|
2016-06-10, 01:46
|
#23 |
"David"
Jul 2015
Ohio
11×47 Posts |
I managed to partially repair one of my Titans with the same problem, however the other two remain damaged.
My experience has been that it is the mosfet responsible for powering the memory chips that blows up, hence why TF work never exposes the problem. |
|
|
|
2016-06-16, 19:38
|
#24 |
Feb 2012
1100101012 Posts |
This is the end, or is it?
The bottom MOSFET welded itself to the traces, no matter how hard I heated it up with an air gun at 450 degrees, it would not let go. I heated it so much, everything around the MOSFET got de-soldered.
Eventually the trace burned through with part of it still attached to the MOSFET. Both inductor coils are fine. The other MOSFET appears to be ok as well, though I did not fully test it. I carefully welded everything around back in place (those little condensers are a lot of fun), and I replaced both coils and the top MOSFET with new ones. There is no sane way to replace the bottom MOSFET---where the top part of it goes (with the narrow underside) the traces are gone. Does it make sense to try starting the board like that, with half of the power circuit is gone; what would happen? |
|
|
|
|
2016-06-17, 08:09
|
#25 | |
|
Romulan Interpreter
"name field"
Jun 2011
Thailand
41·251 Posts |
Quote:
For my future titans "yet to crash" (I can't imagine all will live longer than myself, hehe) I am thinking to a method to "polish" (like in sandpaper button and a drilling or milling machine) the mosfets out. They are not "welded" to the board themselves, but this board has a freaking good thermal dissipation. I wrote about this long ago in the hardware thread. Practically, in the second you lift the soldering iron out from the PCB, the melted tin freezes solid. This is also very risky if the tools are not adequate, you can easily create cold soldering and/or non-wetting soldering. |
|
|
|
|
|
2016-06-17, 08:17
|
#26 | |
|
Romulan Interpreter
"name field"
Jun 2011
Thailand
283316 Posts |
Quote:
 I missed these last posts... So, you have 3, plus 2 of mine plus one from TObject makes 6 "victims" of cudaLucas. This accentuates my initial supposition about a design problem there, and both air and water cards from EVGA are affected. Now, the question is, do we push it too hard? Because if not, than we may be able to get some "discounts" from EVGA for future purchases... 
|
|
|
|
|
|
2016-06-17, 17:52
|
#27 |
|
Feb 2012
34·5 Posts |
I am not kidding you---it welded itself to the trace. I unsoldered the other one, no problem; I pick them off with tweezers without taking the heat off. The bad one would not budge; I used so much heat, there is a crater now on the PCB where the bad one used to be; the smaller traces evaporated (or otherwise gone), and part of the bigger trace is still attached to the MOSFET (that trace burned through on the exposed part that leads to the quadra-hole via).
BTW, I tried to put the board in the computer; it is still shorting power. Probably the PCB is shorted. Last fiddled with by TObject on 2016-06-17 at 18:01 |
|
|
|
|
2019-10-18, 17:35
|
#28 |
|
Oct 2019
3 Posts |
I know this in an old thread. I didn't want to wake this old one, but I hoped to get the attention of people who put in their hardwork on resolving this matter. I will also try to private message LaurV after I put this here hoping to hear from him.
I have 2 titans which died last month. First one smokes and trips the psu and after few weeks the second one takes similar path. I thought I was at fault due to overheat, though these cards were on watercooling and temps were around 60 degrees max. Assuming I will have a mobile phone repair shop do replacement. What are the part numbers? I have the following info so far. 1. Mosfet - NTMFD4901NF (Since I have two cards I will order 4 + 2 extra) 2. Inductor R33 - ???? 3. Inductor 1R0 - ???? 4. Resistor 5M0 - ???? |
|
|
|
|
2019-10-18, 17:59
|
#29 |
|
Romulan Interpreter
"name field"
Jun 2011
Thailand
41·251 Posts |
I approved the post, and also got your PM.
The resistor is 5 milli-ohm, it is a shunt for measuring the current, I don't think it is broken, but if you measure it with an ohmmeter will show short, this is normal. The coils are 330uH and 1mH. (sorry for fast style, here is 1:00 AM and I am going to bed right now in few minutes, real life kept me busy till this "early" time in the morning, hehe) Last fiddled with by LaurV on 2019-10-18 at 18:01 |
|
|
|
|
2019-10-19, 17:48
|
#30 |
|
Oct 2019
38 Posts |
I am attaching the pictures of the Titan and a fully working 780. So someone can find it helpful.
I have 2 titans and the pictures posted is the one having the worst issue. The second one only has 1 inductor burnt out. I haven't posted the picture of that one. After hours of searching I found the parts or I think I found them. The issue is with the size of the package which means everything in this specific project. If possible can you please confirm the following parts are correct so I can order them? 1. Mosfet - Mosfet 2. Inductor R33 - R33 3. Inductor 1R0 - 1R0 4. Resistor 5M0 - 5M0 https://www.mersenneforum.org/attach...1&d=1571506168 https://www.mersenneforum.org/attach...1&d=1571506172 https://www.mersenneforum.org/attach...1&d=1571506172 https://www.mersenneforum.org/attach...1&d=1571506172 https://www.mersenneforum.org/attach...1&d=1571506381 |
|
|
|
 |
| Thread Tools | |
Similar Threads
|
||||
| Thread | Thread Starter | Forum | Replies | Last Post |
| Titan Black | ATH | Hardware | 15 | 2017-05-27 22:38 |
| Nvidia announces Titan Xp card | ixfd64 | GPU Computing | 10 | 2017-05-17 15:19 |
| Nvidia announces Titan X | ixfd64 | GPU Computing | 20 | 2015-04-28 00:27 |
| Geforce GTX Titan 6GB | ATH | GPU Computing | 295 | 2013-05-12 21:35 |
| 2x AMD 7990 or 2x Nvidia Titan ?? | Manpowre | GPU Computing | 27 | 2013-05-12 10:00 |
