB DDR3 Sodimms 
Hard DriveCoolingOSPower
4GB Sata DOM Passive (except for PSU Fan) Debian GNU/Linux 8.0 Jessie PicoPSU 120Watt with 60Watt EDAC Brick 
CaseOther
1U Supermicro DHCP/DNS/Munin 
  hide details  
Reply
post #66 of 68
Quote:
Originally Posted by EpicAMDGamer View Post

Quote:
Originally Posted by latelesley View Post

no, because it doesn't quite work like you think....each phase would do the powering out of step with each other, they don't all switch at the same time. in each cycle, each phase is taking a turn a 3rd of the time. so your PWM is really just working 2/3rds of the time.

You can imagine it like a 3 cylinder engine in a car, and cylinder one is misfiring because its not getting the pulse to the spark plug. What you are taking about with wiring the FETS together, is just like linking the "dead" spark plug to one of the plugs on the other cylinders - it would get it to spark, but at the wrong time, and the engine would still run wrong. Its the same with the dead phase, you could get it switching the FETs, but it would be at the wrong time, and cause unstable voltage (probably with a higher ripple).

The normal cycle would fire phase 1, phase 2, phase 3, back to phase 1 etc in that cycle.

By bridging you'd have no phase, phase 1 & 2, phase 3, no phase etc etc

so all you'd do is increase the pulse on the second part of the cycle, as it fired two phases at once, and it would still have the dead spot on the first phase of the cycle. It would probably confuse the PWM controller too, and that could cause things to go funky on the voltage output.

I see...

..And have another question, what exactly are these things that I thought were inductors or coils? This one broke when i originally removed it and I don't think its an inductor or coil or that its even operational anymore:



And here's a bit of a better pic of the VRM's on the GPU side.

 

 

those are definitely inductors.. of the solid variety. Not the cheap 'iron ferrite' bar with copper coil wrap.. 

 

*the one in your pic is definitely damaged and not usable, even if readings are normal

*from the traces on the PCB it looks as if the 3x GPU phases are parallel to each other and not in series

*if the amperage rating of each phase is too low.. the GPU will not be able to function as the single phase is be in constant state of OCP [with only 1x phase working]

*the slave IC's that control each phase, are controlled by the main IC.. which you pointed out has a damaged component. 

*from the readings you are showing, only a single phase is operational..

*to fix it; you will need to repair the damaged components and double check that the main IC is still programmed/set for 3x phase operation [you can do this with the DS, by reading on which pins control phasing operation.. many times the # of phases is programmed just like VID table]

 

 Hope that helps man. Sorry, Ive been delayed with replies.. working way too much.

post #67 of 68
Quote:
Originally Posted by latelesley View Post

no, because it doesn't quite work like you think....each phase would do the powering out of step with each other, they don't all switch at the same time. in each cycle, each phase is taking a turn a 3rd of the time. so your PWM is really just working 2/3rds of the time.

You can imagine it like a 3 cylinder engine in a car, and cylinder one is misfiring because its not getting the pulse to the spark plug. What you are taking about with wiring the FETS together, is just like linking the "dead" spark plug to one of the plugs on the other cylinders - it would get it to spark, but at the wrong time, and the engine would still run wrong. Its the same with the dead phase, you could get it switching the FETs, but it would be at the wrong time, and cause unstable voltage (probably with a higher ripple).

The normal cycle would fire phase 1, phase 2, phase 3, back to phase 1 etc in that cycle.

By bridging you'd have no phase, phase 1 & 2, phase 3, no phase etc etc

so all you'd do is increase the pulse on the second part of the cycle, as it fired two phases at once, and it would still have the dead spot on the first phase of the cycle. It would probably confuse the PWM controller too, and that could cause things to go funky on the voltage output.

 

sorry for double post.. but +rep to the above. Very good analogy of phasing operation. 

post #68 of 68
Quote:
Originally Posted by EpicAMDGamer View Post

Well thank you everyone for all of your help but this seems to be a lost cause.

With ALL of those things soldered back on I still recieve no core core voltage and no memory voltage and no post.

I'm gonna give up on this card, at this point its more trouble than its worth.

Man keep trying, you've got some good information in this thread to work off of. Scavenge an inductor off a bricked component and start replacing the things that got toasted. At the very least you are gaining valuable soldering skills which is what you need to avoid going through this again. Pay attention to values on the caps and resistors as they are important, as well as the polarity.
EVOLVE
(13 items)
 
  
CPUMotherboardGraphicsRAM
|i5 750-voltage may vary| |Foxconn Katana[Non GTI]| 380 4Gb |4Gb/Tridents@Clocked| 
Hard DriveOSMonitorPower
|[V3/60Gb][500Gb]| |Windows 7 64 bit| |LG-EW224| |Antec HCG750| 
Case
|Tech Station| 
  hide details  
Reply
EVOLVE
(13 items)
 
  
CPUMotherboardGraphicsRAM
|i5 750-voltage may vary| |Foxconn Katana[Non GTI]| 380 4Gb |4Gb/Tridents@Clocked| 
Hard DriveOSMonitorPower
|[V3/60Gb][500Gb]| |Windows 7 64 bit| |LG-EW224| |Antec HCG750| 
Case
|Tech Station| 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Graphics Cards Volt Mods
Overclock.net › Forums › Volt Mods › General Volt Mods › Graphics Cards Volt Mods › Calling All Volt-Modders, 8800 GTS Volt Mod Gone Wrong, HELP